Article

Cite This: ACS Nano 2018, 12, 5580−5587 www.acsnano.org

Dramatic Increase in Polymer Glass Transition

Temperature under Extreme
Nanoconfinement in Weakly Interacting
Nanoparticle Films
Haonan Wang,† Jyo Lyn Hor,‡ Yue Zhang,† Tianyi Liu,† Daeyeon Lee,*,‡ and Zahra Fakhraai*,†
†
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Department of Chemistry, University of Pennsylvania, Philadelphia, Pennsylvania 19104, United States

‡
Department of Chemical and Biomolecular Engineering, University of Pennsylvania, Philadelphia, Pennsylvania 19104, United States
Downloaded via UNIV OF PENNSYLVANIA on December 24, 2020 at 16:58:51 (UTC).

*
S Supporting Information

ABSTRACT: Properties of polymers in polymer nano-

composites and nanopores have been shown to deviate
from their respective bulk properties due to physical
confinement as well as polymer−particle interfacial
interactions. However, separating the confinement effects
from the interfacial effects under extreme nanoconfinement
is experimentally challenging. Capillary rise infiltration
enables polymer infiltration into nanoparticle (NP)
packings, thereby confining polymers within extremely
small pores and dramatically increasing the interfacial
area, providing a good system to systematically distinguish
the role of each effect on polymer properties. In this study,
we investigate the effect of spatial confinement on the glass
transition temperature (Tg) of polystyrene (PS) infiltrated into SiO2 NP films. The degree of confinement is tuned by
varying the molecular weight of polymers, the size of NPs (diameters between 11 and 100 nm, producing 3−30 nm average
pore sizes), and the fill-fraction of PS in the NP films. We show that in these dense NP packings the Tg of confined PS,
which interacts weakly with SiO2 NPs, significantly increases with decreasing pore size such that for the two molecular
weights of PS studied the Tg increases by up to 50 K in 11 nm NP packings, while Tg is close to the bulk Tg in 100 nm NP
packings. Interestingly, as the fill-fraction of PS is decreased, resulting in the accumulation of the polymer in the contacts
between nanoparticles, hence an increased specific interfacial area, the Tg further increases relative to the fully filled films
by another 5−8 K, indicating the strong role of geometrical confinement as opposed to the interfacial effects on the
measured Tg values.
KEYWORDS: capillary rise infiltration, glass transition, nanoporous film, ellipsometry, nanoconfinement, polymer glass,
polymer nanocomposite

hin polymer films and polymer nanoparticle (NP) enhanced mobility,14,20,26 which can result in lower average

T composites have become increasingly important in

modern technology, such as microelectronics,1 separa-
tion membranes,2 protective coatings,3 and photovoltaic cells.4
At the nanometer scale, the reduction in size of the polymers in
these systems results in a deviation of physical properties, such
as the physical aging rate,5−9 viscosity,10 and the glass transition
temperature (Tg),11−16 from their bulk values. These deviations
from bulk properties could be due to either chain and
segmental confinement effects17,18 or the increased role of
interfacial effects on the properties of the system.16,19,20 In
particular, Tg values of supported polymer thin films have been
shown to be influenced by both the substrate and the free
surface.12,21−25 The free surfaces of polymer films show
film Tgs. The effects are quite significant when the film
thickness is reduced below ∼30 nm in both polymer and small-
molecule glass systems.11,14,23,27−30 When the polymer has
strong interactions with the substrate (such as hydrogen
bonding), the substrate−polymer interactions can dominate
and the Tg increases with decreasing film thickness.22,23,31,32
The competing interfacial effects can even give rise to two
distinct Tgs for ultrathin films at low cooling rates.33
Received: February 19, 2018
Accepted: May 24, 2018
Published: May 24, 2018

© 2018 American Chemical Society 5580 DOI: 10.1021/acsnano.8b01341

ACS Nano 2018, 12, 5580−5587
ACS Nano
Figure 1. Schematic illustration of sample preparation of (a) NP-top CaRI, (b) NP-top UCaRI, and (c) PS-top CaRI films.
Polystyrene (PS) on a silicon substrate with native oxide (Si/
SiO2) is one of the most investigated systems with weak
polymer−substrate interactions. Decreasing Tg of PS with film
thickness was observed in the early 1990s.12 Since then, a large
number of both experimental and theoretical studies have
focused on the confinement and interfacial effects on PS glassy
dynamics.12,28,34−40 In supported PS films, Tg typically
decreases with decreasing film thickness due to strongly
enhanced dynamics at the free surface. After the removal of
the free surface, Tg recovers to the bulk value.17,28 As such,
most studies have attributed the observation of Tg changes in
supported PS thin films to the free surface effects that are the
dominant factor in these films. Even the formation of highly
adsorbed PS layers on a native silicon oxide substrate has a
moderate effect on Tg in this geometry.41−43
Distinguishing interfacial effects from pure confinement
effects is even more challenging when it comes to PS/SiO2
nanocomposites. Increase,44 decrease,45 or no change46 in
composite Tg with increasing SiO2 NP loading have all been
observed. In these systems, the highest possible NP loadings are
around 40% volume fraction,44,45,47,48 limiting the degree of
confinement. Furthermore, the dispersion conditions and
possible aggregation of SiO2 NPs can complicate data
interpretation.
Recently, a new method of making polymer NP films with
extremely high NP loadings has been reported via capillary rise
infiltration (CaRI).49 In this method, an NP film is placed on
top of a polymer film. During annealing at high temperature,
capillary forces drive the polymer to infiltrate into the
interstices between the NPs without disturbing the dense
(∼63% volume fraction) NP film packing. The viscosity and
dynamics of unentangled polymers during infiltration have been
studied using experiments and molecular dynamics simulations
and have shown an extreme slow-down of the dynamics under
nanoconfinement.49−51 We have recently shown that the
viscosity increase in unentangled systems is independent of
the polymer/NP interactions and correlates well with the
increased Tg.51
Here, we studied the glass transition temperature of PS/SiO2
CaRI films with various NP diameters, PS molecular weights,
and PS fill-fractions in undersaturated CaRI (UCaRI)52 films.
Spectroscopic ellipsometry was used to measure the thickness
5581
Article
and the index of refraction of CaRI films as a function of
temperature to determine Tg of PS confined in these films
(Tg(confined)). The results show that under these extreme
confinement conditions, Tg(confined) is significantly increased
for CaRI films with NP diameters smaller than 100 nm. For
UCaRI films, where there is insufficient PS to fully fill the
interparticle space, PS forms capillary bridges between NPs and
resides in the most confined regions, and Tg is further
increased, despite having a larger free surface area. The
dewetting of PS on the NPs under these conditions indicates
that the interfacial effects are still minimal in these systems, and
the Tg increase can be strongly attributed to the geometrical
confinement effects due to modifications in segmental
relaxation dynamics in small spaces.
RESULTS AND DISCUSSION
The sample preparation schemes for CaRI and UCaRI films in
various geometries (PS-top or NP-top) are schematically
shown in Figure 1. More details of sample preparation can be
found in the Methods section and in the Supporting
Information (SI). Tgs of these films were determined via
spectroscopic ellipsometry (SE) upon cooling the films at 10
K/min. This is the nominal cooling rate typically used in
differential scanning calorimetry measurements and typically
corresponds to a segmental relaxation time of ∼102 seconds for
bulk polymers.53−58 The thickness of the NP layer, in all three
geometries shown in Figure 1, was found to be insensitive to
the temperature changes and was fixed at a constant initial value
to avoid overfitting (Figure S5 of SI). This is consistent with
the previous observations that, once the NP film is formed, it
does not expand upon polymer infiltration. This is due to
strong interactions between the NPs.49 As such, the changes in
the index of refraction (n) vs temperature (T) were monitored
to measure Tg(confined) (Figure 2).
Figure 2 shows an example of the calculated thickness and
refractive index change with temperature for a PS(8K)/SiO2(11
nm) NP-top, CaRI sample, with hPS = 242 ± 1 nm and hNP =
204 ± 1 nm. The Tg of each layer was determined as the
intersection of the linear fitting of the glass and the supercooled
liquid (SCL) regions in the plot of n vs T. The Tgs of the PS
and NP films were determined to be Tg(PS) = 362 ± 3 K and
Tg(confined) = 403 ± 4 K, respectively. We note that because
DOI: 10.1021/acsnano.8b01341
ACS Nano 2018, 12, 5580−5587
ACS Nano
Figure 2. (a) Index of refraction n (at λ = 632.8 nm) vs T for the PS
(black) and NP (red) films of a PS(8K)/SiO2(11 nm) NP-top CaRI
film. The inset shows the geometry of the sample. The Tgs of these
layers are measured to be Tg(PS) = 362 ± 3 K and Tg(confined) =
403 ± 4 K, respectively. (b) Normalized thickness vs temperature
for PS (black) and NP (red) films. The thicknesses of these layers
were measured to be hPS = 242 ± 1 nm and hNP = 204 ± 1 nm,
respectively. hNP was held constant during fitting. The inset shows
the mean-squared error (MSE) vs T, which shows reliable fitting
throughout this temperature range. Two inflection points are
observed in the hPS vs T plot. The first inflection point occurs at the
Tg(confined) upon cooling, where material is no longer taken up
into the NP film at lower temperatures.
of water absorption, there is a deviation from linear behavior of
the refractive index of the NP layer at low temperatures.
Therefore, only the linear region was fitted, as shown in Figure
2a.
In Figure 2b two inflection points can be observed at 363 ± 2
and 401 ± 7 K in the plot of hPS vs T. These values are the same
as Tg(PS) and Tg(confined) within error, as measured by the
index of refraction, respectively. This is because as the film is
cooled towards Tg(confined) from above (T > 403 K), the PS
chains in both layers are in the SCL state and can further
infiltrate into the NP film, while hNP remains constant due to
the rigidity of the nanoparticle film. This generates an apparent
expansion coefficient in the PS layer that is larger than the
expansion coefficient of the SCL of bulk PS. Below
Tg(confined) (T < 403 K), the PS layer recovers the expansion
coefficient of bulk material and goes through a glass transition
upon further cooling below Tg(PS).
Measurements of both the index of refraction and the film
thickness as shown in Figure 2 indicate an increase of ∼41 K in
Tg of PS(8K) upon confinement in SiO2(11 nm) NP films. In
this system the average pore diameter is estimated to be ∼3
nm,59 and the polymer radius of gyration is roughly 2.5 nm.60
To investigate the changes of Tg (ΔTg) upon further
confinement, CaRI films with various NP sizes were tested,
with PS of two molecular weights. Figure 3 shows the Tg of
PS(8K) and PS(2M), NP-top CaRI films, in NP packings with
various sizes (11−100 nm). In the original CaRI geometry
(NP-top), Tg values are fairly similar to the corresponding bulk
values in SiO2(100 nm) NP films, increasing dramatically with
decreasing NP size for both molecular weights, up to a ΔTg =
5582
Article
Figure 3. Tg as a function of NP diameter for CaRI films at two
different geometries (NP-top with open symbols and PS-top with
solid symbols), with PS with two molecular weights, PS(8K)
(black) and PS(2M) (red). Black and red dashed lines show the
bulk Tg values of PS(8K) (Tg = 362 ± 2 K) and PS(2M) (Tg = 373
± 2 K), respectively, as measured for thick PS films. Error bars
represent the standard error of at least three samples. Thicknesses
of the NP and the residual PS layers are hNP ≈ 250 nm and hPS ≈
200 nm, respectively.
41 ± 3 K increase for PS(8K)/SiO2(11 nm) and a ΔTg = 23 ±
3 K increase for PS(2M)/SiO2(11 nm). However, we note that
in this geometry the polymer and NP films have mismatching
expansion coefficients upon heating and the polymer infiltration
results in the formation of cracks in the NP films during
infiltration (see Figure S6). As such, there are some free
polymers in the cracked areas and the measured Tg(confined) is
a weighted average between the confined Tg and the Tg of PS in
the cracked regions. In contrast, in the PS-top geometry where
cracks are absent, the measured Tg(confined) is dramatically
larger, up to ΔTg = 51 ± 3 K for SiO2(11 nm)/PS(8K) and
ΔTg = 57 ± 3 K for SiO2(11 nm)/PS(2M).
The observations of up to an ∼57 K increase in Tg in CaRI
films (for SiO2(11 nm)/PS(2M) PS-top sample) are quite
remarkable. SiO2 is typically considered as a substrate with
weak interactions with PS. As such, in previous studies, the free
surface effects were considered as the dominant factor in the
observed Tg reduction of PS thin films. In these studies after the
free surface was removed, bulk Tg was recovered, but still no
increase in Tg for films as thin as ∼7 nm was observed.28 The
conventional method of making polymer nanocomposites is by
evaporating the solvent of a polymer−NP mixture suspension.
In these types of nanocomposite systems, low accessible NP
loadings and the tendency of NPs to aggregate can reduce the
effect of interfacial area and the spatial confinement. The force
generated by extremely confined chains precludes NPs from
creating regions of significant confinement, unless the polymer
chain is depleted from that region. As a result, Tg of PS either is
reduced or is within a few degrees from the bulk value in
conventional nanocomposites.44−46
To put the degree of CaRI film confinement in context, the
thinnest films studied in Tg measurements are typically 5−10
nm. PS in anodic aluminum oxide (AAO) templates can be
confined down to a diameter of 55 nm.61 In CaRI films, NPs
are closely packed with strong interparticle interactions before
the infiltration process. The space occupied by NPs is ∼63% of
the CaRI film volume (SI Table S1), which is very close to
random close packing of uniform spheres (64%). As such, the
PS/SiO2 interfacial area and spatial confinement is significantly
larger than achievable by co-mixing of the polymer and NPs. By
DOI: 10.1021/acsnano.8b01341
ACS Nano 2018, 12, 5580−5587
ACS Nano
changing the diameter of NPs from 100 nm to 11 nm, the
average pore diameter can be varied from 30 nm down to 3 nm
(calculated as ∼0.3 of the NP diameter59), with much smaller
areas close to the NP junctions. Moreover, there is almost no
free surface in these films, and virtually all of the polymer chains
are close to a SiO2 surface.
There are a few possible origins for the dramatic increase of
Tg in these highly confined CaRI films. Interfacial effects at the
SiO2/PS boundary, geometrical confinement, and a change in
the boundary conditions from isobaric (P,N,T) to isochoric
(V,N,T) can result in changes in the properties of PS in these
highly confined systems.
We can first rule out the effect of the boundary conditions.
As shown in Figure 2b, PS can freely expand in and out of the
film upon heating/cooling at temperatures above Tg(confined).
This effect is observed as an apparent change in the expansion
coefficient of the adjacent residual PS layer. As such, it is
unlikely that the boundary conditions are isochoric (V,N,T)
during these experiments. Similarly, in the NP-top geometry,
PS can freely expand to the free surface, as well as the bottom
PS surface, yet significantly increased Tg values are still
observed. This is consistent with previous studies where it
was observed that isochoric confinement for PS nanoparticles,
of 100 nm diameter or larger, does not significantly affect Tg.17
Another possibility is the formation of a highly adsorbed PS
layer during the infiltration/annealing process. Previous studies
have shown that during annealing at elevated temperatures,
large molecular weight PS can be irreversibly adsorbed on a
SiO2 surface with slower dynamics. These studies have shown
that the Tg of the interfacial layer is below the bulk Tg before
adsorption and gradually recovers to the bulk Tg with the
growth of the irreversibly adsorbed layer.43,62 The irreversibly
adsorbed PS layer takes hours of annealing to grow at the
infiltration temperature of 433 K used in this study.63 Thus, by
controlling the annealing time, we are able to examine the
potential influence of the adsorbed layer on Tg(confined)
values. Figures S9 and S10 show that extended annealing of the
samples (3 h at 433 K) does not significantly affect the value of
Tg, and thus there is no visible evidence that the growth of a
highly adsorbed layer has an influence on the observed values of
Tg(confined) for either PS(8K) and PS(2M) samples.
An interfacial layer with slower dynamics is also expected to
broaden the width of the glass transition64,65 or even result in
two glass transitions.33,64−67 In the CaRI samples, the width of
the glass transition can be calculated from the first derivative of
the change of the index of refraction with temperature. Figure
S11 shows T+ and T−, the high and low onsets of the glass
transition, as a function of NP diameter for CaRI films. It is
observed that there is no significant difference in the transition
width (ΔT = T+ − T−) between confined PS and bulk PS,
which indicates that the range of dynamical heterogeneity of PS
inside the NP packings is similar to the bulk within our ability
to resolve.
To further confirm whether spatial or interfacial effects are
responsible for Tg changes, a series of undersaturated (UCaRI),
NP-top films (Figure 1b) of PS(8K)/SiO2(25 nm) with various
PS fill-fractions were studied.52 In this geometry, free surfaces
are introduced due to underfilling. Unlike the porous glass
systems used in the past,66 underfilling of NP films forces the
polymer to be confined to the narrowest regions of the pores
due to capillary effects and at the same time have free surfaces.
As such, one can differentiate the role of interface/free surface
vs confinement on the measured Tg(confined). If interfacial
5583
Article
effects are dominant, one would expect the Tg(confined) to
decrease. In contrast, because of the dewetting on the SiO2
surface, PS is expected to form capillary bridges and occupy
more constricting regions in the interstices of the NPs, further
spatially confining the chains, resulting in increased
Tg(confined). This is because, while the PS chains can
eventually fill all the pores of 100 nm NPs with enough
material, they fill the smaller pores (i.e., the regions of higher
curvature) first and form capillary bridges between SiO2 NPs.52
The inset of Figure 4 shows scanning electron miscroscopy
Figure 4. Tg(confined) at a cooling rate of 10 K/min vs fill-fraction
for PS(8K)/SiO2(25 nm), NP-top UCaRI films with ∼200 nm
thickness. The fill-fraction is defined as the ratio of PS volume in
UCaRI vs the PS volume in fully filled CaRI NP films. Error bars
represent average ±1 standard deviation of linear fits for three
separate cooling ramps. The insets show the schematic of the
UCaRI sample and the SEM image of a capillary bridge observed in
a PS(8K)/SiO2(100 nm) NP-top UCaRI film (∼50% PS fill-
fraction). Scale bar is 100 nm.
(SEM) images of the capillary bridges formed for a PS(8K)/
SiO2(100 nm) NP-top UCaRI film with ∼50% PS fill-fraction.
Here, the fill-fraction is defined as the ratio of PS content
compared to the PS content of the fully filled CaRI films. Thus,
100% fill-fraction is analogous to 37% total PS volume in the
NP film. Larger SEM images of both CaRI and UCaRI films
can be viewed in Figure S8. Figure 4 shows Tg values of UCaRI
films with various fill-fractions. As the fill-fraction is decreased,
Tg(confined) further increases. This observation indicates that
the stronger geometric confinement has a much stronger effect
than the increased free surface in changing the Tg values in
CaRI films. The T+ and T− values for UCaRI samples of two
UCaRI samples are also shown in Figure S11. It is observed
that both T+ and T− shift uniformly with decreasing PS fill-
fraction within error, and the width of the transition remains
constant, consistent with the negligible effect of free surfaces,
which would have broadened the width of the transition.16,68
Geometric confinement is thus the most likely explanation
for the dramatically increased Tg(confined) values. Considering
the configurational entropy theory of glass transition,69,70 Tg
should increase when the material is confined in small pores, as
the confinement decreases the configurational entropy.71
Previous simulations72−77 have shown that hard walls can
slow the dynamics close to interfaces, thus increasing the local
Tg. Geometric curvature of the hard wall was also considered as
an important factor that can significantly increase the Tg61 and
even induce ordering of bonds and affect the glassy structure
near the substrate.13,78 However, it is worth noting that in
previous studies of the thermal glass transition of molecules
DOI: 10.1021/acsnano.8b01341
ACS Nano 2018, 12, 5580−5587
ACS Nano
confined in controlled pore glass (CPG) or AAO templates by
differential scanning calorimetry (DSC), decreasing Tg with
decreasing pore size was observed.35,55,71,79,80 Some studies also
observed two Tgs, one below the bulk Tg (due to the “size
effect”) and the other above the bulk Tg value,66,67,81 where the
higher Tg was attributed to the adsorbed layer on the surface of
the pores. In these studies, there is very weak pore size
dependence on the higher Tg value,66 which is qualitatively
different from the data reported here. We note that these DSC
measurements were performed upon heating. The out-of-
equilibrium glass can show lower Tg due to stress induced by
thermal expansion mismatch82 and potentially reduced density,
which could result in reduction and broadening of the Tg
transition.65 In the current study, we measure the glass
transition upon cooling from the equilibrium state, and thus
thermal expansion mismatch does not affect the results as the
polymer can move out of the NP film as observed in Figure 2.
We have also ruled out the Tg broadening as shown in Figure
S11. As such, the strong increase in Tg under isobaric
conditions can be attributed to the entropic confinement
effects. Since the glass transition is primarily due to the arrest of
segmental dynamics in polymers, the significant increase in Tg
transition indicates significantly slower segmental dynamics in
these systems, which is consistent with our previous measure-
ments of increased viscosity in unentangled polymers in
CaRI.51 Future studies that directly measure the relaxation
dynamics in these systems can elucidate the exact nature of the
confinement effects in these systems.
In summary, polymers confined in NP packings using CaRI
are great model systems to systematically study the role of
confinement effects vs interfacial effects on the glass transition
temperature values measured upon cooling under isobaric
conditions. In this study, we demonstrated significantly
increased Tg values of PS in densely packed SiO2 NP films,
with weak interfacial interactions. Tg(confined) of PS increases
with decreasing pore size (increasing confinement), as well as
by underfilling the polymer in UCaRI films. In this weakly
interacting system, extreme geometric confinement presents a
larger effect on the segmental relaxation dynamics and therefore
Tg than the free surface effects that can act to reduce Tg. The
dramatic increase in Tg has potential impact in applications
where nanocomposites are used at elevated temperatures, but a
glassy behavior is required.
METHODS
Preparation of PS/SiO2 CaRI Films. One-sided-polished silicon
wafers (100) purchased from Virginia Semiconductor were cut to
approximately 1 cm × 1 cm squares. Polystyrene (Mn = 7.5 kg/mol,
PDI = 1.06 (PS(8K)), and 1900 kg/mol, PDI = 1.18 (PS(2M)),
Polymer Source, Inc.) solutions were prepared by dissolving PS in
toluene. SiO2(25 nm) NPs were purchased dispersed in water (Ludox
TM-50, Sigma-Aldrich). SiO2(11 nm) and SiO2(100 nm) NPs were
purchased dispersed in isopropyl alcohol (IPA) (free samples from
Nissan Chemical). The suspensions were further diluted with their
respective solvents. No other treatment was applied to the NPs.
To generate the nanoporous CaRI films, a PS layer was first spin-
coated (Laurell, WS-400BZ-6NPP/Lite spin-coater) onto the silicon
wafer. Then the PS film was room-air plasma-treated for ∼2 s
(Expanded Plasma Cleaner PDC-001, Harric Plasma) to render the
film surface hydrophilic (without any effects on polymer properties as
shown in SI Figure S9). For SiO2 in IPA, this latter step was not
necessary. The SiO2 NP suspensions were then spin-coated onto the
PS layer to form a bilayer film. The bilayer films were then annealed at
423−453 K to induce PS infiltration and homogeneous distribution
throughout the voids of the SiO2 NP packings. These films are
5584
Article
designated as NP-top CaRI films (Figure 1a). In some experiments, in
order to prevent crack formation, an inverted geometry was used,
where the NP film was spin-coated first, followed by the PS film. These
samples are designated as PS-top CaRI films (Figure 1c). Fully filled
CaRI films are prepared such that they have a residual PS layer after
full infiltration. CaRI films were made with 11, 25, and 100 nm
diameter SiO2 NPs in both PS-top and NP-top geometries.
The porosity of fully filled NP films was calculated based on the
thickness change of the PS layer after infiltration and was determined
to be 37 ± 3% (see Table S1). The average pore size was calculated to
be 0.3d, where d is the diameter of the NPs.59 As such, the volume
occupied by the NPs is 63%.
Undersaturated CaRI films (UCaRI) were prepared where the
initial thickness of the PS layer was varied such that the desired fill-
fraction could be achieved after all of the polymer was infiltrated in the
system, leaving no residual PS layer under the NP film (Figure 1b).
The PS fill-fraction, the ratio of PS volume compared to the fully filled
PS volume, was defined as (hPS)/(37%hSiO2), where hPS is the PS-layer
thickness before infiltration and hSiO2 is the thickness of the NP layer.
However, due to the formation of cracks in the NP-top geometry, this
method is subject to error. Alternatively, the fill-fraction can be
obtained from the index of refraction (Figure S4). Figure S4 shows a
strong correlation between PS fill-fraction, calculated from the initial
PS thickness, and the refractive index of the NP film after infiltration,
which is consistent with previous observations.83 Using a linear
regression, we can obtain the fill-fraction from n. This method is used
for the data shown in Figure 4.
UCaRI samples were prepared with 25 nm SiO2 NPs and NP-top
geometry only, except for SEM images, where a 100 nm diameter was
used to highlight capillary bridges. For UCaRI films with fill-fractions
of >70%, the assumption was made that the polymer is uniformly
distributed in the NP films and a single-layer model with uniform
index was used in ellipsometry measurements. The model is compared
with an alternative model with a gradient in the refractive index, and it
is observed that for high filling fractions the uniform index assumption
is reasonable (see Figure S7a). Lower fill-fractions where this
assumption fails have not been included in this study (more details
in the SI and Figure S7b).
Ellipsometry Tg Measurements. The thickness and refractive
index values were measured using spectroscopic ellipsometry (M-2000
V, J.A. Woollam). The raw ellipsometry data Ψ(λ) and Δ(λ) were fit
to a Cauchy model (n(λ) = A + B/λ2, k(λ) = 0), where n and k are the
real and imaginary parts of the index of refraction (Figure S2).
Depending on the sample, either a two-layer (for CaRI) or one-layer
(UCaRI) Cauchy model on the Si substrate with 1 nm native oxides
was used. Optical constants of the Si substrate and the native oxide
layer were taken from a previous publication.84 The properties of both
the PS residual layer and the NP films were modeled to fit the
ellipsometry data. Different models were compared to make sure all
the key parameters were included without overfitting of the insensitive
ones (Figure S1). Figure S2 shows typical SE data and the results of
the model used to fit the data. Samples were mounted onto a
temperature-controlled stage (Linkam THMS 600) that was attached
to the ellipsometer. The ellipsometry sampling rate was 1 s with high-
accuracy zone averaging. Heating and cooling ramps under dry
nitrogen flow were run for each sample (Figure S5). The cooling rate
was held at 10 K/min. All Tg measurements are reported upon cooling.
Scanning Electron Microscopy. SEM images were taken using a
JEOL 7500F HRSEM at University of Pennsylvania Materials
Research Science & Engineering Center (MRSEC). Before imaging,
each sample was sputter-coated with a thin gold/palladium layer using
a Cressington sputter-coater 108 to prevent charging. The samples
were imaged at an accelerating voltage of 5.0 kV, an emission current
of 20 μA, and a working distance of approximately 8 mm.
ASSOCIATED CONTENT
*
S Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acsnano.8b01341.
DOI: 10.1021/acsnano.8b01341
ACS Nano 2018, 12, 5580−5587
ACS Nano
Additional ellipsometry parameters and model selection
for different samples; typical SE data and model;
calculation of porosity of the NP films; characterization
of NP-top UCaRI films and PS-top CaRI films; images of
cracks of CaRI films; SEM images of CaRI and UCaRI
films; evaluation of the effect of plasma treatment on
properties (PDF)
AUTHOR INFORMATION
Corresponding Authors
*E-mail: daeyeon@seas.upenn.edu.
*E-mail: fakhraai@sas.upenn.edu.
ORCID
Haonan Wang: 0000-0003-2047-5380
Tianyi Liu: 0000-0002-4282-3254
Daeyeon Lee: 0000-0001-6679-290X
Zahra Fakhraai: 0000-0002-0597-9882
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was financially supported by an NSF CAREER grant
(DMR-1350044, Z.F. and H.W.) and a University of
Pennsylvania Materials Research Science and Engineering
Center (MRSEC) grant (NSF DMR-1720530) and partially
funded by NSF-1449337 (D.L. and J.L.H.) and Corning
Research and Development Corporation (CRDC) (H.W.). We
thank J. Feng and P. Mazumder from CRDC for helpful
discussions. MRSEC facilities were used for SEM imaging
(NSF DMR-1720530).
REFERENCES
(1) Noh, Y.-Y.; Zhao, N.; Caironi, M.; Sirringhaus, H. Downscaling of
Self-aligned, All-printed Polymer Thin-film Transistors. Nat. Nano-
technol. 2007, 2, 784−789.
(2) Yave, W.; Car, A.; Wind, J.; Peinemann, K.-V. Nanometric Thin
Film Membranes Manufactured on Square Meter Scale: Ultra-thin
Films for CO2 Capture. Nanotechnology 2010, 21, 395301.
(3) Cho, W. K.; Kong, B.; Choi, I. S. Highly Efficient Non-Biofouling
Coating of Zwitterionic Polymers: Poly((3-(methacryloylamino)-
propyl)-dimethyl(3-sulfopropyl)ammonium hydroxide). Langmuir
2007, 23, 5678−5682.
(4) Yang, X.; Loos, J.; Veenstra, S. C.; Verhees, W. J. H.; Wienk, M.
M.; Kroon, J. M.; Michels, M. A. J.; Janssen, R. A. J. Nanoscale
Morphology of High-Performance Polymer Solar Cells. Nano Lett.
2005, 5, 579−583.
(5) Huang, Y.; Paul, D. R. Physical Aging of Thin Glassy Polymer
Films Monitored by Optical Properties. Macromolecules 2006, 39,
1554−1559.
(6) Priestley, R. D. Structural Relaxation of Polymer Glasses at
Surfaces, Interfaces, and In Between. Science 2005, 309, 456−459.
(7) Pye, J. E.; Rohald, K. A.; Baker, E. A.; Roth, C. B. Physical Aging
in Ultrathin Polystyrene Films: Evidence of a Gradient in Dynamics at
the Free Surface and Its Connection to the Glass Transition
Temperature Reductions. Macromolecules 2010, 43, 8296−8303.
(8) Baker, E. A.; Rittigstein, P.; Torkelson, J. M.; Roth, C. B.
Streamlined Ellipsometry Procedure for Characterizing Physical Aging
Rates of Thin Polymer Films. J. Polym. Sci., Part B: Polym. Phys. 2009,
47, 2509−2519.
(9) Pye, J. E.; Roth, C. B. Above, Below, and In-between the Two
Glass Transitions of Ultrathin Free-standing Polystyrene Films:
Thermal Expansion Coefficient and Physical Aging. J. Polym. Sci.,
Part B: Polym. Phys. 2015, 53, 64−75.
5585
Article
(10) Bodiguel, H.; Fretigny, C. Reduced Viscosity in Thin Polymer
Films. Phys. Rev. Lett. 2006, 97, 266105.
(11) Forrest, J. A.; Mattsson, J. Reductions of the Glass Transition
Temperature in Thin Polymer Films: Probing the Length Scale of
Cooperative Dynamics. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat.
Interdiscip. Top. 2000, 61, R53−R56.
(12) Keddie, J. L.; Jones, R. A. L.; Cory, R. A. Size-Dependent
Depression of the Glass Transition Temperature in Polymer Films.
Europhys. Lett. 1994, 27, 59−64.
(13) Batistakis, C.; Lyulin, A. V.; Michels, M. A. J. Slowing Down
versus Acceleration in the Dynamics of Confined Polymer Films.
Macromolecules 2012, 45, 7282−7292.
(14) Ediger, M. D.; Forrest, J. A. Dynamics near Free Surfaces and
the Glass Transition in Thin Polymer Films: A View to the Future.
Macromolecules 2014, 47, 471−478.
(15) Rittigstein, P.; Priestley, R. D.; Broadbelt, L. J.; Torkelson, J. M.
Model Polymer Nanocomposites Provide an Understanding of
Confinement Effects in Real Nanocomposites. Nat. Mater. 2007, 6,
278−282.
(16) Glor, E. C.; Fakhraai, Z. Facilitation of Interfacial Dynamics in
Entangled Polymer Films. J. Chem. Phys. 2014, 141, 194505.
(17) Zhang, C.; Guo, Y.; Shepard, K. B.; Priestley, R. D. Fragility of
an Isochorically Confined Polymer Glass. J. Phys. Chem. Lett. 2013, 4,
431−436.
(18) Zheng, W.; Simon, S. L. Confinement Effects on the Glass
Transition of Hydrogen Bonded Liquids. J. Chem. Phys. 2007, 127,
194501.
(19) Fakhraai, Z.; Forrest, J. A. Probing Slow Dynamics in Supported
Thin Polymer Films. Phys. Rev. Lett. 2005, 95, 025701.
(20) Fakhraai, Z.; Forrest, J. A. Measuring the Surface Dynamics of
Glassy Polymers. Science 2008, 319, 600−604.
(21) Roth, C. B.; McNerny, K. L.; Jager, W. F.; Torkelson, J. M.
Eliminating the Enhanced Mobility at the Free Surface of Polystyrene:
Fluorescence Studies of the Glass Transition Temperature in Thin
Bilayer Films of Immiscible Polymers. Macromolecules 2007, 40,
2568−2574.
(22) van Zanten, J. H.; Wallace, W. E.; Wu, W. Effect of Strongly
Favorable Substrate Interactions on the Thermal Properties of
Ultrathin Polymer Films. Phys. Rev. E: Stat. Phys., Plasmas, Fluids,
Relat. Interdiscip. Top. 1996, 53, R2053−R2056.
(23) Park, C. H.; Kim, J. H.; Ree, M.; Sohn, B.-H.; Jung, J. C.; Zin,
W.-C. Thickness and Composition Dependence of the Glass
Transition Temperature in Thin Random Copolymer Films. Polymer
2004, 45, 4507−4513.
(24) Ellison, C. J.; Torkelson, J. M. The Distribution of Glass-
transition Temperatures in Nanoscopically Confined Glass Formers.
Nat. Mater. 2003, 2, 695−700.
(25) Ellison, C. J.; Mundra, M. K.; Torkelson, J. M. Impacts of
Polystyrene Molecular Weight and Modification to the Repeat Unit
Structure on the Glass Transition-Nanoconfinement Effect and the
Cooperativity Length Scale. Macromolecules 2005, 38, 1767−1778.
(26) Qi, D.; Fakhraai, Z.; Forrest, J. A. Substrate and Chain Size
Dependence of Near Surface Dynamics of Glassy Polymers. Phys. Rev.
Lett. 2008, 101, 096101.
(27) Glor, E. C.; Composto, R. J.; Fakhraai, Z. Glass Transition
Dynamics and Fragility of Ultrathin Miscible Polymer Blend Films.
Macromolecules 2015, 48, 6682−6689.
(28) Sharp, J. S.; Forrest, J. A. Free Surfaces Cause Reductions in the
Glass Transition Temperature of Thin Polystyrene Films. Phys. Rev.
Lett. 2003, 91, 235701.
(29) Zhang, Y.; Glor, E. C.; Li, M.; Liu, T.; Wahid, K.; Zhang, W.;
Riggleman, R. A.; Fakhraai, Z. Long-range Correlated Dynamics in
Ultra-thin Molecular Glass Films. J. Chem. Phys. 2016, 145, 114502.
(30) Zhang, Y.; Fakhraai, Z. Decoupling of Surface Diffusion and
Relaxation Dynamics of Molecular Glasses. Proc. Natl. Acad. Sci. U. S.
A. 2017, 114, 4915−4919.
(31) Mundra, M. K.; Ellison, C. J.; Behling, R. E.; Torkelson, J. M.
Confinement, Composition, and Spin-coating Effects on the Glass
Transition and Stress Relaxation of Thin Films of Polystyrene and
DOI: 10.1021/acsnano.8b01341
ACS Nano 2018, 12, 5580−5587
ACS Nano
Styrene-containing Random Copolymers: Sensing by Intrinsic
Fluorescence. Polymer 2006, 47, 7747−7759.
(32) Fryer, D. S.; Peters, R. D.; Kim, E. J.; Tomaszewski, J. E.; de
Pablo, J. J.; Nealey, P. F.; White, C. C.; Wu, W. Dependence of the
Glass Transition Temperature of Polymer Films on Interfacial Energy
and Thickness. Macromolecules 2001, 34, 5627−5634.
(33) Glor, E. C.; Angrand, G. V.; Fakhraai, Z. Exploring the
Broadening and the Existence of Two Glass Transitions Due to
Competing Interfacial Effects in Thin, Supported Polymer Films. J.
Chem. Phys. 2017, 146, 203330.
(34) Klonos, P.; Kyritsis, A.; Pissis, P. Interfacial and Confined
Dynamics of PDMS Adsorbed at the Interfaces and in the Pores of
Silica−gel: Effects of Surface Modification and Thermal Annealing.
Polymer 2016, 84, 38−51.
(35) Alexandris, S.; Papadopoulos, P.; Sakellariou, G.; Steinhart, M.;
Butt, H.-J.; Floudas, G. Interfacial Energy and Glass Temperature of
Polymers Confined to Nanoporous Alumina. Macromolecules 2016, 49,
7400−7414.
(36) Cangialosi, D.; Alegría, A.; Colmenero, J. Effect of
Nanostructure on the Thermal Glass Transition and Physical Aging
in Polymer Materials. Prog. Polym. Sci. 2016, 54−55, 128−147.
(37) Kourki, H.; Famili, M. H. N.; Mortezaei, M.; Malekipirbazari,
M.; Disfani, M. N. Highly Nanofilled Polystyrene Composite: Thermal
and Dynamic Behavior. J. Elastomers Plast. 2016, 48, 404−425.
(38) DeMaggio, G. B.; Frieze, W. E.; Gidley, D. W.; Zhu, M.; Hristov,
H. A.; Yee, A. F. Interface and Surface Effects on the Glass Transition
in Thin Polystyrene Films. Phys. Rev. Lett. 1997, 78, 1524.
(39) Starr, F. W.; Schrøder, T. B.; Glotzer, S. C. Effects of a
Nanoscopic Filler on the Structure and Dynamics of a Simulated
Polymer Melt and the Relationship to Ultrathin Films. Phys. Rev. E:
Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 2001, 64, 021802.
(40) Sussman, D. M.; Tung, W.-S.; Winey, K. I.; Schweizer, K. S.;
Riggleman, R. A. Entanglement Reduction and Anisotropic Chain and
Primitive Path Conformations in Polymer Melts under Thin Film and
Cylindrical Confinement. Macromolecules 2014, 47, 6462−6472.
(41) Napolitano, S.; Rotella, C.; Wübbenhorst, M. Can Thickness
and Interfacial Interactions Univocally Determine the Behavior of
Polymers Confined at the Nanoscale? ACS Macro Lett. 2012, 1, 1189−
1193.
(42) Napolitano, S.; Capponi, S.; Vanroy, B. Glassy Dynamics of Soft
Matter under 1d Confinement: How Irreversible Adsorption Affects
Molecular Packing, Mobility Gradients and Orientational Polarization
in Thin Films. Eur. Phys. J. E: Soft Matter Biol. Phys. 2013, 36, 61.
(43) Burroughs, M. J.; Napolitano, S.; Cangialosi, D.; Priestley, R. D.
Direct Measurement of Glass Transition Temperature in Exposed and
Buried Adsorbed Polymer Nanolayers. Macromolecules 2016, 49,
4647−4655.
(44) Jouault, N.; Vallat, P.; Dalmas, F.; Said, S.; Jestin, J.; Boué, F.
Well-Dispersed Fractal Aggregates as Filler in Polymer-Silica Nano-
composites: Long-Range Effects in Rheology. Macromolecules 2009,
42, 2031−2040.
(45) Bansal, A.; Yang, H.; Li, C.; Cho, K.; Benicewicz, B. C.; Kumar,
S. K.; Schadler, L. S. Quantitative Equivalence between Polymer
Nanocomposites and Thin Polymer Films. Nat. Mater. 2005, 4, 693−
698.
(46) Sen, S.; Xie, Y.; Bansal, A.; Yang, H.; Cho, K.; Schadler, L. S.;
Kumar, S. K. Equivalence between Polymer Nanocomposites and Thin
Polymer Films: Effect of Processing Conditions and Molecular Origins
of Observed Behavior. Eur. Phys. J.: Spec. Top. 2007, 141, 161−165.
(47) Moll, J.; Kumar, S. K. Glass Transitions in Highly Attractive
Highly Filled Polymer Nanocomposites. Macromolecules 2012, 45,
1131−1135.
(48) Bera, O.; Pilić, B.; Pavličević, J.; Jovičić, M.; Holló, B.; Szécsényi,
K. M.; Špirkova, M. Preparation and Thermal Properties of
Polystyrene/silica Nanocomposites. Thermochim. Acta 2011, 515, 1−5.
(49) Huang, Y.-R.; Jiang, Y.; Hor, J. L.; Gupta, R.; Zhang, L.; Stebe,
K. J.; Feng, G.; Turner, K. T.; Lee, D. Polymer Nanocomposite Films
with Extremely High Nanoparticle Loadings via Capillary Rise
Infiltration (CaRI). Nanoscale 2015, 7, 798−805.
5586
Article
(50) Shavit, A.; Riggleman, R. A. The Dynamics of Unentangled
Polymers during Capillary Rise Infiltration into a Nanoparticle
Packing. Soft Matter 2015, 11, 8285−8295.
(51) Hor, J. L.; Wang, H.; Fakhraai, Z.; Lee, D. Effect of Polymer-
nanoparticle Interactions on the Viscosity of Unentangled Polymers
under Extreme Nanoconfinement during Capillary Rise Infiltration.
Soft Matter 2018, 14, 2438−2446.
(52) Hor, J. L.; Jiang, Y.; Ring, D. J.; Riggleman, R. A.; Turner, K. T.;
Lee, D. Nanoporous Polymer-Infiltrated Nanoparticle Films with
Uniform or Graded Porosity via Undersaturated Capillary Rise
Infiltration. ACS Nano 2017, 11, 3229−3236.
(53) Hensel, A.; Schick, C. Relation between Freezing-in Due to
Linear Cooling and the Dynamic Glass Transition Temperature by
Temperature-modulated DSC. J. Non-Cryst. Solids 1998, 235−237,
510−516.
(54) Fukao, K.; Miyamoto, Y. Slow Dynamics near Glass Transitions
in Thin Polymer Films. Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat.
Interdiscip. Top. 2001, 64 DOI: 10.1103/PhysRevE.64.011803.
(55) Napolitano, S.; Glynos, E.; Tito, N. B. Glass Transition of
Polymers in Bulk, Confined Geometries, and near Interfaces. Rep. Prog.
Phys. 2017, 80, 036602.
(56) Angell, C. A. Formation of Glasses from Liquids and
Biopolymers. Science 1995, 267, 1924−1935.
(57) Ediger, M. D.; Angell, C. A.; Nagel, S. R. Supercooled Liquids
and Glasses. J. Phys. Chem. 1996, 100, 13200−13212.
(58) Moynihan, C. T.; Macedo, P. B.; Montrose, C. J.; Montrose, C.
J.; Gupta, P. K.; DeBolt, M. A.; Dill, J. F.; Dom, B. E.; Drake, P. W.;
Easteal, A. J.; Elterman, P. B.; Moeller, R. P.; Sasabe, H.; Wilder, J. A.
Structural Relaxation in Vitreous Materials. Ann. N. Y. Acad. Sci. 1976,
279, 15−35.
(59) Bertei, A.; Nucci, B.; Nicolella, C. Effective Transport Properties
in Random Packings of Spheres and Agglomerates. Chem. Eng. Trans.
2013, 32, 1531−1536.
(60) Terao, K.; Mays, J. W. On-line Measurement of Molecular
Weight and Radius of Gyration of Polystyrene in a Good Solvent and
in a Theta Solvent Measured with a Two-angle Light Scattering
Detector. Eur. Polym. J. 2004, 40, 1623−1627.
(61) Chen, J.; Li, L.; Zhou, D.; Wang, X.; Xue, G. Effect of Geometric
Curvature on Vitrification Behavior for Polymer Nanotubes Confined
in Anodic Aluminum Oxide Templates. Phys. Rev. E 2015, 92, 032306.
(62) Napolitano, S.; Wübbenhorst, M. The Lifetime of the
Deviations from Bulk Behaviour in Polymers Confined at the
Nanoscale. Nat. Commun. 2011, 2, 260.
(63) Housmans, C.; Sferrazza, M.; Napolitano, S. Kinetics of
Irreversible Chain Adsorption. Macromolecules 2014, 47, 3390−3393.
(64) Alba-Simionesco, C.; Dosseh, G.; Dumont, E.; Frick, B.; Geil,
B.; Morineau, D.; Teboul, V.; Xia, Y. Confinement of Molecular
Liquids: Consequences on Thermodynamic, Static and Dynamical
Properties of Benzene and Toluene. Eur. Phys. J. E: Soft Matter Biol.
Phys. 2003, 12, 19−28.
(65) Morineau, D.; Xia, Y.; Alba-Simionesco, C. Finite-size and
Surface Effects on the Glass Transition of Liquid Toluene Confined in
Cylindrical Mesopores. J. Chem. Phys. 2002, 117, 8966−8972.
(66) Park, J.-Y.; McKenna, G. B. Size and Confinement Effects on the
Glass Transition Behavior of Polystyrene/o-terphenyl Polymer
Solutions. Phys. Rev. B: Condens. Matter Mater. Phys. 2000, 61,
6667−6676.
(67) Li, Q.; Simon, S. L. Surface Chemistry Effects on the Reactivity
and Properties of Nanoconfined Bisphenol M Dicyanate Ester in
Controlled Pore Glass. Macromolecules 2009, 42, 3573−3579.
(68) Kim, S.; Hewlett, S. A.; Roth, C. B.; Torkelson, J. M.
Confinement Effects on Glass Transition Temperature, Transition
Breadth, and Expansivity: Comparison of Ellipsometry and Fluo-
rescence Measurements on Polystyrene Films. Eur. Phys. J. E: Soft
Matter Biol. Phys. 2009, 30 DOI: 10.1140/epje/i2009-10510-y.
(69) Gibbs, J. H.; DiMarzio, E. A. Nature of the Glass Transition and
the Glassy State. J. Chem. Phys. 1958, 28, 373−383.
(70) DiMarzio, E. A.; Gibbs, J. H. Chain Stiffness and the Lattice
Theory of Polymer Phases. J. Chem. Phys. 1958, 28, 807−813.
DOI: 10.1021/acsnano.8b01341
ACS Nano 2018, 12, 5580−5587
ACS Nano Article

(71) Alcoutlabi, M.; McKenna, G. B. Effects of Confinement on

Material Behaviour at the Nanometre Size Scale. J. Phys.: Condens.
Matter 2005, 17, R461−R524.
(72) Hocky, G. M.; Markland, T. E.; Reichman, D. R. Growing Point-
to-Set Length Scale Correlates with Growing Relaxation Times in
Model Supercooled Liquids. Phys. Rev. Lett. 2012, 108, 225506.
(73) Berthier, L.; Kob, W. Static Point-to-set Correlations in Glass-
forming Liquids. Phys. Rev. E 2012, 85 DOI: 10.1103/Phys-
RevE.85.011102.
(74) Scheidler, P.; Kob, W.; Binder, K. The Relaxation Dynamics of a
Simple Glass Former Confined in a Pore. Europhys. Lett. 2000, 52,
277−283.
(75) Voyiatzis, E.; Rahimi, M.; Müller-Plathe, F.; Böhm, M. C. How
Thick Is the Polymer Interphase in Nanocomposites? Probing It by
Local Stress Anisotropy and Gas Solubility. Macromolecules 2014, 47,
7878−7889.
(76) Scheidler, P.; Kob, W.; Binder, K. Static and Dynamical
Properties of a Supercooled Liquid Confined in a Pore. J. Phys. IV
2000, 10, Pr7-33−Pr7-36.
(77) Scheidler, P.; Kob, W.; Binder, K.; Parisi, G. Growing Length
Scales in a Supercooled Liquid Close to an Interface. Philos. Mag. B
2002, 82, 283−290.
(78) Watanabe, K.; Kawasaki, T.; Tanaka, H. Structural Origin of
Enhanced Slow Dynamics near a Wall in Glass-forming Systems. Nat.
Mater. 2011, 10, 512−520.
(79) Jackson, C. L.; McKenna, G. B. The Glass Transition of Organic
Liquids Confined to Small Pores. J. Non-Cryst. Solids 1991, 131−133,
221−224.
(80) Zhang, J.; Liu, G.; Jonas, J. Effects of Confinement on the Glass
Transition Temperature of Molecular Liquids. J. Phys. Chem. 1992, 96,
3478−3480.
(81) Lopez, E.; Simon, S. L. Trimerization Reaction Kinetics and Tg
Depression of Polycyanurate under Nanoconfinement. Macromolecules
2015, 48, 4692−4701.
(82) Pye, J. E.; Roth, C. B. Physical Aging of Polymer Films
Quenched and Measured Free-Standing via Ellipsometry: Controlling
Stress Imparted by Thermal Expansion Mismatch between Film and
Support. Macromolecules 2013, 46, 9455−9463.
(83) Tao, P.; Li, Y.; Rungta, A.; Viswanath, A.; Gao, J.; Benicewicz, B.
C.; Siegel, R. W.; Schadler, L. S. TiO2 Nanocomposites with High
Refractive Index and Transparency. J. Mater. Chem. 2011, 21, 18623.
(84) Herzinger, C. M.; Johs, B.; McGahan, W. A.; Woollam, J. A.;
Paulson, W. Ellipsometric Determination of Optical Constants for
Silicon and Thermally Grown Silicon Dioxide via a Multi-sample,
Multi-wavelength, Multi-angle Investigation. J. Appl. Phys. 1998, 83,
3323−3336.

5587 DOI: 10.1021/acsnano.8b01341

ACS Nano 2018, 12, 5580−5587