Polymer nanocomposite films with extremely high nanoparticle loadings via capillary rise infiltration (CaRI).

View Article Online View Journal

Nanoscale Accepted Manuscript

This article can be cited before page numbers have been issued, to do this please use: Y. Huang, Y. Jiang, J. L. Hor, R. Gupta, L. Zhang, K. J. Stebe, G. Feng, K. Turner and D. Lee, Nanoscale, 2014, DOI:

This is an Accepted Manuscript, which has been through the Royal Society of Chemistry peer review process and has been accepted for publication. Accepted Manuscripts are published online shortly after acceptance, before technical editing, formatting and proof reading. Using this free service, authors can make their results available to the community, in citable form, before we publish the edited article. We will replace this Accepted Manuscript with the edited and formatted Advance Article as soon as it is available. You can find more information about Accepted Manuscripts in the Information for Authors. Please note that technical editing may introduce minor changes to the text and/or graphics, which may alter content. The journal’s standard Terms & Conditions and the Ethical guidelines still apply. In no event shall the Royal Society of Chemistry be held responsible for any errors or omissions in this Accepted Manuscript or any consequences arising from the use of any information it contains.

www.rsc.org/nanoscale

Page 1 of 9

Nanoscale View Article Online

DOI: 10.1039/C4NR05464D

Nanoscale

RSCPublishing

ARTICLE

Received 00th January 2012, Accepted 00th January 2012 DOI: 10.1039/x0xx00000x www.rsc.org/

Yun-Ru Huang,a, e, Yijie Jiang, b Jyo Lyn Hor, a Rohini Gupta, a Lei Zhang, a, d Kathleen J. Stebe, a Gang Feng,c Kevin T. Turner b and Daeyeon Lee a, * Polymer nanocomposite films (PNCFs) with extremely high concentrations of nanoparticles are important components in energy storage and conversion devices and also find use as protective coatings in various applications. PNCFs with high loadings of nanoparticles, however, are difficult to prepare because of the poor processability of polymer-nanoparticle mixtures with high concentrations of nanoparticles even at an elevated temperature. This problem is especially exacerbated when anisotropic nanoparticles are the desired filler materials. Here we report a straightforward method for generating PNCFs with extremely high loadings of nanoparticles. Our method is based on what we call capillary rise infiltration (CaRI) of polymer into a dense packing of nanoparticles. CaRI consists of two simple steps: 1) the preparation of a two-layer film, consisting of a porous layer of nanoparticles and a layer of polymer and 2) annealing of the bilayer structure above the temperature that imparts mobility to the polymer (e.g., glass transition of the polymer). The second step leads to polymer infiltration into the interstices of the nanoparticle layer, reminiscent of the capillary rise of simple fluid (e.g., water) into a narrow capillary or a packing of granules. We use in situ spectroscopic ellipsometry and a three-layer Cauchy model to follow the capillary rise of polystyrene (molecular weight = 8,000) into the random network of nanoparticles. The infiltration of polystyrene into a densely packed TiO 2 nanoparticle layer is shown to follow the classical Lucas-Washburn type of behaviour. We also demonstrate that PNCFs with densely packed anisotropic TiO2 nanoparticles can be readily generated by spin coating anisotropic TiO2 nanoparticles atop a polystyrene film and subsequently thermally annealing the bilayer film. We show that CaRI leads to PNCFs with modulus, hardness and scratch resistance that are far superior to the properties of films of the component materials. In addition, CaRI fills in cracks that may exist in the nanoparticle layer, leading to the healing of nanoparticle films and the formation of defect-free PNCFs. We believe this approach is widely applicable for the preparation of PNCFs with extremely high loading of nanoparticles and potentially provides a unique approach to study capillarity-induced transport of polymers under extreme confinement.

Introduction Polymer nanocomposite films (PNCFs), composed of nanoparticles and polymers, combine the unique electronic, mechanical, plasmonic, catalytic and optical properties of nanoparticles with the flexibility and processability of polymers, resulting in nanostructured films with synergistic properties.1-15 PNCFs with extremely high loadings of nanoparticles (> 50 vol%; that is, the majority phase of the PNCFs is the nanoparticle phase), in particular, are increasingly becoming important due to their desirable mechanical and transport properties. For example, recent studies have shown that

This journal is © The Royal Society of Chemistry 2013

nanocomposites containing highly anisotropic nanoparticles such as nanoplatelets with a small amount of polymer, mimicking the structure of natural nanocomposites such as nacre, exhibit exceptional strength, stiffness and toughness at the same time.16-20 In the area of solid state dye-sensitized solar cells, a significant amount of efforts is currently devoted to generating photoanodes made of densely packed TiO2 nanoparticles with polymeric electrolytes filling the interstices of the TiO2 nanoparticle network.21-23 Although several techniques to fabricate PNCFs with high loadings of nanoparticles have been developed, each method presents some

Nanoscale., 2013, 00, 1‐3 | 1


Published on 17 November 2014. Downloaded by McGill University on 20/11/2014 18:09:11.

Cite this: DOI: 10.1039/x0xx00000x

Polymer Nanocomposite Films with Extremely High Nanoparticle Loadings via Capillary Rise Infiltration (CaRI)

Nanoscale

Page 2 of 9 View Article Online

challenges and limitations that must be overcome. Generating PNCFs with high loadings of nanoparticles using conventional melt compounding or solution-based processes, for example, is challenging because of the difficulty of mixing and stabilizing nanoparticles in a polymer matrix.24-26 Nanoparticles have a strong tendency to aggregate due to their large interparticle forces and high specific area, making it difficult to distribute nanoparticles uniformly throughout the polymer matrix.27, 28 In addition, the processability of polymer-nanoparticle mixtures with high contents of anisotropic particles becomes extremely challenging due to very high viscosity and elasticity of these composites at even moderate filler fractions. In situ polymerization, which involves mixing monomers and nanoparticles followed by polymerizing the monomers, takes advantages of low-viscosity monomer to generate PNCFs. The method nevertheless is not applicable to a wide range of polymers and may suffer from incomplete polymerization and difficulty in controlling the molecular weight and uniformity of polymers in PNCFs.29 Layer-by-layer (LbL) assembly of oppositely charged nanoparticles and polymers provides a versatile approach in generating high nanoparticle-fraction PNCFs with high uniformity and conformity; however, the process tends to be time-consuming and also is typically limited to water soluble charged nanoparticles and polymers.30-32 An alternative method that enables the formation of high nanoparticle-content PNCFs using commonly available polymers would be highly desirable for a variety of applications. In this work, we develop a simple process to generate PNCFs with a uniform distribution of nanoparticles at extremely high filler concentrations. We are inspired by commonly observed phenomena of water wicking into porous media such as packings of granules and sands via capillarity.33 Our method is based on generating a twolayer film composed of a polymer layer and a layer of nanoparticles, and subsequently annealing the bilayer structure at elevated temperature that imparts mobility to the polymer. This process leads to polymer infiltration into the interstices of the nanoparticle layer via capillary action. Interestingly, capillary rise of polymers has previously been used to fill well-defined cylindrical pores in anodized alumina membranes to generate polymer nanowires;34 however, it has not been used to generate PNCFs composed of nanoparticles and polymers. Our new approach, we believe, is complimentary to the techniques discussed above and enables the straightforward generations of PNCFs with very high loadings of nanoparticles using common nanoparticles and polymers. We show that PNCFs with densely packed nanoparticles can be readily generated and that we can monitor the formation of PNCFs in situ based on spectroscopic ellipsometry. PNCFs generated based on CaRI exhibit superior mechanical properties and scratch/wear resistance compared to films made of the individual components. We also demonstrate that CaRI can “heal” nanoparticle films with cracks that exist prior to the annealing process.

Experimental Materials TiO2 nanoparticles with aspect ratio (AR) around 1, 2, 4, and 6 were synthesized using a previously reported gel-sol method.35 The minor axis for AR1 particles was 2a = 24 ± 3 nm, the major axis was 2b = 31 ± 6 nm, and AR = 1.31 ± 0.39; for AR2 particles: 2a = 28 ± 3 nm, 2b = 73 ± 10 nm, AR = 2.58 ± 0.52; for AR4 particles: 2a = 28 ± 4 nm, 2b = 134 ± 22 nm, AR = 4.75 ± 0.60; for AR6 particles: 2a = 32 ± 5 nm, 2b = 204 ± 33 nm, AR = 6.41 ± 1.33. Polystyrene (average molecular weight,

2 | J. Name., 2012, 00, 1‐3

Journal Name DOI: 10.1039/C4NR05464D Mn = 8000 g/mol, polydispersity index (PDI) = 1.10) was purchase from Polymer Source, Inc. PS/TiO2 PNCFs fabrication Silicon wafers were cleaned by soaking in piranha solution at 75°C in a water bath for at least 15 mins, rinsed with deionized water and acetone, and dried with nitrogen. To fabricate the PNCFs, a PS film was first deposited on the substrate using a WS-400BZ-6NPP/Lite spin coater from Laurell Technologies Corporation, followed by an oxygen plasma treatment for two seconds to render the surface of PS hydrophilic. A TiO2 NP film was subsequently spin coated on top of the plasma-treated PS layer at a rotational speed of 1000 rpm for 3 mins to form a bilayer film. The concentration of PS solution was 15 wt%, diluted with toluene, for depositing PS layer and the concentrations of TiO2 suspension were 50 wt%, 50 wt%, 40 wt% and 30 wt%, diluted with deionized water, for AR1, AR2, AR4 and AR6 TiO2 NP layer, respectively. The bilayer film was then annealed in an oven, which had been heated to the set temperature (130°C), until the interstices in the TiO2 NP layer was completely filled with PS via capillary infiltration. Characterization The in situ monitoring of PS infiltration into the nanopores in the TiO2 NP layer was performed using an alpha-SE spectroscopic ellipsometer with a heating stage. The sample temperature was controlled with a resolution of 0.1°C using the heating stage, Linkam THMS350V. The test sample was placed on the sample block holder (22 nm in diameter) and exposed to atmosphere to enable ellipsometry measurements. The Linksys32 interface was used to program the temperature controller, namely input the desired temperature, cooling/heating rate, and hold time as well as read the actual temperature. For the experiment mentioned in this work, the samples were heated to the desired temperature (120°C, 125°C, or 130°C) at a rate of 30°C /min (maximum possible with THMS350V). The dynamic data were obtained from 380 nm to 900 nm at an incident angle of 70°C and were fitted using a multilayer model established in the Complete EASE software package provided by J.A. Woollam. The cross-sectional scanning electron microscopy (SEM) images were taken using a FEI-600 Quanta to characterize the morphology of PS in the voids and the thickness of each layer before and after annealing. Each sample was coated with 4 nm of iridium before imaging to prevent charging. The SEM images were captured at an accelerating voltage of 15 kV and the working distances were 10 mm. The modulus and hardness of PS film, TiO2 NP films and PS/TiO2 PNCFs (~2.5 μm in thickness) were determined using a Nano IndenterTM G200 from AgilentTM Technologies, Inc. with continuous stiffness measurement (CSM). A Berkovich tip was used for the indentation and the area function of the indenter tip was calibrated using fused silica. The thermal drift rate of the indenter was stabilized to the value less than 0.1 nm/s before performing any indentation. The CSM harmonic displacement (amplitude) was set at 1.2 nm. For all indentations, the loading of constant strain rate of 0.05 s-1 was used and the indentation depth of 2.5 μm was performed. Nine indentations were performed on each sample and the depth profiles of the mechanical properties were analysed using Agilent AnalystTM




ARTICLE


Journal Name


software to determine the statistical averages of hardness and modulus in the indentation depth range of 200-300 nm. The Poisson’s ratio for PS film, TiO2 film and PS/TiO2 PNCF were assumed to be 0.45, 0.22 and 0.45, respectively. The effect of on the finite thickness of the film on the indentation modulus was corrected using a previous reported method,36 which is valid for indentation depth less than 50% of the film thickness. The substrate effect on the indentation hardness is generally negligible if the indentation depth is less than 30% of the film thickness.37, 38 Thus, the hardness was directly determined using the Oliver-Pharr method.39 Nanoindenter scratch tests were conducted using a Hysitron TI 950 TriboIndenter®. A diamond Berkovich tip was brought into contact with the sample and, for each scratch, the tip was moved in a sliding motion in contact with the surface under a compressive load of 500 μN at a speed of 2 μm/s. The length of a scratch line was 10 μm. Three different locations without observable defects under microscope on each sample were chosen randomly and 4 scratches with a separation of 5 μm were made on each location. To capture the topography of the scratched surface, a 20 x 20 μm2 image was taken by raster scanning on the surface under 0.5 μN using the nanoindenter directly. The wear tests were carried out by sliding an ultrananocrystalline diamond UNCD probe in a raster form on a 500 × 500 nm2 region using a Dimension Icon® AFM. The load was fixed at 500 nN and the speed was 4 μm/s. Tapping mode images were taken before and after both 1 and 5 wear scans over an area of 1 × 1 μm2. The wear tests were repeated at three different locations on each sample to get statistical averages of height losses. Scratch tests using AFM were also done under the same loading and speed conditions. Each scratch line was 1.5 μm in length and included 128 passes of the tip across the surface. Two different locations were chosen randomly on each sample and 4 parallel lines with 400 nm in distance were scratched on each location.

Results and Discussion A polymer nanocomposite film (PNCF) with a high fraction of nanoparticles is prepared by generating a bilayer film consisting of a nanoparticle layer on top of a polymer layer, followed by annealing the bilayer structure at an elevated temperature. The materials we use for the study are monodisperse polystyrene (PS, Mn = 8000 g/mol) and titanium dioxide (titania) nanoparticles (TiO2 NPs). TiO2 NPs with different aspect ratios are synthesized using a hydrothermal method.35, 40 In this work, we use a low-molecular weight of PS to show the feasibility of this approach; however, we have confirmed that this method can be used with high molecular weight of PS (at least up to Mn = 183,000 g/mol) and those results will be reported in our subsequent papers. To produce a polymer/nanoparticle bilayer film, we spin coat a PS layer from toluene-based solution onto a silicon substrate and then a TiO2 NP layer from an aqueous suspension of TiO2 NPs (~ 25 nm in diameter) onto the PS layer. Because of the insolubility of PS in water, the subsequent spin coating of a TiO2 NP layer does not disrupt or dissolve the pre-existing PS layer. The thickness of each layer can be readily controlled by either changing the concentration of PS or TiO2 NPs, and/or by varying the spin speed (i.e., RPM) during coating. The bilayer structure is annealed at a temperature above the glass transition temperature (Tg) of PS in a vacuum


ARTICLE DOI: 10.1039/C4NR05464D oven. The interconnected nanopores in the TiO2 NP layer are gradually filled with PS melt via capillary action as schematically illustrated in Figure 1. A clear front of PS “invading” the TiO2 nanoparticle layer can be seen in a high magnification SEM image (see Figure S1 in Supplementary Information), indicating that the process is extremely similar to the capillary rise of simple liquid into a packing of sand or granules.41, 42 By reducing the temperature to below the Tg of PS, the infiltrated polymer solidifies in the nanopores and a PS/TiO2 PNCF with high loadings of nanoparticle is obtained.

Figure 1. (a) Schematic illustration showing the process to make a polymer nanocomposite film (PNCF) by capillary rise infiltration (CaRI) of polystyrene (PS) into the nanopores of TiO2 nanoparticle (NP) film. (b) Scanning electron microscopy (SEM) images showing CaRI of a bilayer film consisting of PS (Mn = 8,000) and a TiO2 NP layer. From left to right, the SEM images are taken after the bilayer is annealed at 130 ˚C for 0, 2 and 20.5 hrs, respectively. Changes in the thickness of the bottommost PS layer after 8 hours as determined by SEM are negligible. TiO2 NPs in this example are nanoellipsoids with an aspect ratio of 4. SEM image of the top surface of a sample prepared using the same method is available in ESI. To study the dynamics of PS infiltration into the interstices of the 25 nm aspect ratio (AR) 1 TiO2 NP layer, we use in situ spectroscopic ellipsometry to measure the changes in the optical properties and thickness of the materials.43, 44 Assuming that the infiltration process is based on capillarity-induced wicking, we can divide the system undergoing CaRI into three layers: a neat PS layer, a PS/TiO2 nanocomposite layer, and a neat TiO2 NP layer. By monitoring the thickness of the bottommost PS layer, it is possible to deduce the amount of polymer that has wicked into nanoparticle layer. We use a three-layer Cauchy model to interpret the amplitude ratio (ψ) and the phase difference (∆) from the in situ spectroscopic ellipsometry and follow the CaRI process. The refractive index (n) of each Cauchy layer as a function of wavelength (λ) is represented by: n( )  A 

B

2



C

4

(1)

where A, B, and C are optical constants of the Cauchy model. To enable the modelling, we determine the optical constants (A, B, and C) of spin-coated neat PS and neat 25 nm AR1 TiO2 NP film on Si wafer substrates and use those constants for the three-layer Cauchy modelling (see Supplementary Information). We also obtain these constants for the PS-filled TiO2 nanoparticle layer by determining the constants for the

J. Name., 2012, 00, 1‐3 | 3


Page 3 of 9

Nanoscale

Journal Name DOI: 10.1039/C4NR05464D

fully infiltrated TiO2 nanocomposite layer (see Supplementary Information for details). Figure 2a shows the time-dependent changes in the thickness of each layer upon annealing a bilayer film composed of 320 nm of PS and 320 nm of TiO2 NP layers at different temperatures. The temperature of the sample is raised from the room temperature to a target temperature at a rate of 30°C/min. At the beginning of the annealing process, the thickness of each layer remains unchanged because the temperature is lower than the Tg of PS (~ 87°C for PS with Mn = 8,000 g/mol).45, 46 When the annealing temperature is raised above the Tg of PS, the thicknesses of the PS layer and the TiO2 NP layer start to decrease, whereas the thickness of PS/TiO2 nanocomposite layer increases, indicating PS is filling the voids in the TiO2 NP layer. The reduction in the thickness of bottommost neat PS layer stops once the thickness of TiO2 NP layer plateaus at ~ 0 nm. This result is a strong indication that the driving force for the infiltration of PS into the TiO2 NP network is indeed capillary force since the driving force would vanish once the interconnected nanopores are completely filled with PS. Additionally, we find that the sum of the thickness of the uppermost neat TiO2 NP layer and the middle PS/TiO2 nanocomposite layer is essentially the same as the initial thickness of the neat TiO2 NP layer throughout the annealing process. This consistency is a strong indication that the infiltration of PS into the interstices of the TiO2 NP layer does not cause any swelling in the nanoparticle layer and thus the interparticle distance most likely remains unchanged from that of the original neat TiO2 NP layer.

reported values of μ and σ for 8k PS, the estimation of the contact angle based on the prefactor of the Lucas-Washburn equation gives 90˚. Such a contradiction strongly suggests that the bulk values of μ and σ that we use for the lumped constant may not be applicable for this system; that is, these values may be significantly affected by the confinement of the polymer in nanopores (R < 5 nm).51-54

Capillary rise of common liquid such as water into porous media like a packing of sand has been studied extensively. These processes can be modelled using the modified LucasWashburn equation, which defines the height of liquid rise into a porous medium as a function of time47, 48:

h2 

 R cos  t 4 2 

(2)

where h is the height of liquid rise, σ the surface tension, R the mean pore radius, θ the contact angle, t the time, τ is the tortuosity of the porous network and μ the viscosity of liquid.49 This model clearly shows that the viscosity of the fluid has a significant impact on the capillary rise dynamics. Since the viscosity of PS depends sensitively on the temperature above Tg, we study the effect annealing temperature on the dynamics of PS infiltration into TiO2 nanoparticle layers.50 As the annealing temperature is increased above Tg, the dynamics of capillary rise is significantly accelerated as seen in Figure 2b. When the curves are plotted in a log-log plot, the slopes of all three curves show that these phenomena are in fact consistent with the Lucas Washburn behaviour of common liquids undergoing capillary rise into porous media, as evidenced by the ½ scaling. The log-log plot intercept is related to the factors inducing CaRI, as indicated by the prefactor in the modified LucasWashburn equation (i.e.,  R cos  ) . We use parameters that 4 2  readily available for σ, R, τ and μ in the literature to estimate the contact angle (θ) of PS on TiO2 NPs (see Supplementary Information for details). The fact that we observe capillary rise implies that the contact angle of 8k PS on the TiO2 nanoparticle is less than 90°. However, when we use the

4 | J. Name., 2012, 00, 1‐3

Figure 2. (a) Changes in the thicknesses of polystyrene (PS), nanocomposite, and TiO2 layer as a function of time measured using in situ spectroscopic ellipsometry at 130°C. (b) Timedependent change in the height of composite layers at different annealing temperature. The initial thicknesses of TiO2 layers used for 130, 125 and 120 ˚C annealing are 320, 335 and 280 nm, respectively. These three samples were prepared separately. Inset: the log-log plot of composite height versus time shows consistency with the modified Lucas-Washburn model. To test the possibility of generating high nanoparticle-fraction PNCFs with anisotropic nanoparticles, we use ellipsoidal TiO2 NPs with different aspect ratios as nanofillers. The TiO2 NPs with aspect ratio (AR) around 1, 2, 4 and 6 are used and denoted as AR1, AR2, AR4 and AR6, respectively.43 Our previous work has shown that spin coating of these TiO2 nanoellipsoids leads to densely packed nanoparticle films with randomly oriented nanoparticles. The cross-sectional scanning electron microscopy (SEM) images of PNCFs, as seen in Figure 3, show that the PS indeed can successfully infiltrate into the



ARTICLE



Page 5 of 9


ARTICLE DOI: 10.1039/C4NR05464D


nanopores of TiO2 layers made with spherical or ellipsoidal nanoparticles. Small residual PS layers in all of these annealed films do not disappear even if the samples are annealed for a long period of time (> 1 week) at 130°C, indicating that the complete filling of the nano-interstices have been achieved via CaRI.

CaRI PNCFs with high nanoparticle fraction are expected to have high modulus and hardness as well as excellent damage tolerance such as wear/scratch resistance, making them attractive coatings for the protection of surfaces and materials against mechanical stresses.57 We test and compare the mechanical properties of the CaRI PNCFs to neat TiO2 NP and PS films using nanoindentation.39 Each sample is prepared by annealing a bilayer film with a polymer layer that is slightly thicker (~ 5 – 10 %) than the amount that would be necessary to fully saturate the interstices of the nanoparticle layer. This approach ensures that the interstices are completely filled with polymer and also minimizes the effect of the residual film on the mechanical characterization. The comparison of modulus and hardness of these PNCFs to the neat nanoparticle films made of TiO2 nanoellipsoids with different aspect ratios and PS films are summarized in Figure 4a-b. Regardless of the aspect ratio of TiO2 NPs, the hardness and modulus of the PNCFs are significantly higher than those of neat TiO2 NP and neat PS films. Furthermore, the modulus and hardness of the PNCF with different aspect ratio TiO2 NPs follow the similar order as those of neat TiO2 NP films, indicating that the nanostructure of the PNCF has a significant impact on the mechanical properties of the composite. The hardness and modulus measured from PS/AR1 TiO2 PNCF show the largest values that we are aware of for PS/TiO2 NP composite systems that have been studies to date10, 27, 58-61, clearly illustrating the power of making high filler fraction PNCF based on CaRI.

Figure 3. Cross-sectional scanning electron microscopy (SEM) images of PNCFs generated by CaRI of PS into (a) AR1 (b) AR2 (c) AR4 (d) AR6 TiO2 NP layers. The volume fraction of the TiO2 NP layer made with various AR of particles is readily determined by examining the thickness change of PS layer with respect to the initial thickness of TiO2 NP layer; that is, the porosity (p) of the TiO2 layer can be expressed as: p

where

LPSi

and

LPS f

LPSi  LPS f LTiO2

(3)

are the initial and final thickness of PS

layer, respectively, and

LTiO2

is the initial thickness of TiO2

NP layer. As summarized in Table 1, AR2 NP film has the highest volume fraction (0.65), whereas AR6 NP film has the lowest volume fraction of 0.41. The measured volume fractions of nanoparticles in PS/TiO2 PNCFs are in excellent agreement with previous computational studies and experimental work that described the volume fraction of random packing of prolate ellipsoids.40, 55, 56 Our results clearly demonstrate that using CaRI, we can obtain PNCFs with extremely high fractions of nanofillers that are forming a densely packed percolated network in polymer matrix and that changing the shape of nanoparticles provides a way to control the volume fraction of nanoparticles in the PNCFs. Table 1. Volume fraction of TiO2 NP layer as determined from the thickness of PS layers before and after thermal annealing. AR1 AR2 AR4 AR6 Volume 62.8 ± 0.4 64.9 ± 0.1 58.4 ± 0.1 41.6 ± 0.4 fraction


Figure 4. (a) Modulus and (b) hardness of PS, TiO2 NP films and CaRI PNCFs obtained from nanoindentation measurements with a Berkovich indenter tip. Every point is determined by a statistical average of nine continuous stiffness measurements (CSM) between d = 200-300 nm, where d is the depth from the surface of the testing film. Plan-view images of (c) pure PS film, (d) pure AR1 TiO2 film, and (e) PS/AR1 TiO2 PNCF film after nanoindenter scratch tests. Each image shows four scratches that were made under the same loading condition (500 μN normal load). We also characterize and compare the nanoscale scratch and wear resistance of CaRI PNCFs to those of neat PS and TiO2 NP films. Due to significantly different surface roughness we observe in the films made of non-spherical nanoparticles, we focus our characterizations on PNCFs with AR1 TiO2 nanoparticles. Nanoindenter scratch tests are conducted using a

J. Name., 2012, 00, 1‐3 | 5


Journal Name

Nanoscale


nanoindenter on three different samples: neat PS film, AR1 TiO2 film and PS/TiO2 PNCF. A diamond Berkovich tip is brought into contact with the sample and, for each scratch, the tip is slid across a surface under a load of 500 μN and at a speed of 2 μm/s. 12 scratches are conducted on each sample using the same protocol at random locations. Although slightly rougher than polymer-filled nanoparticle films, neat nanoparticles films have sufficient uniformity for us to reliably conduct our scratch and wear tests. As observed in the images of the scratched surfaces (Figure 4c-e), each scratch exhibits a groove with two pile-up pads along the edges. The neat PS film shows significant deposition of debris along the edges, whereas the neat TiO2 NP film shows fairly clean grooves. Interestingly, PS/TiO2 PNCF shows an intermediate behaviour. As seen in Table 2, PS/TiO2 PNCF has the shallowest scratch depth, thus demonstrating better scratch resistance compared than the neat TiO2 NP and neat PS. The measured lateral force normalized by the cross-sectional area of the scratch, which is reported in Table 2, is largest for the PS/TiO2 PNCF composite. The lateral force per area is a measure of the stress required to deform and displace the material during the scratch, thus this result suggests the CaRI PNCF has better mechanical properties than the neat PS and the neat NP films. Table 2. Average scratch depth and lateral forces of neat PS film, AR1 TiO2 film, and PS/AR1 TiO2 PNCF using nanoindenter. Average scratch Average lateral force depth (nm) per area (nN/nm2) 54.6 ± 12.5 20 ± 7.6 Neat PS film 45.1 ± 6.9 26 ± 7.3 AR1 TiO2 film PS/AR1 TiO2 38.6 ± 6.6 31 ± 8.2 PNCF CaRI PNCFs also show improved nanoscale wear resistance when compared to neat TiO2 NP film. Wear tests are carried out in a contact mode atomic force microscopy (AFM) by sliding an ultrananocrystalline diamond (UNCD) probe over an area of 500 × 500 nm2 with 256 lines back and forth on the surface. As summarized in Table 3, the average height losses for the neat AR1 TiO2 film are larger than that of the PS/AR1 TiO2 PNCF for both 1 and 5 wear scans. The PS/AR1 TiO2 PNCF has approximately 39% and 18% less reduction in the film thickness than the neat AR1 TiO2 film after 1 and 5 wear scans, respectively. Interestingly, PNCFs show better wear resistance (~ 40%) even though the reduction in the height of the films is smaller than the size of TiO2 NPs (~ 30 nm). We also confirm that the scratch tests performed using AFM with a much smaller normal force than the nanoindenter scratch tests show that the scratch resistance of PNCF is better than neat TiO2 NP films even if the depths of scratches are less than the size of TiO2 NPs (Supplementary Information). These results clearly indicate that CaRI leads to PNCFs with high nanofiller fraction, which imparts them with enhanced stiffness and hardness as well as nanoscale wear and scratch resistance. Table 3. Average height losses of AR1 TiO2 film and PS/AR1 TiO2 PNCF after 1 and 5 wear scans. Sample Height loss after 1 Height loss after 5 wear scan (nm) wear scans (nm) Neat AR1 TiO2 4.33 ± 1.13 7.30 ± 1.91 film PS/AR1 TiO2 3.09 ± 0.40 6.23 ± 1.09 PNCF

6 | J. Name., 2012, 00, 1‐3


Figure 5. During CaRI, PS infiltration occurs into cracks that are in the film prior to the annealing process. Cross-sectional SEM images of a PNFC with cracks (a) before and (b) after annealing. One remarkable feature of CaRI is that the capillary rise of polymers can occur into microscopic defects such as pinholes and cracks in nanoparticle layers; that is, polymers are able to infiltrate and fill in these microscopic openings and “heal” the nanoparticle layer during the annealing process. It is well known that when a thick layer of particle film is deposited from a suspension, lateral stress develops due to the drying and evaporation of the medium and can result in cracking of the colloidal film.62-64 These cracks are highly undesirable features for practical applications of PNCFs, as they significantly degrade the transport, optical and mechanical properties of these structures. The cross-sectional SEM images in Figure 5 clearly show that polymer infiltration occurs into pre-existing cracks in the nanoparticle layer and completely closes the cracks. We believe this healing effect is another versatile feature of CaRI that make it an extremely attractive process for generating PNCFs with high volume fractions of nanomaterials.

Conclusions and Outlook We have successfully fabricated PNCFs with high loadings of nanoparticles using the CaRI method. The dynamics of CaRI of a low molecular weight (Mn = 8k) polystyrene into the porous network of spherical TiO2 NP film is described by the LucasWashburn model. CaRI allows for a straightforward preparation with PNCFs with anisotropic nanoparticles and also heals nanoparticle layers with pre-existing microscale defects such as cracks. We show the mechanical properties and damage tolerance of CaRI PNCFs are significantly superior to those of the films made of individual components that make up the PNCFs. We believe that CaRI provides a versatile method that enables the generation of high filler fraction PNCFs with different sets of polymers and nanoparticles. Such PNCFs will have useful functionality for advanced applications in energy (e.g., dyesensitized solar cells, batteries), separations (e.g., membranes), coatings and display devices. We also believe that CaRI can be a scalable method to generate PNCFs with extremely high volume fraction of nanoparticles as long as uniform bilayers of nanoparticle and polymer can be generated using a variety of film deposition techniques. In addition to these practical aspects, CaRI presents unique opportunities to address a number of fundamental questions. For example, our results indicate that CaRI of polymer into nanoporous media is influenced by a confinement effect. Thus it will be interesting to study how confinement influences the




ARTICLE

Page 7 of 9



important parameters such as surface tension and viscosity of polymers and in turn the CaRI dynamics. By changing the size of nanoparticles or the molecular weight of the polymer, it will be possible to study the effect of confinement on the dynamics of CaRI as well as the effect of chain entanglement on the dynamics of CaRI.


Acknowledgements This work was primarily supported by the PENN MRSEC DMR-1120901, NSF CBET-1449337 and an NSF CAREER Award (DMR-1055594). K. T. T and G.F. acknowledge the support from NSF CMMI-1200019 and Pennsylvania Keystone Innovation Starter Kit (KISK) grant in nanotechnology, respectively. a

Department of Chemical and Biomolecular Engineering, University of

Pennsylvania, Philadelphia, Pennsylvania 19104, United States b

Department of Mechanical Engineering and Applied Mechanics,

University of Pennsylvania, Philadelphia, Pennsylvania 19104, United States c

Department

of

Mechanical

Engineering,

Villanova

University,

Villanova, Pennsylvania 19085, United States d

Department of Mechanical Engineering, University of Alaska Fairbanks,

Fairbanks, Alaska 99775, United States e

Present address: Department of Chemical Engineering, Carnegie Mellon

University * Corresponding author: Daeyeon Lee ([email protected]) Electronic

Supplementary

Information

(ESI)

available:

High

magnification SEM images of polystyrene infiltrating TiO2 nanoparticle layer, in situ spectroscopic ellipsometry determination of three layers during CaRI, contact angle estimation based on the Lucas-Washburn equation, atomic force microscopy images of samples after wear tests and height profiles of samples after scratch test. See DOI: 10.1039/b000000x/

Notes and references 1. J. M. Garcés, D. J. Moll, J. Bicerano, R. Fibiger and D. G. McLeod, Advanced Materials, 2000, 12, 1835-1839. 2. A. Chandra, L.-S. Turng, P. Gopalan, R. M. Rowell and S. Gong, Composites Science and Technology, 2008, 68, 768-776. 3. S. C. Tjong, in Polymer Composites with Carbonaceous Nanofillers, John Wiley & Sons Inc., 2012, DOI: 10.1002/9783527648726.ch9, pp. 351-379. 4. Rajesh, T. Ahuja and D. Kumar, Sensors and Actuators B-Chemical, 2009, 136, 275-286. 5. H. Liu and T. J. Webster, Journal of Biomedical Materials Research Part A, 2010, 93A, 1180-1192. 6. A. C. Balazs, T. Emrick and T. P. Russell, Science, 2006, 314, 11071110. 7. G. M. Odegard, T. C. Clancy and T. S. Gates, Polymer, 2005, 46, 553562. 8. T.-I. Yang and P. Kofinas, Polymer, 2007, 48, 791-798. 9. P. Murugaraj, D. Mainwaring and N. Mora-Huertas, Journal of Applied Physics, 2005, 98. 10. A. Chandra, L. S. Turng, S. Q. Gong, D. C. Hall, D. F. Caulfield and H. J. Yang, Polymer Composites, 2007, 28, 241-250. 11. K. I. Winey and R. A. Vaia, MRS Bulletin, 2007, 32, 314-322. 12. K. I. Winey, T. Kashiwagi and M. F. Mu, Mrs Bulletin, 2007, 32, 348353. 13. L. S. Schadler, L. C. Brinson and W. G. Sawyer, JOM, 2007, 59, 53-60. 14. J. H. Fendler, Nanoparticles and Nanostructured Films: Preparation, Characterization, and Applications, John Wiley & Sons, 1998.


15. K. Ariga, Q. Ji, J. P. Hill, Y. Bando and M. Aono, NPG Asia Mater, 2012, 4, e17. 16. L. J. Bonderer, A. R. Studart and L. J. Gauckler, Science, 2008, 319, 1069-1073. 17. Z. Y. Tang, N. A. Kotov, S. Magonov and B. Ozturk, Nature Materials, 2003, 2, 413-U418. 18. A. Sellinger, P. M. Weiss, A. Nguyen, Y. F. Lu, R. A. Assink, W. L. Gong and C. J. Brinker, Nature, 1998, 394, 256-260. 19. S. Deville, E. Saiz, R. K. Nalla and A. P. Tomsia, Science, 2006, 311, 515-518. 20. E. Munch, M. E. Launey, D. H. Alsem, E. Saiz, A. P. Tomsia and R. O. Ritchie, Science, 2008, 322, 1516-1520. 21. A. C. Arango, S. A. Carter and P. J. Brock, Applied Physics Letters, 1999, 74, 1698-1700. 22. S. Nejati and K. K. S. Lau, Nano Letters, 2010, 11, 419-423. 23. U. Bach, D. Lupo, P. Comte, J. E. Moser, F. Weissortel, J. Salbeck, H. Spreitzer and M. Gratzel, Nature, 1998, 395, 583-585. 24. B. M. Novak, Advanced Materials, 1993, 5, 422-433. 25. N. Jouault, P. Vallat, F. Dalmas, S. Said, J. Jestin and F. Boue, Macromolecules, 2009, 42, 2031-2040. 26. T. G. Gopakumar, J. A. Lee, M. Kontopoulou and J. S. Parent, Polymer, 2002, 43, 5483-5491. 27. T. P. Selvin, J. Kuruvilla and T. Sabu, Materials Letters, 2004, 58, 281289. 28. M. Hashimoto, H. Takadama, M. Mizuno and T. Kokubo, Journal of Materials Science-Materials in Medicine, 2007, 18, 661-668. 29. C. L. Lu, Z. C. Cui, Y. Wang, Z. Li, C. Guan, B. Yang and J. C. Shen, Journal of Materials Chemistry, 2003, 13, 2189-2195. 30. J. B. Schlenoff and G. Decher, eds., Multilayer Thin Films: Sequential Assembly of Nanocomposite Materials, Wiley-VCH, Weinheim, 2012. 31. S. Srivastava and N. A. Kotov, Accounts of Chemical Research, 2008, 41, 1831-1841. 32. D. D. Kulkarni, I. Choi, S. S. Singamaneni and V. V. Tsukruk, ACS Nano, 2010, 4, 4667-4676. 33. M. Lago and M. Araujo, Journal of Colloid and Interface Science, 2001, 234, 35-43. 34. M. F. Zhang, P. Dobriyal, J. T. Chen, T. P. Russell, J. Olmo and A. Merry, Nano Letters, 2006, 6, 1075-1079. 35. T. Sugimoto, X. P. Zhou and A. Muramatsu, Journal of Colloid and Interface Science, 2003, 259, 43-52. 36. J. Hay and B. Crawford, Journal of Materials Research, 2011, 26, 727738. 37. S. M. Han, R. Saha and W. D. Nix, Acta Materialia, 2006, 54, 15711581. 38. R. Saha and W. D. Nix, Acta Materialia, 2002, 50, 23-38. 39. W. C. Oliver and G. M. Pharr, Journal of Materials Research, 1992, 7, 1564-1583. 40. L. Zhang, G. Feng, Z. Zeravcic, T. Brugarolas, A. J. Liu and D. Lee, ACS Nano, 2013, 7, 8043-8050. 41. J. Bachmann, S. K. Woche, M. O. Goebel, M. B. Kirkham and R. Horton, Water Resources Research, 2003, 39. 42. A. Siebold, A. Walliser, M. Nardin, M. Oppliger and J. Schultz, Journal of Colloid and Interface Science, 1997, 186, 60-70. 43. Y.-R. Huang, J. T. Park, J. H. Prosser, J. H. Kim and D. Lee, Journal of Materials Chemistry C, 2014, 2, 3260-3269. 44. E. Langereis, S. B. S. Heil, H. C. M. Knoops, W. Keuning, M. C. M. van de Sanden and W. M. M. Kessels, Journal of Physics D-Applied Physics, 2009, 42. 45. L. J. An, D. Y. He, J. K. Jing, Z. G. Wang, D. H. Yu, B. Z. Jiang, Z. H. Jiang and R. T. Ma, European Polymer Journal, 1997, 33, 1523-1528. 46. P. Claudy, J. M. Letoffe, Y. Camberlain and J. P. Pascault, Polymer Bulletin, 1983, 9, 208-215. 47. E. W. Washburn, Physical Review, 1921, 17, 273-283. 48. N. Fries and M. Dreyer, Journal of Colloid and Interface Science, 2008, 320, 259-263. 49. R. Masoodi, K. M. Pillai and P. P. Varanasi, AIChE Journal, 2007, 53, 2769-2782. 50. T. G. Fox and P. J. Flory, Journal of the American Chemical Society, 1948, 70, 2384-2395. 51. M. Alcoutlabi and G. B. McKenna, Journal of Physics-Condensed Matter, 2005, 17, R461-R524. 52. A. Bansal, H. C. Yang, C. Z. Li, K. W. Cho, B. C. Benicewicz, S. K. Kumar and L. S. Schadler, Nature Materials, 2005, 4, 693-698.

J. Name., 2012, 00, 1‐3 | 7


Journal Name

Nanoscale


53. J. A. Forrest and K. Dalnoki-Veress, Advances in Colloid and Interface Science, 2001, 94, 167-196. 54. J. A. Forrest, K. DalnokiVeress and J. R. Dutcher, Physical Review E, 1997, 56, 5705-5716. 55. A. Donev, I. Cisse, D. Sachs, E. Variano, F. H. Stillinger, R. Connelly, S. Torquato and P. M. Chaikin, Science, 2004, 303, 990-993. 56. A. Wouterse, S. R. Williams and A. P. Philipse, Journal of PhysicsCondensed Matter, 2007, 19, 406215. 57. K. Ariga, T. Mori and J. P. Hill, Advanced Materials, 2012, 24, 158-176. 58. J. Zhang, X. Wang, L. Lu, D. Li and X. Yang, Journal of Applied Polymer Science, 2003, 87, 381-385. 59. M. T. Byrne, J. E. McCarthy, M. Bent, R. Blake, Y. K. Gun'ko, E. Horvath, Z. Konya, A. Kukovecz, I. Kiricsi and J. N. Coleman, Journal of Materials Chemistry, 2007, 17, 2351-2358. 60. S. P. Thomas, S. Thomas and S. Bandyopadhyay, Composites Part A: Applied Science and Manufacturing, 2009, 40, 36-44. 61. Z. Wang, G. Li, H. Peng, Z. Zhang and X. Wang, J Mater Sci, 2005, 40, 6433-6438. 62. W. P. Lee and A. F. Routh, Langmuir, 2004, 20, 9885-9888. 63. J. H. Prosser, T. Brugarolas, S. Lee, A. J. Nolte and D. Lee, Nano Letters, 2012, 12, 5287-5291. 64. K. B. Singh and M. S. Tirumkudulu, Physical Review Letters, 2007, 98, 218302.



ARTICLE


8 | J. Name., 2012, 00, 1‐3


Page 9 of 9

Nanoscale

Journal Name

View Article Online


TOC Figure



Polymer nanocomposite films with extremely high concentrations of nanoparticles are fabricated using capillary rise infiltration (CaRI).


J. Name., 2012, 00, 1‐3 | 9

Nanoparticle assemblies in supramolecular nanocomposite thin films: concentration dependence.

Rapid synthesis of flexible conductive polymer nanocomposite films.

graphene nanocomposite films.

Capillary wave dynamics of thin liquid polymer films.

Confinement enhances dispersion in nanoparticle-polymer blend films.

DOPA-polymer composites.

Sequential Infiltration Synthesis of Doped Polymer Films with Tunable Electrical Properties for Efficient Triboelectric Nanogenerator Development.

Sintering Inhibition of Silver Nanoparticle Films via AgCl Nanocrystal Formation.

Controlling Hydrogel Mechanics via Bio-Inspired Polymer-Nanoparticle Bond Dynamics.

Graphene Oxide Nanocomposite Films.

Environmental release of core-shell semiconductor nanocrystals from free-standing polymer nanocomposite films.

Gold nanorod length controls dispersion, local ordering, and optical absorption in polymer nanocomposite films.

Quantification of nanoparticle release from polymer nanocomposite coatings due to environmental stressing.

Nanocomposite Hydrogels: 3D Polymer-Nanoparticle Synergies for On-Demand Drug Delivery.

Ribbon curling via stress relaxation in thin polymer films.

Selectively Patterning Polymer Opal Films via Microimprint Lithography.

Silver nanoparticle-mesoporous oxide nanocomposite thin films: a platform for spatially homogeneous SERS-active substrates with enhanced stability.

Free-Standing Conducting Polymer Films for High-Performance Energy Devices.

Bio-inspired high-performance antireflection and antifogging polymer films.

Creep behavior of starch-based nanocomposite films with cellulose nanofibrils.

Supermacroporous polymer monolith prepared with polymeric porogens via viscoelastic phase separation for capillary electrochromatography.

Novel hybrid organic thermoelectric materials:three-component hybrid films consisting of a nanoparticle polymer complex, carbon nanotubes, and vinyl polymer.

Durable, superoleophobic polymer-nanoparticle composite surfaces with re-entrant geometry via solvent-induced phase transformation.

polymer-coated nanocrystals with extremely high stability used as a color conversion material in LEDs.