Self-Assembled Copolymer Adsorption Layer-Induced Block Copolymer Nanostructures in Thin Films.

In polymer thin films, the bottom polymer chains are irreversibly adsorbed onto the substrates creating an ultrathin layer. Although this thin layer (only a few nanometers thick) governs all film properties, an understanding of this adsorbed layer remains elusive, and thus, its effective control has yet to be achieved, particularly in block copolymer (BCP) thin films. Herein, we employ self-assembled copolymer adsorption layers (SCALs), transferred from the air/water interfacial self-assembly of BCPs, as an effective control of the adsorbed layer in BCP thin films. SCALs replace the natural adsorbed layer, irreversibly adsorbing onto the substrates when other BCP is additionally coated on the SCALs. We further show that SCALs guide the thin film nanostructures because they provide topological restrictions and enthalpic/entropic preferences for a BCP self-assembly. The SCAL-induced self-assembly enables unprecedented control of nanostructures, creating novel nanopatterns such as spacing-controlled hole/dot patterns, dotted-line patterns, dash-line patterns, and anisotropic cluster patterns with exceptional controllability.

Introduction

Control of the polymer adsorption at an interface is of particular significance in polymer science because it is believed to determine all physical properties of a material.[1,2] Polymer chains are spontaneously adsorbed onto a substrate when an attractive interaction is present between the polymer chains and the substrate where the adsorption energy per chain is proportional to the number of polymer segments, which can be extremely strong.[3] Because of the intrinsic but strong interactions, polymer chains very close to the substrate are irreversibly adsorbed, often creating an ultrathin layer.[4−7] Polymer chains in the adsorbed layer entail qualitatively and quantitatively different molecular motions with respect to those in the bulk, which in turn, influences the molecular motions of the adjacent polymer chains, resulting in a significant change in the physical properties including the glass transition temperature (Tg),[8−10] relaxation time,[11] physical aging,[12] wettability,[13] crystallization,[14] and viscosity[15] of the film.

An adsorbed layer also exists even when polymer chains have stronger inter- and intramolecular interactions such as those in block copolymer (BCP) systems.[16−19] The BCP can be self-assembled in thin films[20] exhibiting interesting periodic nanostructures,[21,22] which have been widely exploited with the potential for use in bottom-up nanofabrication.[23,24] In BCP thin films, the interfacial control is the most important because it can govern the entire film structure. Thus, numerous efforts have been made to control the BCP self-assembly, changing the surface chemistry of the substrate, e.g., grafting of polymer brushes on a substrate[25] or applying chemoepitaxy.[26] Chemoepitaxy[26−29] has been considered as one of the effective ways to direct the BCP self-assembly and has been widely studied. In chemoepitaxy, the underlying chemical patterns are created on the basis of the polymer chemisorption and cross-linking; thus, a top-down method controls a pattern dimension, integrated with e-beam lithography and the etching process, e.g., trim etching of e-beam patterns. Whereas the physically adsorbed ultrathin layer near the substrate also influences the self-assembling behavior of the BCPs,[18,19] the presence of an adsorbed layer has not been fully elucidated and remains elusive. Therefore, no rigorous attempts have been made to control the adsorbed layer, which is believed to be naturally formed and uncontrollable.

Herein, we introduce an effective method for creating and controlling an ultrastable adsorbed layer in BCP thin films employing an air/water interfacial self-assembly (ISA) of BCPs.[30−34] BCPs from an ISA, which create periodic structures of varying sizes and shapes,[35] are transferred from the interface to the substrate, and we found that the self-assembled BCPs are irreversibly adsorbed onto the substrates, becoming a stable layer, which is defined as a self-assembled copolymer adsorption layer (SCAL). A SCAL prevents the formation of a natural adsorbed layer upon additional BCP coatings, and further guides the self-assembly of BCP nanostructures. In contrast to the chemoepitaxy, the dimension of the underlying SCAL structure is precisely controlled by a bottom-up method based on the ISA of BCPs with inherent controlling parameters, e.g., molecular weights or relative block fraction of BCPs. We show that a SCAL-induced self-assembly of BCP exhibits distinctive nanostructures, which have not been found in a conventional BCP self-assembly and are highly controllable with high precision in terms of the shape and size of the SCALs.

Results and Discussion

Scheme illustrates a strategy for creating SCALs from the ISA of BCPs and achieving a SCAL-induced self-assembly of BCPs. In the first step of Scheme , the amphiphilic BCP is readily self-assembled at the air/water interface by spreading BCP solutions onto the water surface. An ISA occurs quickly, and the morphology obtained is fixed as the spreading solvent quickly evaporates.

Scheme 1

Formation of Self-Assembled Copolymer Adsorption Layer (SCAL) Transferred from Interfacial Self-Assembly (ISA) of BCPs and SCAL-Induced BCP Nanostructures in Thin Films

Amphiphilic BCPs[36−41] are known to form various morphologies of dot-, strand-, and planar-shapes at the air/water interface. Whereas the types of morphologies are generally dependent on the relative block ratio,[41,42] varying the spreading area at a fixed block ratio of BCP provides more versatility, regarding not only the type of morphology[35] but also the periodicity and type of array.[34,43] Among the many amphiphilic BCPs self-assembled at the air/water interfaces, polystyrene-b-poly(2-vinylpyridine) (PS-b-P2VP) was chosen because its phase behavior is well-known[35,44] and exhibits more various structures with a high-quality ordering at a fixed molecular weight.[35]

As shown in Scheme , the self-assembled PS-b-P2VP at the interface is then transferred onto the Si wafer using the Langmuir–Blodgett (LB) technique. The transferred films maintain their original structure at the interface and are strongly adsorbed onto the substrate, forming as a SCAL.

When the additional BCP of polystyrene-b-poly(methyl methacrylate) (PS-b-PMMA) is spin-coated, the SCAL of PS-b-P2VP can guide the self-assembly of PS-b-PMMA because a SCAL provides enthalpic and entropic restrictions in the self-assembly of PS-b-PMMA. Two types of PS-b-PMMA are employed and are known to form a cylinder and lamella in bulk (Table S1). The PS-b-PMMAs were spin-coated onto SCALs and self-assembled through thermal annealing, as shown in the last step of Scheme . We characterized their various SCAL-induced BCP nanostructures using scanning electron microscopy (SEM), atomic force microscopy (AFM), X-ray photoelectron spectroscopy (XPS), time-of-flight secondary ion mass spectrometry (ToF-SIMS), and grazing-incidence small-angle X-ray scattering (GISAXS). Details of the sample preparations and characterization methods are provided in the Supporting Information.

SCAL Stability

An important prerequisite for a SCAL-induced BCP self-assembly is to ensure a good stability of the SCAL. Because BCPs are self-assembled through thermal and solvent vapor annealing requiring a severe change in temperature and exposure to an organic solvent, the SCAL needs to remain stable and maintain its original nanostructure against the additional spin-coating of BCPs and further processing for the SCAL-induced self-assembly. Thus, we first confirmed the chemical and thermal stability of SCAL consisting of PS-b-P2VP with a molecular weight of 44-b-18.5 kg/mol (44–18.5k).

The PS-b-P2VP (44–18.5k) forms dots at the air/water interface, the structure of which after the transfer to the substrate is shown on the left side of Figure a. Then, to test the stability of SCAL adsorption, substrates with PS-b-P2VP (44–18.5k) SCALs were thermally annealed (TA) at 190 °C for 24 h and/or directly immersed in various solvents including deionized water, ethanol, heptane, tetrahydrofuran (THF), toluene, isopropanol (IPA), hexane, and chloroform for 1 h. As shown in Figure a, the SCAL of PS-b-P2VP (44–18.5k) remains stable, preserving the original structure under all conditions; PS-b-P2VP (44–18.5k) SCALs still showed a well-defined nanostructure after thermal annealing at 190 °C for 24 h under a vacuum, even after both thermal annealing and toluene washing.

Figure 1

Stability of SCALs. (a) SEM images of PS-b-P2VP (44–18.5k) SCALs before (pristine) and after immersing in various solvents and/or thermal annealing (TA). (b) Schematic illustration of film transfer from the air/water interface to substrates. (c) N 1s XPS spectra of P2VP homopolymer (37.5k) adsorbed layers obtained from Guiselin’s approach (the first and second) and transferred from the air/water interface (from the third to the sixth) and that of bulk P2VP (the seventh). (d) Calculated adsorption fraction of pyridine and the thickness of the layer.

The superior stability of a SCAL was also confirmed after an additional spin-coating of PS-b-PMMA. PS-b-PMMA (64–35k) was spin-coated onto PS-b-P2VP (44–18.5k), followed by toluene washing. We found that the inherent structure of PS-b-P2VP (44–18.5k) SCAL remained, preventing the creation of a naturally adsorbed layer from PS-b-PMMA (64–35k), as shown in Figure S1. The chemical and thermal stabilities in the SCALs were also confirmed through a GISAXS analysis (Figure S2). The scattering profiles from all SCALs showed well-defined first- and second-order peaks at a ratio of 1:, ensuring that the hexagonal arrays of SCAL nanostructures were maintained against the chemical/thermal treatment. The excellent stability of SCALs was also checked using other BCP types of PS-b-PMMA (Figure S3).

Whereas the high stability of a SCAL can certainly be beneficial in guiding the self-assembly of PS-b-PMMA, the origin of such an excellent stability of a SCAL remains questionable. As illustrated in Figure b, PS-b-P2VP self-assembles at the air/water interface, and P2VP (PS) chains are stretched (aggregated) to maximize (minimize) the contact area on the water surface. This molecular architecture is transferred to the substrates without any structural changes. Prior to a thermal annealing step in SCAL-induced self-assembly, the molecular monolayer of P2VP significantly contributes to the stability of SCAL as compared to the PS aggregates because the stretched P2VP chains have a maximized number of adsorption sites, but PS aggregates are weakly adsorbed to the substrates (Figure S4) despite faster adsorption kinetics of PS.[45] During the thermal annealing step in SCAL-induced self-assembly, PS aggregates are irreversibly adsorbed on the substrate increasing the stability of the SCAL.

We hypothesized that the degree of P2VP chain stretching prepared from an LB can be systemically varied with the surface pressure, π, changing the adsorption affinity to the substrate. P2VP (37.5 kg/mol) was prepared in an air/water interface at a π of 2, 5, 10, and 15 mN/m and transferred to the Si wafer forming an adsorbed layer, which is labeled P2VP(LB). A surface pressure–area (π–A) isotherm curve of P2VP (37.5k) is shown in Figure S5. In comparison, a natural adsorbed layer in a spin-coated thin film was prepared following Guiselin’s approach,[46] which is labeled as P2VP(spin).

To estimate the degree of H-bonding and predict the chain conformation of P2VP, we obtained the high-resolution XPS spectra of N 1s for various P2VP adsorbed layers. Figure c compares the N 1s spectra of the P2VP (37.5k) adsorbed layer from an air/water interface when varying π and that from Guiselin’s approach.[46] The details of the samples used for the XPS analysis are shown in Table S2.

When the P2VP chains are in the bulk (300 nm thick spin-coated film), a main peak is found in the N 1s spectra at 399.1 ± 0.1 eV, which is attributed to a single nitrogen atom of a pyridine ring (the seventh plot shown in Figure c). However, all other P2VP(LB) layers show an additional peak at 401.5 ± 0.1 eV, which is assigned to the hydrogen bonded nitrogen atoms of the N···H–O feature,[47−50] which implies that the transferred P2VP chains from the air/water interface can strongly adsorb onto a silicon wafer through H-bonding, as indicated in Figure b.

On the other hand, a shoulder peak of N 1s at 397.8 eV is more dominant in P2VP(spin)TA, as indicated with the green arrow in the first and second rows of Figure c, respectively. There are various possibilities for the origin of this peak, and the most probable case is that a hydroxyl group of the native oxide layer interacts with the π bond in the aromatic pyridine ring, thus forming a well-known hydrogen bonding of π···H–O.[51−53] It should be noted that the origins of the H-bonding for adsorbed layers can be different depending on how the adsorbed layers are formed: whether they are transferred from the air/water interface or obtained from a spin-coated film. A thorough and complete XPS analysis is provided in the Supporting Information (see the Discussions section; Figures S6–S8, and Table S2).

The H-bonding-based chain conformation has a significant stability; a peak was shown after a severe thermal annealing at 190 °C for 24 h in a vacuum, which is labeled P2VP(LB)TA (Figure S6). We further estimated how many pyridine segments of P2VP chains are adsorbed onto the substrates in the adsorbed layer. The fraction of adsorbed polymer segments was calculated from the peak area ratio of the XPS N 1s spectra, as plotted in Figure d. By increasing π for P2VP(LB), the adsorption fraction gradually decreases, and the layer thickness increases. The P2VP chains are fully stretched at low π maximizing the contact to the substrate. As π increases, the chains start to be compressed; the two-dimensional conformation of P2VP is disrupted, and the chains are practically folded normal to the interface, resulting in a reduced H-bonding capability with water molecules.

The adsorption fraction of P2VP(LB) was varied with a value of π within the range of approximately 8–20%, implying that a significant amount of segments in P2VP chains are strongly adsorbed, whereas that of the conventional irreversibly adsorbed layer obtained from Guiselin’s approach,[46] P2VP(spin)TA (inset of Figure d), was approximately 10–15%.

SCAL from the Air/Water Interface to the Substrates

The SCAL was prepared from the ISA of PS-b-P2VP (44–18.5k) at the air/water interface and transferred to the substrate using the LB technique. Note that all information of the polymers used for a SCAL and spin-coated BCP is given in Table S1. Given a π–A isotherm curve for the ISA with PS-b-P2VP (44–18.5k), a self-assembled PS-b-P2VP (44–18.5k) was transferred at six different values of π of 1, 5, 10, 15, 20, and 30 mN/m (red triangles in Figure ).

Figure 2

π–area isotherm curve of PS-b-P2VP (44–18.5k).

Figure a shows SEM images of the transferred PS-b-P2VP (44–18.5k) films. Whereas PS-b-P2VP (44–18.5k) has a dot morphology at all values of π, the periodicity of hexagonally packed dots gradually decreases, and an abrupt order transition of the dot patterns, from a hexagonal-to-square array, was found as previously reported.[34,43] As π increases from 1 to 15 mN/m, the P2VP chains are compressed, which reduces the wetting area for P2VP on the substrate and decreases the center-to-center distance, L0, of PS aggregates in SCALs. At 20 mN/m, a local-order transition to the quasi-square array[34,43] was found to develop a plateau in the π–A isotherm curve shown in Figure .

Figure 3

SCALs and SCAL-induced self-assembly for asymmetric BCPs. (a) SEM images of PS-b-P2VP (44–18.5k) SCALs from low (left) to high (right) surface pressure. (b) SEM images and (c) AFM height images of SCAL-induced self-assembly of PS-b-PMMA (64–35k) on the PS-b-P2VP (44–18.5k) SCALs obtained at each pressure of panel a. (d) GISAXS 2D images and (e) 1D profiles of PS-b-P2VP (44–18.5k) SCALs in panel a (upper row) and SCAL-induced self-assembly of PS-b-PMMA (64–35k) in panels b and c (lower row). All data in the same column of panels a–e have the same surface pressure of SCAL, from left to right, 1, 5, 10, 15, 20, and 30 mN/m, respectively.

SCAL-Induced Self-Assembly for Asymmetric BCPs

For the SCAL-induced BCP self-assembly, PS-b-PMMA (64–35k) was spin-coated onto SCALs of PS-b-P2VP with a thickness of 25 nm and thermally annealed at 190 °C for 24 h under a vacuum. Whereas the PS-b-PMMA (64–35k) without a SCAL shows a layered lamellar structure on a bare Si wafer (see Figure S9), it exhibits more versatile nanostructures on a SCAL.

Figure summarizes the SCAL-dependent film structure of PS-b-PMMA when increasing the initial π at the ISA. The structure of a SCAL is first given in the first row with an increasing π (Figure a), and the following film structure of PS-b-PMMA is given in the second and third rows (Figure b,c).

At π = 1 mN/m, PS-b-PMMA forms dotted structures similar to the structure of a SCAL but with apparent holes, found in both the SEM and AFM images, as indicated by the white circles in the first column of Figure b,c, respectively. We found that the center-to-center distance, L0, of the PS-b-PMMA structure and that of the SCAL were the same as 126 nm, as confirmed through GISAXS (the first columns of Figure d,e). The detailed peak index of the GISAXS data is given in Figure S10.

As π increases to 5 mN/m, the L0 of a SCAL is reduced to 105 nm, as is that of the PS-b-PMMA structure. Note that additional dots are found in the original holes, as indicated by the arrows, implying that the inner structure of the dots has changed. At 10 mN/m (the third column of Figure ), most of the features of a SCAL-induced self-assembly are the same as that at 5 mN/m, although L0 was reduced to 96 nm.

A more dramatic change of a SCAL-induced self-assembly was found at 15 mN/m (the fourth column of Figure ). The L0 of the PS-b-PMMA self-assembled structure was no longer matched with that of the underlying SCAL. Both the SCAL and SCAL-induced self-assembly show hexagonal dot arrays; however, the feature size of the SCAL-induced BCP structure was remarkably reduced, as indicated by the white dots in the fourth column of Figure a–c. This is confirmed from GISAXS data in the fourth column of Figure d,e. Whereas a SCAL has an L0 of 83 nm, the L0 in a SCAL-induced self-assembly is suddenly reduced to 53 nm, which is similar to the bulk periodicity of the PS-b-PMMA cylinder.[54] Thus, we hypothesized that the perpendicular PS-b-PMMA cylinders appear in this region, as supported by the film structure analysis described in the following section.

At 20 mN/m, where the SCAL shows a partial square array of dot patterns, PS-b-PMMA still holds perpendicular cylinder structures along the aggregated PS clusters (fifth column of Figure a–c). With an increase of π of 30 mN/m, where the SCAL shows a closely packed square array of dot patterns, some PS-b-PMMA cylinders start to orient parallel to the substrate (the sixth column of Figure b,c).

Film Structure Analysis of SCAL-Induced Self-Assembly

The observed SCAL-induced self-assembly shown in Figure can be classified into four structure regimes: (i) holes, (ii) dot-filled holes, (iii) perpendicular cylinders, and (iv) these with some parallel cylinders, respectively. The expected film structures of each regime are illustrated in Figure a where the morphology evolution of the SCAL is depicted in the upper row in Figure a. While enthalpic/entropic interactions between blocks and the SCAL are important in all regimes, an apparent height variation between PS aggregates and the underlying P2VP in regimes 1 and 2 provides an additional topographic restriction to the self-assembly of overlying BCP as well as the enthalpic interactions. As π increases, the P2VP chains start to be compressed and folded normal to the interface. The resulting thicker P2VP domain reduces the height differences as confirmed in Figure S11. Then, the SCAL-induced BCP self-assembly at higher π (regimes 3 and 4) is mainly influenced by the enthalpic contribution of the interfacial energy contrast between PS/P2VP domains in the SCAL.

Figure 4

Expected film structure of the SCAL-induced self-assembly. (a) Schematic illustration of the formation of the SCAL-induced self-assembly of asymmetric BCPs. Topology data from the white dash-line in AFM height images of the (b) first column and (c) third column of Figure c. (d) SEM cross-section image and (e) ToF-SIMS depth profile of PS-b-PMMA (64–35k) on the PS-b-P2VP (44–18.5k) SCAL of 5 mN/m. (f) SEM cross-section image and (g) ToF-SIMS depth profile of PS-b-PMMA (64–35k) on the PS-b-P2VP (44–18.5k) SCAL of 15 mN/m. The purple shaded area having the maximum Si intensity in panels e and g represents the Si wafer.

In regime 1 of Figure a (π = 1 mN/m in Figure ), the SCAL induces a hole structure of PS-b-PMMA. PMMA prefers to sit on P2VP rather than on PS because of the topological restriction in the SCAL (Figure S11) and the smaller difference in the Flory–Huggins interaction parameter, χ, between the blocks of PMMA/P2VP as compared to PS/P2VP.[55] Thus, PS-b-PMMA forms a single lamellar layer on the P2VP chains of a SCAL, although the PMMA is depleted on the PS aggregates of the SCAL, creating apparent hole structures. The hole structures are partially confirmed with SEM (first column of Figure b) and were more clearly confirmed through the AFM height profiles shown in Figure b.

In regime 2 of Figure a (π = 5–10 mN/m in Figure ), the lateral area of the P2VP chains in the SCAL continuously decreases as π increases. After forming a lamellar layer on P2VP, some of the PS-b-PMMA is left because of the reduced P2VP area. The remaining PS-b-PMMA forms a micelle on the PS aggregates in a SCAL; the holes in regime 1 are filled with PS-b-PMMA micelles in regime 2 creating a bumpy dot, as confirmed in the AFM height profile (Figure c). The micelles were also found in the cross-sectional SEM images, where PS was brightened by staining with ruthenium tetroxide. The morphology transition from the regime 1 to regime 2 is not abrupt but gradual; thus, a few holes are found in regime 2.

The depth profiling with ToF-SIMS also provides a clear sign of a PS-b-PMMA micelle. As the sputter time increases, the depth profiles of PS, PMMA, P2VP, and Si were obtained from the surface of the film to the substrate, as shown in Figure e. The PMMA profile shows two peaks, as indicated by the arrows. The first peak near the surface was found to originate from the core of the micelle structures. The intensity increases when approaching the substrate because of the PMMA of the single lamellar layer. The second peak is located immediately before the maxima of the P2VP peak, confirming that a single lamellar layer of PS-b-PMMA was formed on the P2VP chains in the SCAL.

In regime 3 of Figure a (π = 15 mN/m in Figure ), the decreased height variations with the thickly grown P2VP domain in SCAL attenuated the aforementioned topological restriction effect on the SCAL-induced self-assembly (Figure S11). Consequently, the enthalpic contribution from the interfacial energy contrast between PS/P2VP domains in SCAL becomes dominant in the SCAL-induced self-assembly. The lateral area of the PS aggregates of the SCAL becomes greater than that of the P2VP chains of the SCAL, which provides a neutral wetting condition for the PS-b-PMMA cylinders[56] as confirmed in the contact angle measurement in the Figure S12. Thus, PS-b-PMMA cylinders tend to be oriented perpendicular to the substrate.

We confirmed the film structure using both SEM cross-section images (Figure f) and a ToF-SIMS analysis (Figure g). Whereas the alternating dark/bright domains ensure the formation of perpendicular cylinders, the ToF-SIMS data also show the typical depth profiles of perpendicular PS-b-PMMA cylinders; in addition, the depth profiles of PS and PMMA are constant with an increase in the sputter time, as shown in Figure g.

In regime 4 of Figure a (π = 20–30 mN/m in Figure ), the surface of a SCAL is mostly covered by the PS aggregates at a high π, providing a PS-selective wetting condition as confirmed with the contact angle measurement (Figure S12), which allows PS-b-PMMA cylinders to begin to lie parallel to the substrate.

SCAL-Induced Self-Assembly for Symmetric BCPs

A SCAL-induced self-assembly is also confirmed for other BCPs, which have a symmetric block fraction, thus forming lamellar structures in bulk (see Table S1).

The SCALs transferred at π of 1, 2, 5, 7, 10, and 20 mN/m (blue arrows in Figure ) were employed as an underlying layer for a SCAL-induced self-assembly of symmetric PS-b-PMMA (35–37k). PS-b-PMMA (35–37k) was spin-coated onto SCALs with 23 nm thickness but did not show any self-assembled structures on bare substrates (Figure S13) thermally annealed at 190 °C for 24 h under a vacuum.

Symmetric BCPs exhibited a more versatile SCAL-induced self-assembly. The results are shown in Figure a–f. At a π of 1, 2, and 5, PS-b-PMMA (35–37k) forms dot patterns, as depicted in regime 2 in Figure a. Whereas an asymmetric PS-b-PMMA (64–35k) forms micelles on the PS aggregate of a SCAL only, the exact positions of the symmetric PS-b-PMMA (35–37k) micelles vary with the value of π.

Figure 5

SCAL-induced self-assembly for symmetric BCPs. SEM images of the SCAL-induced self-assembly of PS-b-PMMA (35–37k) on the PS-b-P2VP (44–18.5k) SCALs of (a) 1, (b) 2, (c) 5, (d) 7, (e) 10, and (f) 20 mN/m. (g) Schematic description for the location of PS-b-PMMA (35–37k) micelles and PS aggregates of SCAL shown in panels a–c. (h) Plot of average center-to-center distance for PS-b-P2VP (44–18.5k) SCAL nanostructures, L0,SCAL, as a function of surface pressure. Filled squares are calculated values from GISAXS data, and empty squares are extrapolated values; d is the approximate value of the PS-b-PMMA (35–37k) micelle size.

It should be noted that a SCAL shows a periodic dot pattern and that the center-to-center distances of the dots (PS aggregates), L0,SCAL, monotonically decrease with the value of π, as indicated in Figure h.

At π = 1 mN/m (Figure a), a SCAL induces a dot pattern of PS-b-PMMA (35–37k) composed of a single lamellar layer on the P2VP chains and micelles on the PS aggregates, similar to the structure in regime 2 shown in Figure a, thus forming a hexagonal array of dot patterns. The three micelles form a triangle, as drawn in the inset of Figure a, and the side length of the triangle becomes L0,SCAL.

As π increases, a density multiplication of the dot patterns occurs in relation with the size ratio between the micelle size, d (Figure S14), and L0,SCAL. Here, L0,SCAL continuously decreases with an increase in π; however, the value of d remains constant as the micelle size of the PS-b-PMMA (35–37k) does not vary. As L0,SCAL becomes comparable to the micelle size (L0,SCAL ∼ 2d), additional PS-b-PMMA micelles are formed between the original PS-b-PMMA micelles. At π = 2 mN/m, the additional PS-b-PMMA micelles are formed between two original PS-b-PMMA micelles on the side of the triangle PS aggregates of the SCAL created (inset of Figure b).

The second density multiplication of the dot patterns occurs as the L0,SCAL further decreases at π of 5 mN/m. The additional PS-b-PMMA micelles are formed between three of the original PS-b-PMMA micelles at the positions of the PS aggregates of the SCAL (inset of Figure c). As the side length (L0,SCAL) of the triangle is reduced (L0,SCAL ∼ d < 2d), the micelles are no longer located at the sides but rather at the center of the triangles.

Figure g,h describes the ratio (L0,SCAL/d)-dependent density multiplications for the dot patterns. The PS aggregates of a SCAL create an initial dot pattern, and the size of the unit cells decreases with an increase in π. A density multiplication occurs when L0,SCAL is commensurate with the micelle size of PS-b-PMMA (35–37k), d. When L0,SCAL > 2d, such as shown in Figure a, dot patterns are found only on the PS aggregates of the SCAL. Then, the first density multiplication of dot patterns is found when L0,SCAL ∼ 2d, such as in Figure b. The second density multiplication for the dot patterns is shown in Figure c when L0,SCAL ∼ d.

As π increases to 7 mN/m, a SCAL-induced self-assembly for symmetric PS-b-PMMA forms a dot-filled honeycomb pattern, as indicated in Figure d. At a much higher π of 10–20 mN/m, SCALs begin to provide a neutral wetting condition for the symmetric PS-b-PMMA, thus forming a perpendicular PS-b-PMMA lamellar structure, as shown in Figure f, similar to the perpendicular PS-b-PMMA cylinders formed on the SCALs at high π in regime 3 of Figure a.

Effect of SCAL Morphology on SCAL-Induced Self-Assembly (Strand-Forming SCAL-Induced Self-Assembly)

A SCAL can generate various morphologies from the ISA of the BCPs, such as dot-, strand-, planar-, and perforated planar-morphologies depending on the block ratio of the BCPs and the degree of confinement.[35] Here, we found that both dots and strands also provide an interesting topological constraint for a SCAL-induced self-assembly.

To observe the strand-forming SCAL-induced self-assembly, three strand-forming PS-b-P2VP for a SCAL were employed with the same block ratio but different molecular weights (30–12.5k, 44–18.5k, and 79–36.5k; see Table S1). These BCPs form a dot-morphology without an area constraint, but nicely formed a strand-morphology with a proper confinement, as introduced in our previous study.[35]

Figure shows that the width of the strands of a SCAL can systematically vary with an increase in molecular weight of the PS-b-P2VP, namely, narrow, middle, and wide strands, as shown in Figure b–d, respectively.

Figure 6

Effect of SCAL morphology on the SCAL-induced self-assembly. (a) Schematic illustration of the formation of the SCAL-induced self-assembly of asymmetric BCPs on the strand-forming SCALs. SEM images of spreading-area-dependent morphologically transitioned (b) PS-b-P2VP (30–12.5k), (c) PS-b-P2VP (44–18.5k), and (d) PS-b-P2VP (79–36.5k) SCALs. (e–g) SEM images and (h–j) AFM phase images of the SCAL-induced self-assembly of PS-b-PMMA (64–35k) forming (e, h) a dotted-line pattern, (f, i) a dash-line pattern, and (g, j) an anisotropic cluster pattern on corresponding strand-forming SCALs of panels b–d, respectively. Corresponding AFM height images for panels h–j are in Figure S16.

On these strand-forming SCALs, PS-b-PMMA (64–35k) of 25 nm films was spin-coated and thermally annealed at 190 °C for 24 h under vacuum. Figure e–j shows the self-assembled structure of PS-b-PMMA on strand-forming SCALs. Whereas PS-b-PMMA forms a similar self-assembled structure found in a dot-forming SCAL-induced self-assembly, the self-assembled structures are now formed along the curves of the strand.

The width of the strands alters the self-assembling morphology of the PS-b-PMMA (64–35k), as illustrated in Figure a; a SCAL-induced self-assembly of PS-b-PMMA (64–35k) showed a dotted-line pattern (Figure e,h) on the narrow strand of PS-b-P2VP (30–12.5k), a dash-line pattern (Figure f,i) on the middle strand of PS-b-P2VP (44–18.5k), and anisotropic clusters (Figure g,j) on the wide strand of PS-b-P2VP (79–36.5k).

Their expected film structures are shown in Figure a. The PMMA block of the PS-b-PMMA sits on the P2VP chains of the SCAL, creating a single lamellar layer in the same manner as in the dot-forming SCAL. However, in this strand-forming SCAL, the preferential sitting of PS-b-PMMA creates a trench along the strands. The remaining PS-b-PMMAs are then self-assembled, forming various nanostructures inside the trenches. The micelles may not be formed when an insufficient amount of PS-b-PMMA is present, with only trenches being created (Figure S15).

Here, the width of the trench becomes an important parameter chaining the structures. When the width of the trench is comparable to the size of the micelles, an array of single micelles is formed, showing a dotted-line pattern, as indicated in Figure e,h. As the trench is widened (middle strand), the PS-b-PMMA is self-assembled into cylinders inside the trench parallel to the substrate, thus showing a dash-line pattern, as indicated in Figure f,i. If the width of the trench is sufficiently wide allowing more than the parallel cylinder to be located, a SCAL-induced self-assembly forms multiple micelle-like structures along the strand of the SCAL, thus forming anisotropic clusters, as indicated in Figure g,j. These new complex nanopatterns are rarely found in conventional BCP self-assemblies or require complicated designs, e.g., incommensurate chemoepitaxy.[57]

The interesting width-dependent pattern transition in the strand-forming SCAL was essential to minimize the surface energy of the film. Because the preferential sitting of PS-b-PMMA films at the P2VP side of a SCAL generates an uneven surface with trenches or holes, filling these areas with micelles can relieve the energy penalty.

Conclusion

In this study, we introduced a SCAL as an effective way to control the adsorbed layers in BCP thin films. Whereas physisorption is generally considered to be weaker than chemisorption, we found that a physisorption-based SCAL possesses a significant adsorption capability and stability, which makes the SCAL a superior alternative for the natural adsorbed layer. Furthermore, the suggested strategy of changing the chain conformation of the adsorbed layer allowed us to effectively change the thermodynamic equilibrium of the film structures.

Upon applying the additional BCP coatings, the underlying layer of the SCAL guides the nanostructures in thin films because it provides topological restrictions and enthalpic/entropic preferences for a BCP self-assembly. The SCAL provides a greater level of opportunity to control the adsorbed layer structures more delicately, which in turn reveals an unprecedented yet interesting BCP nanostructure: arrays of spacing-controlled hole/dot patterns, dotted-line patterns, dash-line patterns, and anisotropic cluster patterns, among others, were found.

We believe that a SCAL can provide a new platform to control the interfacial property of polymer thin films and thus generate more versatile nanostructures required in more advanced polymer-based applications.