Abstract
Polymer–inorganic hybrid nanocomposites exhibit enhanced material properties, combining the advantages of both their organic and inorganic subcomponents. Extensive research is being carried out to functionalize polymers towards various improved physicochemical characteristics such as electrical, optical, and mechanical properties for various applications. Vapor-phase material infiltration is an emerging hybridization route, derived from atomic layer deposition, which facilitates uniform incorporation of inorganic entities into a polymer matrix, leading to novel applications in fields such as microelectronics, energy storage, smart coatings, and smart fabrics. In this article, recent advances in employing vapor-phase material infiltration as a hybridization and nanopatterning technique for various application avenues are reviewed.
Similar content being viewed by others
Introduction
Polymers are a class of materials that have found widespread applications due to their easy and low-cost processability and mechanical flexibility. However, compared with their inorganic counterparts, in some instances, polymers often have electrical, optical, catalytic, and mechanical properties that do not meet the performance required for targeted applications. Hybridization of polymers with other, inorganic materials, such as metals and metal oxides, either by simple blending or direct synthesis, has been investigated to improve polymer properties. The motivation behind this approach is to combine the properties of all the subcomponents at the molecular level to generate a new hybrid material with properties unlike those of the starting materials; For example, the ultraviolet (UV) light absorption of poly(methyl methacrylate) (PMMA) increases drastically after hybridization with ZnO, and the hybrid nanocomposite can be used as a flexible and transparent UV shield.1 A number of techniques such as sol–gel synthesis, intercalation method, in situ polymerization, and direct mixing have been reported for functionalization of polymers to form nanocomposites for use as alternative materials with enhanced performance.2
One of the promising hybridization techniques that has emerged recently is infiltration synthesis, a vapor-phase material infiltration process derived from atomic layer deposition (ALD). During infiltration synthesis, polymers are exposed to vapor-phase metalorganic precursors, which infiltrate into the polymer matrix and react with functional groups within the polymer to form hybrids.3 For the hybrid material to exhibit mixed properties of its subcomponents, their uniform distribution throughout the matrix is critical, and this can be achieved via vapor-phase material infiltration using ALD.4 Over recent years, different variations of this methodology have been developed for exposing polymers to metalorganic precursors during the infiltration process including: (a) sequential infiltration synthesis (SIS),3,4 (b) sequential vapor infiltration (SVI),5,6 and (c) multipulse infiltration (MPI).7,8 The crucial difference between conventional ALD and infiltration synthesis lies in the interaction between the target substrate and the precursor. The typical ALD process is governed by precursor adsorption onto the substrate surface, while infiltration synthesis relies upon precursor diffusion into the substrate (i.e., polymer) matrix. The variation in the infiltration process parameters, such as the precursor exposure time, purge time, exposure pressure, and temperature, dictates the difference between these reported infiltration synthesis protocols. Typically, infiltration experiments are conducted in the temperature range of 50°C to 150°C and below the glass-transition temperature (Tg) of the polymer being used. Comparison of the temporal pressure profiles of various infiltration protocols illustrates their differences. Figure 1 shows typical temporal pressure profiles of SIS and SVI in comparison with the standard ALD process. During the normal ALD process, carrier gas such as N2 or Ar flows continuously, and the precursors [metalorganics (e.g., trimethylaluminum (TMA) for AlOx) and oxidants (e.g., H2O, O3)] are briefly pulsed (for tens of milliseconds) alternatively, with each pulse being separated by a short waiting period (of a few seconds) to complete one ALD cycle (Fig. 1, top); these procedures result in precursor adsorption on the substrate surface and ultimately formation of a conformal monolayer of a given material per ALD cycle.3 In contrast, in the SIS process, the precursor pulsing cycles are initiated under a static vacuum (i.e., no carrier gas flow, with the chamber being isolated from pumping) followed by a long exposure period (up to tens of minutes or hours) before reevacuating and purging the chamber; this allows the precursors to infiltrate into the polymer substrate matrix (Fig. 1, middle). The SVI protocol is differentiated from SIS in that the half infiltration cycle of a given precursor is repeated multiple times (with each being separated by an evacuation and purge), followed by similar repeated half infiltration cycles of a counter-reactant (Fig. 1, bottom).5 More detailed mechanistic aspects of each infiltration protocols have been extensively reviewed by Leng and Losego recently.9
Schematic descriptions of temporal pressure profiles comparing the normal ALD (top), SIS (middle), and SVI processes (bottom). A (red) and B (blue) denote organometallic precursors and counter reactants (e.g., oxidants), respectively
Different aspects of infiltration synthesis have been recently reviewed, focusing on (a) applications of hybrids generated by infiltration into natural and synthetic polymers,10 (b) infiltration reaction mechanisms,11 and (c) infiltration processing kinetics and applications based on material properties.9 The present review primarily focuses on highlighting recent advances in use of infiltration synthesis for functional material applications. An overview of the vapor-phase infiltration process as a material hybridization technique for functionalizing polymers is presented first. The subsequent sections concisely summarize the enhancements in polymer properties and various functional applications enabled by infiltration synthesis.
Applications of Materials Infiltration
Direct Material Patterning by Self-assembled Block Copolymer (BCP) Templates
Lithography-based patterning has been a workhorse of the semiconductor industry for decades in achieving higher computing performance by increasing the device areal integration density and keeping Moore’s law going. However, in recent years, this trend has slowed down due to several technological challenges, with cost-effective patterning at sub-30-nm dimensions remaining one of the crucial ones. In general, large-scale, low-cost nanopatterning of functional materials is of great interest for fabrication of not only electronic devices but also other functional devices, including energy harvesters, sensors, and drug delivery agents.12,13,14,15,16 In a typical lithography step, a thin photosensitive polymer layer (i.e., photoresist) is spin-coated onto a substrate (e.g., Si) then exposed to UV light through a mask to chemically modify desired areas of the resist. The exposed (or unexposed) polymer region can be selectively removed by using suitable chemical developers to obtain targeted patterns, which can be further transferred to the bottom substrate by, e.g., metallization or reactive-ion etching.
Nontraditional, self-assembled BCP-based nanopatterning has drawn widespread interest because of its ability to spontaneously form regular patterns with dimensions smaller than 30 nm. However, transforming these small polymeric patterns into useful structured inorganic materials is challenging; for instance, the etch resistance of small thicknesses of BCP thin films (< ~ 30 nm) and other high-resolution polymer resists is insufficient for transfer of the polymer patterns to high-aspect-ratio inorganic patterns via etching-based techniques. Patterning of metal and metal oxides by vapor-phase infiltration synthesis onto self-assembled BCP thin films has been extensively investigated over the past few years as a potential solution to address both the resolution and etch resistivity challenges in lithographic patterning. BCP-based infiltration synthesis is particularly attractive due to the easy material processing, which can be performed on large scale without resorting to expensive conventional lithography facilities.12
A detailed description of the mechanism governing the BCP self-assembly process has been previously reported12,17 and is briefly summarized here. A BCP consists of two chains (conversely, blocks) of polymers that are covalently bound but immiscible with each other. The most commonly used example is the diblock copolymer of polystyrene (PS) and PMMA (commonly referred to as PS-b-PMMA). The two immiscible blocks of a BCP can be driven to phase separate in their molten states by repulsion, which is controlled by the degree of polymerization, composition, and Flory–Huggins interaction parameter (χ). These three parameters control the free energy of the BCP system as expressed mathematically by the Flory–Huggins equation:
where \( \Delta G_{\text{mix}} \) is the free energy of the melt, \( k_{\text{b}} \) is the Boltzmann constant, N is the molecular weight of the polymer block, and f is the volume fraction of a given polymer block. By varying N, f, and T, BCP thin films spin-coated on a substrate with suitable thicknesses can self-assemble into different geometries such as (a) parallel cylinders, (b) vertical cylinders, (c) parallel lamella, (d) vertical lamella, and (e) spheres (Fig. 2).
Various orientations and morphologies of BCP domains with respect to a substrate plane obtained after self-assembly: (a) parallel cylinders, (b) perpendicular cylinders, (c) parallel lamella, (d) perpendicular lamella, and (e) spheres.
Such self-assembled BCP geometries can then be used as templates for domain-selective inorganic loading using infiltration techniques. The most common example of metal oxide infiltration is AlOx infiltration synthesis into the PMMA block of self-assembled PS-b-PMMA thin-film templates. This infiltration process exploits the selective reactivity of the Lewis-acidic metalorganic Al precursor (TMA) with the Lewis-basic carbonyl and ester groups present within the PMMA domain. The relatively inert PS domain does not react with the infiltrating precursor, and this asymmetric reactivity of TMA enables AlOx to be infiltrated into the PMMA block. The selectively infiltrated TMA in the PMMA domain undergoes hydrolysis upon subsequent exposure to H2O vapor, forming AlOx inside the PMMA domain. Further repeated sequential exposures to TMA and H2O generate more -Al-O- groups, increasing the AlOx content inside the PMMA domain. After completion of the infiltration cycles, the newly formed PMMA-AlOx hybrid composite can be further treated by oxygen plasma etching to preferentially remove the organic polymer matrix and finally yield AlOx nanopatterns inherited from the starting self-assembled PMMA domain morphology.3
Numerous reports have demonstrated patterning using BCP self-assembly and SIS of AlOx nanostructures such as nanowires (fingerprint pattern), nanospheres, and nanodots, with controlled feature size below 30 nm. Peng, Darling, and Elam et al. have shown the size tunability of AlOx nanowire patterns generated by self-assembled PS-b-PMMA BCP thin-film templates from 8 nm to 30 nm by controlling the number of SIS cycles (from 1 to 10).3
In addition to AlOx, various other materials can also be patterned by applying the SIS process to self-assembled BCP templates. Peng, Darling, and Elam et al. selectively modified the chemical reactivity of the PMMA block in the self-assembled PS-b-PMMA template for infiltration synthesis of other metal oxides and metals such as ZnO, SiO2, and W, where the weak interaction of material precursors with PMMA was overcome by pre-infiltration synthesis of ~ 1 to 3 cycles of AlOx (i.e., AlOx priming) to create -Al-CH3 and -Al-OH reactive sites within the PMMA block.18 The presence of these reactive sites increased the reactivity of PMMA towards various material precursors, including diethylzinc (DEZ), SiCl4, and WF6.
Kamcev and Nam et al. reported a chemical modification technique to infiltrate ZnO, TiOx, and AlOx into the PS domain, rather than the PMMA domain, of self-assembled PS-b-PMMA templates without relying on AlOx priming, thus generating metal oxide nanopatterns inverted from those obtained from normal infiltration synthesis in unmodified BCP templates.19 Exposure of the self-assembled PS-b-PMMA film to UV radiation (~ 0.8 mW/cm2) in presence of a small amount (< 5%) of oxygen initiated photooxidation, decomposing phenyl rings available in the PS block. This resulted in various new chemical moieties in PS, such as hydroxyl, carboxyl, carbonate, and carbonyl groups, which increased the Lewis basicity of the PS domain and enhanced its reactivity with otherwise weakly reacting organometallic precursors (such as DEZ for ZnO infiltration synthesis). Meanwhile, UV exposure of the PMMA block led to breaking of the ester groups and formation of weakly reactive saturated carbon-carbon bonds. Consequently, the precursor could not infiltrate into the PMMA block, thus augmenting the selective infiltration synthesis of metal oxides in the PS domain of UV-modified PS-b-PMMA BCP templates.
One of the advantages of utilizing BCP self-assembly is the ability to sequentially assemble multiple BCP layers and induce an ordered structure, where adjacent stacked layers are self-registered to each other. Rahman and Black et al. have carried out extensive studies on combining different BCP phases such as cylinders, lamella, and inverse cylinders, generating unique material pattern combinations at nanoscale;20 For example, PS-b-PMMA was first self-assembled into a lamellar phase, followed by selective infiltration of AlOx into the PMMA domain (four cycles of TMA and H2O at 85°C), which renders the layer insoluble in appropriate solvents, thus allowing spin-coating of additional BCP layers on top. Additionally, infiltration synthesis created a subtle modulation of the surface topography, guiding the self-assembly of the subsequent BCP layer (called responsive layering). Repeated responsive layering of multiple self-assembled BCP layers, enabled by the infiltration process, eventually resulted in a library of unique three-dimensional (3D) self-aligned metal oxide nanomorphologies (Fig. 3) once the polymer matrix was removed by oxygen plasma etching.
(a) Scanning electron micrographs (top view) of various nanomorphologies obtained by performing iterative PS-b-PMMA BCP thin-film self-assembly and AlOx infiltration. (b) Schematics depicting anticipated self-aligned 3D nanomorphologies for given BCP layering combinations. BCP layer naming convention: AXY, where A = C (cylinders), L (lamellar), or O (inverse cylinders), with XY = BCP molecular weight in units of kg/mol. Scale bar denotes 100 nm.
Infiltration-Modified Polymers for Lithographic Pattern Transfer
Lithography-based patterning techniques are the cornerstone in the semiconductor industry for patterning Si-based microelectronics devices. The pattern generated on the photoresist is transferred onto substrates via chemical or plasma etching techniques by using the patterned photoresist as an etch mask. Carbon-based photoresists typically show poor etch resistance and are thus insufficient for fabrication of high-aspect-ratio structures on the substrate by etching-based pattern transfer. Traditionally, sacrificial hard mask layers made of metals or metal oxides are used underneath the photoresist to first transfer the patterns into the hard mask and then subsequently use the patterned hard mask to enable high-aspect-ratio etching of substrate materials. Tseng, Darling, and Elam et al. have shown that direct SIS of metal oxides in the patterned photoresist could significantly enhance the etch resistance of photoresists, therefore facilitating etching-based, high-aspect-ratio pattern transfer onto the Si substrate without relying on a sacrificial hard mask underlayer.21,22,23
Specifically, improved etch resistance of PMMA and commercial ZEP520A polymer resists patterned by electron-beam (e-beam) lithography was demonstrated by Tseng, Elam, and Darling et al. via AlOx infiltration using the SIS process;21,22 e-beam-patterned PMMA and ZEP520A resist structures were subjected to five cycles of AlOx SIS (TMA and H2O), and the resulting AlOx-infiltrated hybrid polymer pattern was used as an etch mask to transfer the patterns into the underlying Si substrate (Fig. 4). The etch rate of the PMMA-AlOx hybrid was reduced to 8 nm/min from 300 nm/min for the unmodified PMMA resist. A more modest decrease in etch rate was observed for the ZEP-AlOx hybrid, to 20 nm/min from 100 nm/min for unmodified ZEP. Si trench depths of over 300 nm were realized using the SIS-modified polymer resist, much deeper compared with the 25-nm-deep trenches etched using unmodified resists for given trench width (50 nm). Similar to the AlOx-infiltrated, lithographically patterned polymer resist, self-assembled BCP thin-film templates displayed significantly increased etch resistance once block-selectively infiltrated with AlOx;23 self-assembled PS-b-PMMA thin film modified by AlOx SIS enabled transfer of BCP patterns into various inorganic substrates made of, e.g., Si, indium tin oxide (ITO), and permalloy (Ni0.8Fe0.2).
Fabrication scheme based on SIS-modified polymer resist etch mask: (a) spin-coating resist on Si substrate, (b) pattern exposure by e-beam lithography and (c) development, (d) AlOx infiltration by SIS, (e) AlOx breakthrough plasma etching to reveal substrate, (f) plasma etching of Si substrate to desired depth, and (g) etch mask removal using wet chemical etching.
Infiltration-Synthesized Metal Oxides for Electronic Devices
Although lithography-based, top-down approaches have been successfully implemented for Si-based nanofabrication, patterning semiconducting metal oxides such as ZnO and TiO2 at nanoscale has remained a challenge. Patterning by combining a self-assembled BCP template and the SIS process is an alternative route for achieving nanopatterned metal oxides as discussed above, but utilization of this infiltration-synthesis-based patterning technique for electronic device applications has been scarce to date. Nam et al. demonstrated AlOx- and TiOx-based high-aspect-ratio nanostructures with arbitrary morphologies and sub-40-nm linewidth by applying metal oxide infiltration synthesis in e-beam-patterned, epoxy-based negative-tone SU-8 resist;24 For example, an e-beam-patterned SU-8 structure was infiltrated with TiOx at 95°C using titanium(IV) isopropoxide (TIP) and H2O as precursors, and subsequent removal of the organic polymer matrix by oxygen plasma created TiOx nanostructures with sub-35-nm linewidth and high aspect ratio (> 10). Dandley and Parsons et al. have also demonstrated infiltration synthesis of AlOx structures using e-beam-patterned PMMA templates, having potential applications in large-scale patterning of electronic isolation or contacts in advanced electronic devices.25
The first demonstration of electrical functionality of metal oxide nanostructures generated by infiltration synthesis was based on infiltration synthesis of ZnO in e-beam-patterned SU-8 patterns, as reported by Nam et al.;26 polycrystalline ZnO nanowires of ~ 40 nm width with controllable length and positional registry were patterned by infiltrating DEZ and H2O in SU-8 nanowire patterns and removing the polymer matrix by oxygen plasma and rapid thermal annealing. The electronic properties of these ZnO nanowires were characterized by fabricating field-effect transistors (FETs). Drive currents of ~ 120 nA at drain–source voltage (VDS) of 2.5 V and gate voltage (VG) of 60 V were reported with corresponding current density of 107 mA/m2 through the nanowire cross-section. Carrier concentration of ~ 2.5 × 1019 cm−3 and field-effect electron mobility of ~ 0.07 cm2 V−1 s−1 were estimated for these infiltration-synthesized ZnO nanowires, where the high carrier concentration and relatively low mobility most likely stem from oxygen vacancies and polycrystallinity, respectively.
Nam et al. further demonstrated that infiltration-synthesized ZnO nanowire arrays could be integrated into high-performance UV photodetectors via fully complementary metal–oxide–semiconductor (CMOS)-compatible processes,27 featuring extreme carrier depletion under dark condition (carrier density < 1017 cm−3), with device photoconductive on/off current ratio and corresponding detector responsivity reaching over 106 and ~ 800 A/W, respectively, at wavelength of ~ 350 nm. Temperature-dependent charge transport study revealed that the charge carrier density, as well as the electron mobility, in the infiltration-synthesized ZnO nanowires were modulated by oxygen chemisorption and desorption at the nanowire surface and grain boundaries; the intragrain electron transport was controlled by thermionic emission across the grain-boundary potential barrier, whose height could be affected by incident photons via photoelectrochemical reaction between chemisorbed oxygen and photogenerated holes. The high surface-to-volume ratio of such ultrathin (< ~ 30 nm), polycrystalline (< 10 nm grain size) ZnO nanowires, synthesized and integrated by infiltration synthesis, illustrates the potential of infiltration synthesis for electronic device applications, such as optical and chemical sensors.
Mesoporous Inorganic Materials Derived from Polymers
Mesoporous metal oxides are utilized as active or supporting components in chemical separation, catalysis, as well as energy storage and conversion applications. Gong and Parsons et al. synthesized well-defined Al2O3 and TiO2 mesoporous material via vapor-phase precursor infiltration using the SVI protocol.28 Three different surfactant solutions, using F127 [triblock copolymers of ethylene oxide (EO) and propylene oxide (PO); EO106PO70EO106], cetyltrimethylammonium bromide (CTAB), and sodium dodecyl sulfate (SDS) with suitable solvents, were spin-coated on Si wafer substrates or nonwoven polypropylene (PP) polymer fibers to form polymer templates for metal oxide infiltration. The samples were then subjected to SVI of AlOx (i.e., TMA and H2O) or TiOx (TiCl4 and H2O) and were subsequently annealed at 400°C to remove the organic template and finally generate free-standing, porous metal oxides. It was shown that the surface properties of the generated mesoporous metal oxides could be tailored by varying the type of surfactant and metal precursor, ultimately achieving specific surface area as high as 1000 m2/g with pore size varying between 3 nm and 10 nm. By using the SVI-prepared mesoporous TiO2 material as an anode layer in a dye-sensitized solar cell, a maximum power conversion efficiency of 8% was achieved.
Nanotextures for Antireflective Coatings
Many optical and optoelectronic components and devices, such as smart windows, solar cells, light-emitting diodes (LEDs), lenses, telescopes, and flat-panel displays, require antireflective coatings to ensure efficient light coupling. Inorganic nanotextures fabricated by selective material infiltration into BCPs and pattern transfer by reactive-ion etching have proven useful for such antireflective coating applications.29,30,31,32 These nanotexture coatings change the surface refractive index, reducing the amount of light reflected from the surface when properly optimized.
Berman, Elam, and Darling et al. have explored AlOx nanotextures generated by AlOx SIS in self-assembled BCP thin films with subsequent removal of the organic matrix by oxygen plasma;29 self-assembly of the BCP thin film was promoted by the solvent annealing technique, which enabled selective swelling of targeted BCP domains and consequently thick self-assembled BCP films with increased porosity. By varying the number of AlOx SIS cycles, the refractive index of the resulting nanoporous AlOx coating could be tuned from 1.11 to 1.25, much lower than that of bulk Al2O3 (~ 1.76) and the commonly used antireflective MgF2 coating (~ 1.38).
Rahman and Black et al. have demonstrated antireflective, nanowire-textured Si, Ge, and Ag layers obtained by oblique-angle physical vapor deposition on AlOx nanodot seed layers generated by AlOx infiltration synthesis in vertically aligned cylindrical-phase PS-b-PMMA thin film;30 For example, the generated Ge nanowire textures displayed a refractive index reduced down to 1.8, as opposed to ~ 4.5 for continuous Ge film, resulting in < 10% reflectance in the wavelength range of 400 nm to 800 nm. Infiltration synthesis in self-assembled BCP thin films also permits direct patterning of nanotextured surfaces on target substrates; Rahman and Black et al. have fabricated antireflective nanocone structures on Si31 and SiO2,32 resulting in improved Si solar cell efficiency as well as transparency of window glass; the nanocone structure was created directly on the substrate by transferring AlOx nanodot patterns (Fig. 5a) via inductively coupled plasma reactive ion etching of Si (Fig. 5b). By varying the etching gas composition (comprising SF6:O2 or HBr:Cl2:O2) and etching time, the height and tapering of the nanocone structure could be controlled, increasing the photovoltaic power conversion efficiency of crystalline Si solar cells from 8.7% to 13.1%31 and achieving glass windows with ultrahigh transparency (< 0.2% reflectivity) applicable to solar cells.32
(a) AlOx nanodots fabricated by applying three cycles of AlOx infiltration synthesis on the vertical cylinder pattern of self-assembled PS-b-PMMA thin film followed by oxygen plasma etching of the organic template. (b) Si nanocone structure with AlOx nanospheres at the tip after 100 s of inductively coupled plasma reactive ion etching. (c) Comparison of measured spectral reflectance for flat Si film, 80 nm-thick Si3N4 film, and Si nanocones with height ranging from 32 nm to 230 nm, along with simulated reflectance of 80-nm-thick Si3N4 film on chemically etched inverted Si pyramids. Scale bars denote 100 nm.
Manipulating Surface Hydrophobicity and Hydrophilicity by Infiltration Synthesis
Surface properties of materials, such as self-cleaning, antiicing, and reduced fluid drag, can be controlled by changing the hydrophobicity and oleophobicity of the surface. Control over these properties can be useful for microfluidic applications; Checco and Black et al. have shown that a Si nanocone structure generated by self-assembled PS-b-PMMA BCP thin films and AlOx infiltration synthesis displayed superhydrophobicity enabled by unique combinations of surface texture and hydrophobic surface chemistry,33,34 achieving large advancing water contact angles above 160°. In situ x-ray diffraction studies revealed that the size and geometry of the surface nanotexture fundamentally affected the robustness of the superhydrophobicity under elevated liquid pressure, where gas trapped within the nanotexture and under the water could undergo complete dissolution into the liquid at high pressure. It was found that this mode of collapse of superhydrophobicity could be reversed upon liquid depressurization only for the conical nanocone structure, as controlled by the kinetics of gas–liquid exchange.
Gong and Parsons et al. also reported improvement of the hydrophilicity of polydimethylsiloxane (PDMS) substrates by combining SVI, ALD, and molecular layer deposition (MLD);35 when PDMS samples were coated with Al2O3 (~ 10 nm) by exposure to100 cycles of normal ALD, the surface was initially rendered hydrophilic with contact angle decreased from 116° to 60°, but became hydrophobic once exposed to ambient air due to outdiffusion of uncrosslinked oligomers through cracks in the Al2O3 film. This issue could be addressed by combining one cycle of AlOx SVI (5 h TMA and 0.5 h H2O exposure), 100 cycles of normal Al2O3 ALD, and 100 cycles of MLD of TMA and glycidol; this created a robust hydrophilic coating with high mechanical modulus and barrier property against outdiffusion of oligomers, achieving a contact angle of ~ 40° that was stable over 200 h.
Nanoporous Inorganic Membranes for Gas Separation
Ceramic porous membranes are of great interest for high-throughput separation of gaseous species based on their molecular size. Greil and Black et al. fabricated a porous Si3N4 membrane using BCP-based lithography after selective modification of BCP via AlOx infiltration synthesis.36 A self-assembled PS-b-PMMA template with inverted-cylindrical morphology (i.e., PS:PMMA molecular weight ratio of 3:7 with PMMA being the major matrix and PS the minor cylindrical phase) was infiltrated with AlOx to form nanoporous AlOx film on a Si3N4 membrane (Fig. 6a). This AlOx nanoporous pattern was transferred to the underlying Si3N4 by reactive etching (Fig. 6b), whose pore size could be reduced to make it suitable for gas separation by applying a conformal Al2O3 coating via normal ALD (Fig. 10c). The resulting membrane exhibited high gas separation selectivity of 7:1 between acetone and ethyl acetate vapor.
(a) Nanoporous AlOx thin film formed on Si3N4 membrane by block-selective infiltration on self-assembled BCP thin film with inverted cylindrical morphology and subsequent polymer removal by oxygen plasma. (b) Nanoporous patterns transferred to Si3N4 membrane by reactive ion etching. (c) Nanoporous Si3N4 membrane with reduced pore size by conformally coating ~6-nm-thick Al2O3 by normal ALD. Scale bar denotes 100 nm.
Enhancing Optical Properties of Polymers by Infiltration Hybridization
Past studies have shown that the optical properties of polymers can be improved by physically mixing them with metal oxides for application in UV shields,37 UV and visible-light absorbers,38 and photovoltaic devices.39 Uniform impregnation of metal oxides within given polymer matrices is critical for not only fully exploiting the properties of both the organic and inorganic material but also creating new material functionalities.
Ocola et al. demonstrated that ZnO-infiltrated PMMA hybrid thin film generated via SIS exhibited enhanced photoluminescence (PL);4,40 PMMA films of various thicknesses were infiltrated with ZnO at 95°C by first applying two cycles of AlOx priming and, subsequently, up to 12 cycles of ZnO SIS using DEZ and H2O as precursors. Energy-dispersive spectroscopy (EDS) analysis revealed uniformly infiltrated ZnO into the PMMA matrix as deep as 300 nm. The ZnO-infiltrated PMMA hybrid composite film exhibited orders-of-magnitude enhanced PL intensity at wavelength of ~ 360 nm, nearly matching the bandgap of ZnO, compared with untreated PMMA. It was suggested that direct band-to-band emission from molecular ZnO (such as dimers and trimers) embedded in the PMMA was responsible for the PL observed in the hybrid PMMA film with a small number of SIS cycles. It was further shown that the PL intensity could be adjusted by patterning the starting PMMA template prior to ZnO SIS.4
Akyildiz and Jur et al. discovered that AlOx infiltration in poly(ethylene terephthalate) (PET) fabrics using the SVI protocol generated new photoluminescent hybrid materials;6 under excitation at wavelength of 355 nm, the AlOx-infiltrated hybrid PET exhibited strong visible emission in the wavelength range of ~ 400 nm to 550 nm, which was absent for the untreated PET sample. As the SVI temperature was increased from 90°C to 150°C, the PL peak wavelength of the hybrids also shifted, from ~ 425 nm to ~ 472 nm, resulting in the changed apparent luminescent color (Fig. 7). Formation of Al-OH chemical groups upon AlOx SVI was suggested to be responsible for the visible emission at 425 nm, while the redshift of the PL peak wavelength to 472 nm in the sample SVI-treated at 150°C was attributed to the reduced amount of AlOx infiltration at higher SVI temperature due to a greater extent of near-surface reaction between the C=O functionality of PET and infiltrating TMA. Such ability to tune the optical properties of PET fabrics by material infiltration could open up potential applications in flexible electronics, wearable devices, and optoelectronic sensors.
(a) Comparison of PET fabric samples before and after AlOx infiltration by SVI process at various temperatures (60°C, 90°C, and 150°C), under ambient light (upper row) or UV illumination (lower row). (b) PL spectra obtained from untreated PET fabric and AlOx-SVI-treated at 90°C and 150°C, respectively (excitation wavelength 355 nm).
Photocatalytic activity is another material property closely related to the optoelectronic characteristics of given materials, as it is strongly influenced by the light absorption and subsequent generation of free charge carriers with reasonable lifetime, which can participate in electrochemical reactions.41,42 Akyildiz and Jur et al. have found that AlOx-SVI-treated PET fabric exhibited significantly enhanced photocatalytic activity compared with untreated pristine PET, enabling photoreduction of aqueous metal salt ions and subsequent deposition of metal particles (Au or Ag) on the hybrid fabric surface once irradiated by UV light.42 There existed an optimal SVI temperature (90°C) for maximal photocatalytic activity, at which the internal PL quantum efficiency was also maximal (~ 30% at wavelength of ~ 372 nm), accompanied by the most strongly enhanced light absorption by the hybrid. Deposition of metallic patterns on the hybrid PET fabric surface was also demonstrated by using masked or rastered UV irradiation.42
Enhanced Triboelectric Properties of Metal-Oxide-Infiltrated Polymers
Triboelectric properties of polymeric materials have potential for use in triboelectric power generators, in which tribological mechanical energy is converted to electrical energy. Specifically, when two polymeric materials having different electron affinities come into contact, there is an exchange of electrons due to their different surface potential, generating a harvestable electrical current.43,44 Controlling the surface charge density of these materials is important for the performance of triboelectric generators. Yu and Wang et al. have shown that AlOx SIS into triboelectric polymers can increase the power output from associated triboelectric generators43 (Fig. 8); when five cycles of AlOx SIS were applied to polymeric triboelectric materials (e.g., PDMS), their charge storage capacity increased as evidenced by the enhanced dielectric constant (e.g., from ~ 2.4 to ~ 2.7 for PDMS before and after AlOx SIS), which was associated with increased electron density and decreased electron affinity. This resulted in a near tenfold enhanced triboelectric voltage output when the SIS-modified PDMS and the untreated control came into contact. Also improved was the mechanical resilience of the AlOx-infiltrated polymers, another material property critical for practical triboelectric power generation.
(a) Schematic representation of charge redistribution between pristine and AlOx-infiltrated (doped) PDMS films upon contact. (b) Triboelectric voltage generated due to charge transfer between two unmodified PDMS (left) and between AlOx-modified PDMS and untreated PDMS (right).
Enhanced Mechanical Properties of Infiltration-Modified Polymers
Polymer fibers and nanofiber composites have potential applications in various areas such as protective clothing, electret filters, wearable flexible sensors, microelectronic interconnects, tissue engineering, implant devices, optical sensors, and biomaterial scaffolds.45 Generally, polymeric materials have high extensibility but suffer from low stiffness and fracture toughness. Hybridization of polymer fibers with inorganic materials by vapor-phase material infiltration is one route for improving their mechanical properties.
Lee and Knez et al. first demonstrated the enhanced mechanical properties of a polymeric fiber (natural dragline silk) by infiltrating various metal oxides (AlOx, TiOx, and ZnO) via the MPI process;7 For instance, AlOx-infiltrated silk fibers showed a twofold increase in failure strength and strain, with a threefold increased fracture toughness compared with nontreated control fiber.46 These improved mechanical properties were explained by consumption of hydrogen bonds by the infiltrating TMA, inducing formation of supplemental amorphous regions; native silk fiber is composed of stiff crystals (so-called β-sheet crystals) and a flexible amorphous matrix of amino acids linked by hydrogen bonds. When the fibers were subjected to AlOx MPI, the infiltrating TMA consumed hydrogen bonds near crystal boundaries, converting parts of the crystalline region to additional amorphous matrix (as confirmed by Raman spectroscopy and x-ray diffraction), likely making the crystals more subjected to a more uniform shear force under tensile loading. Similarly, the mechanical stability of collagen membranes8 and cellulose47 could be improved by metal infiltration.
Padbury and Jur et al. conducted a detailed study to understand the change in mechanical properties of AlOx-infiltrated PET fibers using the SVI protocol;48 PET fibers with different crystallinity, produced by controlling the draw ratio during their melt spinning, were subjected to various AlOx SVI cycles. The tensile strength and elongation at break of AlOx-infiltrated PET fibers increased by 16.5% and 55%, respectively, due to formation of -O-Al-O- intrachain complexes. With increasing fiber crystallinity, the extent of the mechanical property improvement was observed to decrease because diffusion of the precursor (e.g., TMA) into the polymer matrix through the tortuous diffusion pathways became more difficult, reducing the amount of AlOx infiltration.
Hybrid polymer–inorganic nanocomposites can also be important for micro/nanoelectromechanical systems (MEMS/NEMS) resonators and actuators,49 supercapacitors,50 battery electrodes,51 and nanoimprinting52 due to their enhanced mechanical resilience. Using nanomechanical characterization (Fig. 9a), Dusoe, Lee, and Nam et al. demonstrated that AlOx-infiltrated patterned SU-8 nanopillars could exhibit a unique combination of high, metal-like yield strength with modest, foam-like Young’s modulus (e.g., ~ 500 MPa and 7 GPa, respectively for a 16-cycle-infiltrated sample; Fig. 9b), leading to one of the highest elastic strain energy storage capacity (i.e., modulus of resilience) known for high-strength engineering materials, surpassing that of bulk metallic glass (Fig. 9c).53 Considering its lithographic patternability combined with excellent modulus of resilience and low density, the developed AlOx-infiltrated hybrid nanocomposite has potential application in NEMS/MEMS device components.
(a) False-color scanning electron micrograph of in situ nanomechanical testing of AlOx-infiltrated SU-8 nanopillar. Variation of (b) yield strength (σy) and Young’s modulus (E), and (c) modulus of resilience (R), with respect to the number of AlOx infiltration cycles.
Enhanced Oil Adsorption in Polymeric Foams
Material infiltration within polymeric foams has also proven effective for generating new functional surface properties. Barry, Darling, and Elam et al. have demonstrated enhancement of the oil absorption properties of commercially available polyurethane and polyimide foams by infiltrating AlOx and ZnO via the SIS process.54 The infiltrations were carried out at 80°C with the number of SIS cycles varying from 0 to 10. In situ infrared spectroscopy and mass spectrometry revealed the reaction of TMA and DEZ with reactive functional groups within the foams. The SIS-treated foams were grafted with oleophilic silane compounds to functionalize them towards improved oil sorption properties. Sorption test results confirmed that the SIS-treated foams exhibited a superior level of oil-to-water selectivity of 6:1, compared with 1:1 for the untreated foam. Additionally achieved were a 30-fold increase in the sorption capacity for the SIS-treated polyurethane foam and a 90-fold increase for the treated polyimide foam, highlighting the unique surface functionality enabled by vapor-phase material infiltration in polymeric foams. The developed oleophilic foam could be useful for addressing oil spills, which have a devastating environmental impact (e.g., crude oil spills in the ocean).
Application of Infiltration Synthesis in Organic Electronics
Organic semiconductors, such as poly(3-hexylthiophene) (P3HT), have attracted significant attention due to their unique advantages over inorganic counterparts, such as low-temperature solution-based processing, applicability in flexible devices, light weight, and tunable optical transparency.55 Properties of organic semiconductors can be further enhanced by hybridization with other materials for improving device performance.56,57,58 Vapor-phase materials infiltration techniques have been successfully applied to organic semiconductors by hybridizing them with metallic components to enhance their electronic properties. Wang and Knez et al. reported increased electrical conductivity of P3HT after infiltration with MoCl5,56 when 35-µm-thick spin-coated P3HT films were exposed to 100 infiltration cycles of MoCl5 at 70°C, their conductivity increased to ~ 3 S/cm, an impressive five-decade enhancement compared with the value of 1.44 × 10−5 S/cm for untreated pristine P3HT.
Inorganic-infiltrated organic semiconductors have also been investigated for potential application as an active layer in photovoltaic devices. Obuchovsky and Frey et al. have fabricated bulk heterojunction (BHJ) solar cells based on P3HT infiltrated with ZnO;57 150-nm-thick P3HT films were exposed sequentially to DEZ and H2O, up to 75 cycles, using an ALD-type exposure protocol with intermediate N2 purging at 60°C. Under the optimal number of ZnO infiltration cycles (55), the photoconductive transient time was increased to ~ 100 µs, from ~ 0.25 µs for untreated P3HT, indicating immobilization of electrons within shallow traps, preventing electron–hole recombination. Based on the ZnO-infiltrated P3HT, power conversion efficiency of 0.12% has been achieved.
Moshonov and Frey et al. also conducted a similar study utilizing self-assembled P3HT-b- poly(ethylene oxide) (PEO) BCP films by exposing them to up to 80 ALD cycles using DEZ and H2O at 60°C.58 The PEO domain of BCP was found to be preferentially infiltrated with ZnO, creating a P3HT-ZnO hybrid layer. Cross-sectional scanning electron micrographs confirmed ZnO infiltration within the BCP polymer film (Fig. 10). Subsequently, this layer was utilized as the active layer in hybrid solar cells. The short-circuit current density of the films that were exposed to 80 cycles (0.095 mA/cm2) was ~ 3 times higher than that obtained for films exposed to only 60 cycles (0.035 mA/cm2). This was attributed to formation of continuous crystalline ZnO domains, which support electron transport.58
Cross-sectional backscattered scanning electron micrographs of P3HT homopolymer (left column) and P3HT-b-PEO BCP (right column) thin films, with respect to varying ZnO infiltration ALD cycles.
Conclusion
Vapor-phase material infiltration has emerged as a new organic–inorganic hybridization technique derived from ALD, useful for incorporating inorganic materials into organic polymer matrices. Based on the polymer matrix and infiltrating inorganic material, various polymer properties can be enhanced, generating hybrid nanocomposites with new functional applications. The infiltration technique has been exploited for applications in unconventional material nanopatterning, antireflective and superhydrophobic coatings, photocatalysis, PL, oil sorption, gas filters, triboelectric generators, and unconventional synthesis of semiconductors and organic electronics. With further improvement of the understanding of the underlying mechanism, infiltration synthesis is expected to find numerous new innovative applications in the future.
References
Y. Zhang, S. Zhuang, X. Xu, and J. Hu, Opt. Mater. 36, 169 (2013).
W.S. Khan, N.N. Hamadneh, and W.A. Khan, Polymer nanocomposites—synthesis techniques, classification and properties, in Science and applications of Tailored Nanostructures, ed. Professor Paolo Di Sia (Cheshire, United Kingdom: One Central Press, 2016), p. 50.
Q. Peng, Y.C. Tseng, S.B. Darling, and J.W. Elam, Adv. Mater. 22, 5129 (2010).
L.E. Ocola, D.J. Gosztola, A. Yanguas-Gil, H.-S. Suh, and A. Connolly, Proc. SPIE 9755, 97552C-1 (2016).
H.I. Akyildiz, R.P. Padbury, G.N. Parsons, and J.S. Jur, Langmuir 28, 15697 (2012).
H.I. Akyildiz, M. Lo, E. Dillon, A.T. Roberts, H.O. Everitt, and J.S. Jur, J. Mater. Res. 29, 2817 (2014).
S.-M. Lee, E. Pippel, U. Gösele, C. Dresbach, Y. Qin, C.V. Chandran, T. Bräuniger, G. Hause, and M. Knez, Science 324, 488 (2009).
S.M. Lee, E. Pippel, O. Moutanabbir, I. Gunkel, T. Thurn-Albrecht, and M. Knez, ACS Appl. Mater. Interfaces 2, 2436 (2010).
C.Z. Leng and M.D. Losego, Mater. Horizons 4, 747 (2017).
K. Gregorczyk and M. Knez, Prog. Mater Sci. 75, 1 (2016).
G.N. Parsons, S.E. Atanasov, E.C. Dandley, C.K. Devine, B. Gong, J.S. Jur, K. Lee, C.J. Oldham, Q. Peng, J.C. Spagnola, and P.S. Williams, Coord. Chem. Rev. 257, 3323 (2013).
C.T. Black, K.W. Guarini, R. Ruiz, E.M. Sikorski, I.V. Babich, R.L. Sandstrom, and Y. Zhang, IBM J. Res. Dev. 51, 605 (2006).
E. Mills, J. Cannarella, Q. Zhang, S. Bhadra, C.B. Arnold, and S.Y. Chou, J. Vac. Sci. Technol., B 32, 06FG10 (2014).
H.W. Ra, K.S. Choi, J.H. Kim, Y.B. Hahn, and Y.H. Im, Small 4, 1105 (2008).
J. Guan, N. Ferrell, L. James Lee, and D.J. Hansford, Biomaterials 27, 4034 (2006).
Y. Chen, Microelectron. Eng. 135, 57 (2015).
R.A. Segalman, Mater. Sci. Eng. R Rep. 48, 191 (2005).
Q. Peng, Y.C. Tseng, S.B. Darling, and J.W. Elam, ACS Nano 5, 4600 (2011).
J. Kamcev, D.S. Germack, D. Nykypanchuk, R.B. Grubbs, C.Y. Nam, and C.T. Black, ACS Nano 7, 339 (2013).
A. Rahman, P.W. Majewski, G. Doerk, C.T. Black, and K.G. Yager, Nat. Commun. 7, 1 (2016).
Y.-C. Tseng, Q. Peng, L.E. Ocola, D.A. Czaplewski, J.W. Elam, and S.B. Darling, J. Mater. Chem. 21, 11722 (2011).
Y.-C. Tseng, Q. Peng, L.E. Ocola, D.A. Czaplewski, J.W. Elam, and S.B. Darling, J. Vac. Sci. Technol., B 29, 06FG01 (2011).
Y.C. Tseng, Q. Peng, L.E. Ocola, J.W. Elam, and S.B. Darling, J. Phys. Chem. C 115, 17725 (2011).
C.-Y. Nam, A. Stein, and K. Kisslinger, J. Vac. Sci. Technol., B 33, 06F201 (2015).
E.C. Dandley, P.C. Lemaire, Z. Zhu, A. Yoon, L. Sheet, and G.N. Parsons, Adv. Mater. Interfaces 4, 1 (2017).
C.-Y. Nam, A. Stein, K. Kisslinger, and C.T. Black, Appl. Phys. Lett. 107, 203106 (2015).
C.Y. Nam and A. Stein, Adv. Opt. Mater. 5, 1700807 (2017).
B. Gong, D.H. Kim, and G.N. Parsons, Langmuir 28, 11906 (2012).
D. Berman, S. Guha, B. Lee, J.W. Elam, S.B. Darling, and E.V. Shevchenko, ACS Nano 11, 2521 (2017).
A. Rahman, M. Liu, and C. T. Black, J. Vac. Sci. Technol. B Nanotechnol. Microelectron. Mater. Process. Meas. Phenom. 32, 06FE02 (2014).
A. Rahman, A. Ashraf, H. Xin, X. Tong, P. Sutter, M.D. Eisaman, and C.T. Black, Nat. Commun. 6, 1 (2015).
A.C. Liapis, A. Rahman, and C.T. Black, Appl. Phys. Lett. 111, 183901 (2017).
A. Checco, A. Rahman, and C.T. Black, Adv. Mater. 26, 886 (2014).
A. Checco, B.M. Ocko, A. Rahman, C.T. Black, M. Tasinkevych, A. Giacomello, and S. Dietrich, Phys. Rev. Lett. 112, 216101 (2014).
B. Gong, J.C. Spagnola, and G.N. Parsons, J. Vac. Sci. Technol. A Vacuum Surfaces Film. 30, 01A156 (2012).
S. Greil, A. Rahman, M. Liu, and C.T. Black, Chem. Mater. 29, 9572 (2017).
H.M. Shanshool, M. Yahaya, W.M.M. Yunus, and I.Y. Abdullah, AIP Conf. Proc. 76, 136 (2014).
D. Japić, I. Djerdj, M. Marinšek, and Z.C. Orel, Acta Chim. Slov. 60, 797 (2013).
D.C. Olson, J. Piris, R.T. Collins, S.E. Shaheen, and D.S. Ginley, Thin Solid Films 496, 26 (2006).
L.E. Ocola, A. Connolly, D.J. Gosztola, R.D. Schaller, and A. Yanguas-Gil, J. Phys. Chem. C 121, 1893 (2017).
J. Liqiang, Q. Yichun, W. Baiqi, L. Shudan, J. Baojiang, Y. Libin, F. Wei, F. Honggang, and S. Jiazhong, Sol. Energy Mater. Sol. Cells 90, 1773 (2006).
H.I. Akyildiz, K.L. Stano, A.T. Roberts, H.O. Everitt, and J.S. Jur, Langmuir 32, 4289 (2016).
Y. Yu, Z. Li, Y. Wang, S. Gong, and X. Wang, Adv. Mater. 27, 4938 (2015).
Y. Yu and X. Wang, Extrem. Mech. Lett. 9, 514 (2016).
Z.M. Huang, Y.Z. Zhang, M. Kotaki, and S. Ramakrishna, Compos. Sci. Technol. 63, 2223 (2003).
S.M. Lee, E. Pippel, O. Moutanabbir, J.H. Kim, H.J. Lee, and M. Knez, ACS Appl. Mater. Interfaces 6, 16827 (2014).
K.E. Gregorczyk, D.F. Pickup, M.G. Sanz, I.A. Irakulis, C. Rogero, and M. Knez, Chem. Mater. 27, 181 (2015).
R.P. Padbury and J.S. Jur, J. Vac. Sci. Technol. A Vacuum Surfaces Film. 33, 01A112 (2015).
Q.P. Unterreithmeier, T. Faust, and J.P. Kotthaus, Phys. Rev. Lett. 105, 1 (2010).
S. Ramesh, S. Khandelwal, K.Y. Rhee, and D. Hui, Compos. Part B Eng. 138, 45 (2018).
P. He, K. Zhao, B. Huang, B. Zhang, Q. Huang, T. Chen, and Q. Zhang, J. Mater. Sci. 53, 4482 (2018).
S. Kim, S. Hyun, J. Lee, K.S. Lee, W. Lee, and J.K. Kim, Adv. Funct. Mater. 28, 1800197 (2018).
K.J. Dusoe, X. Ye, K. Kisslinger, A. Stein, S.W. Lee, and C.Y. Nam, Nano Lett. 17, 7416 (2017).
E. Barry, A.U. Mane, J.A. Libera, J.W. Elam, and S.B. Darling, J. Mater. Chem. A 5, 2929 (2017).
A. Facchetti, Chem. Mater. 23, 733 (2011).
W. Wang, C. Chen, C. Tollan, F. Yang, Y. Qin, and M. Knez, J. Mater. Chem. C 5, 2686 (2017).
S. Obuchovsky, I. Deckman, M. Moshonov, T. Segal Peretz, G. Ankonina, T.J. Savenije, and G.L. Frey, J. Mater. Chem. C 2, 8903 (2014).
M. Moshonov and G.L. Frey, Langmuir 31, 12762 (2015).
Acknowledgements
The research was in part carried out at the Center for Functional Nanomaterials (CFN), Brookhaven National Laboratory (BNL), which is supported by the US Department of Energy, Office of Basic Energy Sciences, under Contract No. DE-SC0012704.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Subramanian, A., Tiwale, N. & Nam, CY. Review of Recent Advances in Applications of Vapor-Phase Material Infiltration Based on Atomic Layer Deposition. JOM 71, 185–196 (2019). https://doi.org/10.1007/s11837-018-3141-4
Received:
Accepted:
Published:
Version of record:
Issue date:
DOI: https://doi.org/10.1007/s11837-018-3141-4












