Creation of triblock copolymer thin films by combining vapor annealing with a raster spray

-->
Adapted from "Spatial and Orientation Control of Cylindrical Nanostructures in ABA Triblock Copolymer Thin Films by Raster Solvent Vapor Annealing", ACSNano,

Jonathan E. Seppala, Ronald L. Lewis, III, and Thomas H. Epps, VOL. 6 ’ NO. 11 ’ 9855–9862 ’ 2012)

This nanosciece paper describes a novel approach to annealing polymeric thin films, particularly block copolymer thin films.  Self-assembly is a significant phenomenon in these materials because they open the floodgates for designer nanoscale materials for nanoporous membranes, lithographic masks, and nanopatterning/templating applications (the last two have huge implications for the electronics industry).  These three nanotechnologies often exploit morphologies often found in AB diblock and ABA triblock copolymers (spheres, gyroid, and lamellae) because the thermodynamics of bulk self-assembly is relatively well established.    The morphology of bulk block copolymers is influenced by three major factors: the degree of polymerization (N), the interaction parameter (c), and the volume fraction of the blocks (f). 

With thin films, surface energy becomes an additional factor; it can be exploited by thermal annealing to facilitate copolymer self-assembly via the bestowment of mobility to amorphous regions that are trapped upon casting.  However, thermal annealing is limited to copolymer systems where the components have similar γ‘s and are thermally insensitive.  Another technique is solvent vapor annealing (SVA), which grants mobility by effectively reducing the T_g of the copolymers.  It is a powerful technique, but is limited to small quantities that are typical of research labs. 

The objective is to devise a faster method for large-scale production of block copolymer (BCP) nanotechnologies.  The method must enable control over morphology and orientation of BCP thin films.  The authors propose raster solvent vapor annealing (RSVA): solvent vapor from a bubbler system is directed onto a BCP surface, which then creates a SVA zone.  The zone is modified/expanded by a motorized stage moving in a raster fashion.  

RSVA was performed with a THF-rich vapor stream in single or multiple passes over a 100 nm thick poly(styrene-b-isoprene-b-styrene) (SIS) film, with domains of 29 nm.  The RSVA speeds ranged from 500 µm/s to 3 µm/s.   The RSVA process swelled the films due to hydrolysis, so the film thickness was measured by spectral reflectometry.  The swelling increased the thickness to 160 nm, and eventually dried down to nearly the original thickness, although some samples were reported to have residual solvent. 

Several approaches to RSVA were performed.  One was single-pass, where the stage moved under the nozzle once, with only the speed varying. The as-cast film had a lamellar structure consisting of cylinders oriented parallel to the substrate, but with minimal long-range order.   Varying the speed affected the ordering of the lamellar cylinders.  The slower the speed, the longer the order-range; at 10 µm/s, the cylinders have mostly perpendicular orientation.  The 10 µm/s speed corresponds to an annealing time of 50 s.  The cylinders even looked slightly swollen, which was confirmed by azimuthally integrated 1D profiles from FFTs of the AFM images.    Even reducing the nozzle diameter still induced the ^cylinders, although at lower speeds. 

           

The post-RSVA morphology is an imbroglio of competing forces. The lower g for polyisoprene (32.0 mJ/m² vs. 40.7 mJ/m²), the majority block, enables wetting of both the free and substrate surfaces, which leads to the propensity for the cylinders to possess parallel orientation.  High RSVA speeds do not change the orientation (see 2a–2c) because the surface energy difference was too large for entropy to take effect.  The slow raster allowed enough solvation to lower the differences in g, which lets a) entropic effects to manifest, and to compensate for the stretching experienced by the cylinders during swelling and deswelling. 

              Briefly, Seppela et al tried two more approaches.  One is multiple passes under the nozzle.  Retracing the RSVA pathways altered the cylindrical orientation toward perpendicularity.  The other is a crossed-path approach; two orthogonal passes cross each other, and the result is a domain dominated by perpendicular cylinders.  Both approaches are supported by crisp AFM phase images. 

            This is a wonderfully written paper, but it helps that I have significant background in polymer chemistry.  Note that only THF was used in the vapor stream; it’s natural to ask if this approach has been done with other solvents, and other BCPs.  d_THF = 18.1 (MPa)^1/2, d_polyisoprene = 16.2 (MPa)^1/2, d_polystyrene =  18.6 (MPa)^1/2, so utilizing similar solubilities is a probable reason for the RSVA setup described in the paper.  One should also ask if this technique can be done for other self-assembling polymeric thin films. Seppela et al noted that this annealing method can be altered according to slit geometries, solvent quality, and substrate temperature–indications of much promise for RSVA.  

Confined Crystallization of Polyethylene Oxide in Nanolayer Assemblies

We live in an era of increasing reliance on the very small to satisfy humanity’s endless needs and desires for new technologies.  Nanotechnology manifests itself in numerous scientific fields, and polymer chemistry is no exception.  Polymers are generally amorphous, but polymer crystallinity can be observed if the conditions are right.  Semi-crystalline polymer chains (possesses crystalline and amorphous phases) such as polyethylene and nylon are often used as barrier films in food, medicine, and electronics industries.  A barrier is considered highly efficient if small gas molecules are relegated to permeating through only the amorphous regions of the chains (crystalline regions are impenetrable).   Efficiency can be fine-tuned by varying the polymer-film processing conditions to suit the desired amount of crystallinity and chain orientation.  Polymer films can now be made thin enough to effectively confine the crystallization process to 2D; this leads to surprising results. 

Conventionally, confined polymer chains crystallize into lamellae with thicknesses of ~10-20 nm with spherelitic morphology.   However, this convention is skirted at the nanoscale, as isotropic growth is severely hampered to the point of producing lamellar crystal orientation.  This orientation is usually perpendicular to the layer (edge-on), but parallel orientations have been reported several times in the literature; mechanisms for orientation determination remain mysterious for the time being. 

Normally, researchers prepare 2D crystallization of polymers via solution processes such as spin-coating or Langmuir-Blodgett (LB) techniques, but these are limited by the solvent requirement and the small quantity of material fabricated.  LB techniques enable layered nm morphologies due to microphase separation of dissimilar block copolymers within the thin films.  Alas, block copolymers are notoriously difficult to synthesize and align with respect to the direction of the thin films. 

Enter a new technique known as layer-multiplying extrusion.  It uses forced assembly to create alternating layers of two polymers that number up to the 100,000s.  Almost any melt-processable polymer can be formulated into kilometers of nanolayered films with thicknesses of ~10 nm.  With less material comes an explosion of new previously unknown properties (“less is more”). 

The materials used in this study are polyethylene oxide (PEO, also known as polyethylene glycol), which has the following structure:

                                                   HO-CH₂-(CH₂-O-CH₂-)_n-CH₂-OH

The other is ethylene-co-acrylic acid (EAA), a copolymer with much lower crystallinity than PEO:  

Films with 33, 257, and 1025 alternating EAA and PEO layers were extruded, with various thicknesses and composition ratios, including (EAA/PEO vol/vol) 50/50, 70/30, 80/20, and 90/10.  The nominal PEO layer varied from 3.6 µm to 8 nm. 

The films were subjected to oxygen permeability tests with respect to to layer thickness.  The results are shown below:

Fig. 1 The effect of layer thickness on oxygen permeability. (A) Oxygen permeability of films with equal volume fractions of EAA and PEO. The dashed line indicates P_// calculated from Eq. 1. (B) Oxygen permeability of the PEO layers from films of varying composition calculated from Eq. 2. The dashed line indicates PPEO. The open symbol is for a film with PEO layer breakup. The solid lines are drawn to guide the eyes. 

The plots show a significant decrease in O₂ permeability.  Gas permeability for layered assemblies is modeled by the following equation. 

     (1)

where 𝜙_PEO is the volume fraction of PEO and P_PEO and P_EAA are the permeabilities of PEO and EAA, respectively.  Upon plugging determined values of P_PEO and P_EAA from literature into Eq. (1), the result did not agree with the findings reported in the plot above. Eq. (1) predicts increasing permeability with respect to decreasing PEO thickness, but the data show the opposite trend. Eq. (1) was then modified to account for the apparent sensitivity to P_PEO due to the far lesser permeability of PEO; it still did not agree with the plotted data with the exception of thicker PEO layers as indicated by the dashed line.  Clearly, the PEO nanolayers possess some previously unknown crystalline morphology that bestowed them with staggeringly low permeability.  However, differential scanning calorimetry revealed that the PEO and EAA layers (even the very thin ones) share the same melting enthalpy and melting temperature as the control films; this means that the changes in crystalline morphology granting the PEO nanolayers low permeability was not accompanied by changes in crystallinity nor lamellar thickness. 

Upon examination by AFM, the authors found that the thin 20 nm PEO layers exhibited single lamellae that extended beyond the field of the AFM image.  The single lamellae are said to be very large single crystals.  Reducing the PEO layer thickness to 8 nm then induces breakage, thereby increasing the permeability.  Fig. 2 below shows the AFM image of the 20 nm PEO layer, and an accompanying schematic showing a gas diffusion pathway through the layered assembly.

Fig. 2  AFM phase images of partial cross sections of the layered EAA/PEO films. The PEO layer has substantially higher crystallinity than the EAA layers and hence appears bright in the AFM images. (A) A low-resolution image of an EEA/PEO film with 50/50 composition, 33 alternating layers, and nominal PEO layer thickness of 3.6 mm. (B) A higher-resolution image showing the spherulitic morphology of the 3.6-mm-thick PEO layer. (C) A low-resolution image of an EAA/PEO film with 70/30 composition, 1025 alternating layers and nominal PEO layer thickness of 110 nm. (D) A higher-resolution image of the 110-nm-thick PEO layers showing the oriented stacks of PEO lamellae. (E) A high-resolution image of an EAA/PEO film with 90/10 composition, 1025 alternating layers, and nominal PEO layer thickness of 20 nm showing that the PEO layers crystallized as single, extremely large lamellae. (F) A schematic showing the gas diffusion pathway through the layered assembly with 20-nm- thick PEO layers. The arrows identify the EAA layers and PEO layers. 

The lamellar crystalline region is considered impermeable, with the lamellar fold surfaces constituting the permeable amorphous regions.  As seen in Fig. 2, the gas pathways depend on the frequency of defects such as lamellar edges.  The permeability is now expressed by

   (2)

 where α is the aspect ratio of the impermeable platelets (length/width), and 𝜙 is the volume fraction of impermeable platelets; the platelets are orientated perpendicular to the flux.  For the thinnest PEO layers, the aspect ratio was as high as 120, which meant the lamellae extended up to 2 µm for the 20 nm thick layers.   Gradually thickening the PEO layer relaxed the restrictions on 3D growth, which returned the morphology to spherelitic.   The results were further confirmed by small-angle x-ray scattering (SAXS) and wide-angle x-ray scattering (WAXS). 

This work is a major breakthrough in polymeric applications for nanotechnology because it shows experiment trumping theory, and possibly describes a major advance for gas-barrier films.  Its importance is amply demonstrated by the 51 citations it has generated since its publication in 2009.  Science Magazine accepted the paper because of its reliance on well-established analytical techniques (AFM, differential scanning calorimetry, SAXS, WAXS), and, more importantly, because of its broad significance in the field of nanoscience. 

This significance is underscored by the novel utilization of a relatively new technique–coextrusion–on readily available polymers to engineer nanolayered polymeric formations in sufficient amounts to allow for probing links between the confined crystalline morphology and the properties exhibited.  This opens up new possibilities for packaging methods, i.e., incorporating polymer nanolayers into common polymeric films for less cost, thereby reducing the environmental and energy consequences.