Synthesis and morphology investigations of a novel alkyne-functionalized diblock copolymer

Abstract

We reported the synthesis and morphology of a novel alkyne-functionalized diblock copolymer (di-BCP) poly(methyl methacrylate-random-propargyl methacrylate)-block-poly(4-bromostyrene). The di-BCPs were synthesized by atom transfer radical polymerization and postpolymerization deprotection, with good control over molecular weight and polydispersity index. Microphase separation in bulk di-BCPs was confirmed by thermal analysis, small-angle X-ray scattering, and transmission electron microscopy. Microphase-separated morphologies were also observed in thin films, and the orientation of the microdomains can be conveniently controlled by annealing under different solvents. © 2012 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys, 2012

INTRODUCTION

Nanoscaled surface patterns formed by selective immobilization of functional molecules or nanoparticles are crucial for many applications. For instance, they can be used for diagnostics1,2 or the study and controlling of cell adhesion.3–5 Generally, such patterning can be realized by “top-down” approaches, that is, combining surface modification with a series of lithography methods.6–9 Recently, attempts have also been made to use thin films of microphase-separated block copolymers (BCPs) to obtain such functionalized surface patterns, as BCP thin films have been demonstrated as an efficient “bottom-up” lithographic approach to fabricate nanostructured materials.10–14 For example, nanoscale protein arrays have been achieved by selective absorption of proteins into one block of a BCP thin film.15,16 Nanoparticles can also be placed into nanoporous BCP thin films by using electric fields17 or capillary force.18,19

Recently, alkyne/azide click chemistry, that is, the cycloaddition between a terminal alkyne and an azide to form the triazole structure,20–22 has been demonstrated as an efficient synthetic method to obtain complex structures owing to its high yield and high compatibility with different functional groups. Alkyne/azide click chemistry also provides a convenient approach for surface modification,23–27 which allows the fabrication of patterned surfaces. For example, surface modification via alkyne/azide click chemistry can be combined with microcontact printing to generate functional arrays.26,28 Efforts have also been made to prepare microphase-separated BCPs with alkyne or azide functionalities.29–32 Recently, Stadermann et al.33 reported microphase-separated structures in thin films of alkyne-bearing BCPs, which may be used to develop reactive surface patterns.

In this article, we report the synthesis and morphology studies of a novel alkyne-functionalized diblock copolymer (di-BCP) poly(methyl methacrylate-random-propargyl methacrylate)-block-poly(4-bromostyrene) (P(MMA-r-PgMA)-b-PBrS). The di-BCP was synthesized by atom transfer radical polymerization (ATRP), with good control over molecular weight and polydispersity. Microphase separation between the two blocks was found both in bulk and in thin films, and different microphase-separated structures were achieved by adjusting the composition of the di-BCP. By annealing under solvent vapor, the orientation of the microdomains in thin films can be conveniently controlled. Our results suggest that this alkyne-functionalized di-BCP may be of potential interest for developing functionalized nanostructured materials via alkyne/azide click chemistry.

EXPERIMENTAL

Materials

2-Methylacrylic acid 3-trimethylsilanyl-prop-2-ynyl ester34 and benzyl 2-bromoisobutyrate35 were synthesized according to the literature. 4-Bromostyrene (98%), methyl methacrylate (>98.5%), N, N, N′, N″, N″-pentamethyldiethylenetriamine (PMDETA) (99%), copper(I) bromide (99.999%), anisole (anhydrous, 99.7%), and tetrabutylammonium fluoride (TBAF) (1.0 M in tetrahydrofuran [THF], containing about 5 wt % water) were purchased from Sigma-Aldrich. Acetic acid was purchased from Fisher Scientific. 4-Bromostyrene was passed through an inhibitor remover column (Scientific Polymer Products) to remove the inhibitor. Methyl methacrylate was purified by vacuum distillation to remove the inhibitor. THF was dried over sodium metal and distilled. All the other chemicals were used as received. Si (100) wafers with ∼2 nm native oxide layer were purchased from International Wafer Service and cleaned with oxygen plasma for 5–10 min before use.

Instrumentation

Nuclear magnetic resonance (NMR) spectra were recorded on a Bruker DPX300 or DPX400 spectrometer. Gel permeation chromatography (GPC) was measured in THF relative to poly(methyl methacrylate) or polystyrene standards (Scientific Polymer Products) on a system equipped with a three-column set (Polymer Laboratories 300 × 7.5 mm; 5 µm; pore sizes, 10–5, 10–4, and 10–3 Å) and a refractive-index detector (HP 1047 A) at room temperature with a flow rate of 1 mL/min.

Differential scanning calorimetry (DSC) was performed on a TA Instrument Q200 instrument equipped with a refrigerated cooling system and nitrogen purging (50 mL/min). About 5 mg of sample was sealed in an aluminum hermetic pan and cycled at 10 °C /min. Data from the second cycle were used to minimize the effects of thermal history.

Two-dimensional small-angle X-ray scattering (SAXS) was performed at the beamline X27C, National Synchrotron Light Source (NSLS), Brookhaven National Laboratory (BNL). The wavelength of incident X-ray was 0.1371 nm, and the sample-to-detector distance was 1816.6 mm. Scattering signals were collected by a marCCD 2D detector with a resolution of 79 µm/pixel. Typical exposure time was between 30 and 90 s. All the SAXS data presented here are raw data without background subtraction and empty cell scattering calibration. One-dimensional SAXS profiles were obtained by integration of the corresponding two-dimensional scattering patterns.

To observe the microstructures by transmission electron microscopy (TEM), the samples were embedded into epoxy and microtomed (Leica Ultracut) into ultrathin sections with thicknesses of 30 nm using a diamond knife at room temperature. To enhance contrast, the di-BCPs were stained by exposure to the vapor of a 4 wt % aqueous solution of OsO₄ for 1 h. Bright field TEM measurements were performed with a JEOL 200CX electron microscope operated at an accelerating voltage of 200 kV.

Grazing-incidence SAXS (GISAXS) was performed at the beamline 7.3.3, Advanced Light Source (ALS), Lawrence Berkeley National Laboratory (LBL), Berkeley, CA. The wavelength of incident X-ray was 0.124 nm, and the sample-to-detector distance was 4145.6 mm. Scattering signals were collected by a Pilatus 1M 2D detector with a resolution of 172 µm/pixel. Measurements were performed in the open air with typical exposure time between 30 and 90 s.

Scanning force microscopy (SFM) was performed using a Multiple-mode Nanoscope III SFM instrument (Digital Instruments) in the tapping mode with commercial Si cantilevers. Film thickness was measured by ellipsometry.

Synthesis of Poly(methyl methacrylate-random-3-(trimethylsilyl)prop-2-ynyl methacrylate) Random Copolymer

A typical polymerization procedure is as follows: To a 100-mL Schlenk flask, 3.82 g methyl methacrylate, 2.40 g 2-methylacrylic acid 3-trimethylsilanyl-prop-2-ynyl ester, 130 mg benzyl 2-bromoisobutyrate, 40.0 mg Cu(I)Br, and 17.5 mL anisole were added. The flask was then capped with a rubber septum, subjected to three freeze–pump–thaw cycles and backfilled with nitrogen. In another small vial, 85 µL PMDETA was dissolved in 0.5 mL anisole. The solution was bubbled with nitrogen for 30 min and added to the mixture. The mixture was then heated to 70 °C using an oil bath and stirred at 70 °C for 2 h. Polymerization was terminated by cooling the mixture to 0 °C and exposing the mixture to the open air. The solution was then passed through a neutral aluminum oxide column and precipitated into cold methanol. Polymer was isolated by filtration and dried under vacuum.

Synthesis of Poly(methyl methacrylate-random-3- (trimethylsilyl)prop-2-ynyl methacrylate)-b-PBrS Di-BCP

A typical polymerization procedure is as follows: To a 100-mL Schlenk flask, 0.812 g poly(methyl methacrylate-random-3 -(trimethylsilyl)prop-2-ynyl methacrylate) (P(MMA-r-TMSPYMA)) random copolymer (M_n = 7 kg/mol, MMA: TMSPYMA = 3:1, mole ratio) was dissolved in 11.5 mL anisole. To this mixture, 5.65 g 4-bromostyrene and 17.7 mg Cu(I)Br were added. The flask was then capped with a rubber septum, subjected to three freeze–pump–thaw cycles and backfilled with nitrogen. In another small vial, 40 µL PMDETA was dissolved in 0.5 mL anisole. The solution was bubbled with nitrogen for 30 min and added to the mixture. The mixture was then heated to 90 °C using an oil bath and stirred at 90 °C for 27 h. Polymerization was terminated by cooling the mixture to 0 °C and exposing the mixture to the open air. The solution was then passed through a neutral aluminum oxide column and precipitated into methanol. Polymer was isolated by filtration and dried under vacuum.

Synthesis of P(MMA-r-PgMA)-b-PBrS Di-BCP

A typical procedure is as follows: 2.00 g P(MMA-r-TMSPYMA)-b-PBrS di-BCP (M_n (P(MMA-r-TMSPYMA)) = 7 kg/mol, M_n (PBrS) = 27 kg/mol, DP (MMA) = 42, DP(TMSPYMA) = 14, 0.82 mmol TMSPYMA repeating units, and 2.5 mmol MMA repeating unites) was dissolved in 130 mL dry THF, and 0.4 mL acetic acid (6.7 mmol) was added. The solution was bubbled with nitrogen for 30 min and cooled to 0 °C. In brief, 2.2 mL TBAF solution (2.2 mmol) was then added dropwise at 0 °C. After that, the cooling bath was removed, and the solution was stirred at room temperature overnight. The solution was then concentrated and precipitated into methanol. Polymer was isolated by filtration and dried under vacuum.

Sample Preparation

Bulk di-BCP samples for SAXS and TEM experiments were prepared by drop-casting a 5 wt % solution of the di-BCP in dichloromethane onto Kapton films. Thin film samples were prepared by spin coating a 1.5 wt % solution of the di-BCP in the mixture of toluene and THF (toluene/THF = 7/3 v/v) onto silicon wafers.

RESULTS AND DISCUSSIONS

Synthesis of Di-BCPs

Scheme 1 shows the synthetic pathway of P(MMA-r-PgMA)-b-PBrS di-BCPs. A P(MMA-r-TMSPYMA) random copolymer was first synthesized via ATRP. Molecular weight of the random copolymer can be estimated from ¹H NMR spectra using the protons in the phenyl ring as an internal standard, as has been reported previously.34 The P(MMA-r-TMSPYMA) random copolymer was then used as a macroinitiator for the ATRP of 4-bromostyrene. After the polymerization of 4-bromostyrene, the trimethylsilyl (TMS) protecting groups were quantitatively removed using TBAF. The main reason of using 4-bromostyrene instead of styrene is that although thin films of polystyrene-block-poly(methyl methacrylate) (PS-b-PMMA) di-BCPs have been widely used as templates for the fabrication of nanostructured materials,36,37 PS-b-PMMA is only weakly segregated.38 On the other hand, PBrS-b-PMMA has much stronger nonfavorable segmental interactions, and hence it is easier to generate well-defined microphase-separated structures without the need of reaching very high molecular weight.39 In addition, previously Chen et al.39 reported that, in thin films of PBrS-b-PMMA di-BCPs, the microdomains can be aligned perpendicular to the substrate by solvent annealing, and hence it is possible to use similar strategies to control the orientation of microdomains in thin films of P(MMA-r-PgMA)-b-PBrS di-BCPs.

Scheme 1. Synthesis of the P(MMA-r-PgMA)-b-PBrS di-BCPs.

Two P(MMA-r-TMSPYMA) random copolymers, denoted as Random 1 and Random 2, were synthesized. As summarized in Table 1, the molecular weight calculated from ¹H NMR spectra is 17 kg/mol for Random 1 and 7 kg/mol for Random 2, and for both copolymers, the mole ratio between the TMSPYMA and the MMA repeating units is about 1:3. Each of the random copolymer was then used as the macroinitiator to synthesize one di-BCP. For simplicity, the P(MMA-r-TMSPYMA)-b-PBrS di-BCP synthesized using Random 1 is denoted as BCP1-TMS, and the corresponding P(MMA-r-PgMA)-b-PBrS di-BCP, that is, the final di-BCP after deprotection, is denoted as BCP1. Similarly, BCP2-TMS and BCP2 refer to the di-BCPs synthesized using Random 2 as the macroinitiator. As it is clear from the GPC curves (Fig. 1), after polymerization, there was significant increase in molecular weight, whereas the narrow distribution of molecular weight was still maintained, indicating good control over molecular weight and molecular weight distribution.

Figure 1. GPC profiles of Random 1(i), BCP 1(ii), Random 2 (iii), and BCP 2(iv). Profile curves are offset for clarity.

Table 1. Characteristics of the P(MMA-r-TMSPYMA) Random Copolymers, P(MMA-r-TMSPYMA)-b-PBrS di-BCPs, and P(MMA-r-PgMA)-b-PBrS di-BCPs

Figure 2 shows the typical ¹H NMR spectra of the P(MMA-r-PgMA)-b-PBrS di-BCP. The chemical shift and integration area of all peaks are in good agreement with the proposed structure. Especially, no peak is observed around 0 ppm, which indicates that all the TMS groups have been removed using TBAF. From the ¹H NMR spectra, we calculated the total molecular weight as well as composition of these di-BCPs, and the results are summarized in Table 1. According to the literature, the density of PBrS homopolymer is 1.57 g/cm³,40 and we assumed that the density of both P(MMA-r-TMSPYMA) and P(MMA-r-PgMA) blocks is the same as the density of PMMA (1.18 g/cm³). Therefore, the volume fraction of each block can be also calculated, which is also listed in Table 1.

Figure 2. ¹H NMR spectra of the P(MMA-r-PgMA)-b-PBrS di-BCP.

As summarized in Table 1, the values of molecular weight of these di-BCPs determined from ¹H NMR spectra are close to those measured by GPC. BCP1 and BCP1-TMS have f(PBrS) around 50%, whereas for BCP2 and BCP2-TMS, f(PBrS) is much higher, around 75%. From Table 1 we could also see that, for both BCP1 and BCP2, the mole ratio between the two comonomers is close to the ratio in the macroinitiator and does not change significantly during the deprotection step, indicating no loss of functional groups during both polymerization and postpolymerization deprotection.

Thermal Properties and Bulk Morphologies

Figure 3 shows the DSC traces of the two P(MMA-r-PgMA)-b-PBrS di-BCPs during the second heating cycle. As it can be seen, BCP1 shows two distinctive glass transitions. The first glass transition is at ∼140 °C, which should be ascribed to the glass transition of the PBrS block.41 The other glass transition at ∼110 °C should correspond to the glass transition of the P(MMA-r-PgMA) block. The existence of two glass transitions clearly indicates the microphase separation between the two blocks. For BCP2, glass transition of the P(MMA-r-PgMA) block is not very obvious, probably owing to the much lower volume fraction of the P(MMA-r-PgMA) block and therefore much smaller change of heat capacity during the glass transition. However, as a glass transition is clearly seen at ∼140 °C, it is also an indication of microphase separation in BCP2.

Figure 3. DSC traces of BCP1 (i) and BCP2 (ii) during heating. Exothermic is up. The arrows indicate the glass transitions. Profiles are offset for clarity.

SAXS was then performed to further confirm the microphase separation in the two di-BCPs. As the terminal alkyne group may crosslink at higher temperatures, thermal annealing of the di-BCPs may not generate well-defined microphase-separated morphologies. To overcome this problem, the as-cast samples were annealed at room temperature under the vapor of THF, a good solvent for both blocks. The samples were annealed under THF vapor for 8 h and then dried under vacuum overnight to remove the residual solvent. Figure 4 shows the SAXS profiles of the di-BCP samples after solvent annealing. As it can be seen, both samples show sharp scattering peaks, indicating microphase separation. BCP1 shows the first-order peak at q = 0.19 nm⁻¹ and higher order reflections at q = 2q* and 3q*, indicating the formation of a lamellar microdomain morphology with the domain spacing of ∼33 nm. BCP2 shows the first-order peak at q = 0.22 nm⁻¹ and a higher order reflection at q = 3^1/2q*, characteristic of a hexagonally packed cylindrical morphology with the domain spacing of ∼29 nm. Bulk morphology of BCP2 was also investigated by cross-section TEM, and the image is shown in Figure 5. The alkyne groups in the P(MMA-r-PgMA) block were selectively stained by OsO₄, making this block appears dark. The coexistence of circular and line structures further confirms the formation of hexagonally packed cylindrical microdomains.

Figure 4. SAXS profiles of BCP1 (i) and BCP2 (ii). The samples were annealed under the vapor of THF for 8 h. Profile curves are offset for clarity.

Figure 5. Cross-section TEM image of BCP2. The sample was annealed under the vapor of THF for 8 h.

We also investigated the morphologies of the di-BCPs before the removal of the TMS groups. Previously, it has been demonstrated that the change in chemical structure owing to deprotection can have significant effects on the morphology of BCPs. For example, during thermal annealing, BCPs containing poly(tert-butyl acrylate) block or poly(tert-butyl methacrylate) block can be converted to the corresponding BCPs containing poly(acrylic acid), poly(acrylic anhydride), or poly(methacrylic acid) block, which leads to a series of change in morphology.42–50 However, from the SAXS profiles shown in Figure 6 we can see that the two P(MMA-r-TMSPYMA)-b-PBrS di-BCPs show almost identical morphologies to the morphologies of the corresponding di-BCPs after deprotection, although BCP2-TMS has slightly lower domain spacing than BCP2. These results are completely different from our previous observations in another alkyne-functionalized di-BCP, where the removal of the TMS groups results in a dramatic decrease in the nonfavorable segmental interactions.51,52 It is likely that for P(MMA-r-PgMA)-b-PBrS di-BCPs, the change in chemical structure owing to deprotection does not change the segmental interactions substantially. In addition, the change in the volume fraction of both blocks owing to deprotection of the TMS groups is also small. As a result, the microphase-separated morphologies before and after deprotection are almost the same.

Figure 6. SAXS profiles of BCP1-TMS (i) and BCP2-TMS (ii). The samples were annealed at 180 °C under vacuum for 12 h. Profile curves are offset for clarity.

Thin Film Morphologies

After confirming the microphase separation in bulk samples, we continued to study the microphase-separated structures of these di-BCPs in thin films. Recently, solvent annealing has been demonstrated as an efficient approach to control the orientation of BCP microdomains in thin films.53,54 As mentioned earlier, it has been reported that in thin films of PBrS-b-PMMA di-BCPs, the microdomains can be aligned perpendicular to the substrate by solvent annealing.39 As our P(MMA-r-PgMA)-b-PBrS di-BCP can be considered as a modification in chemical structure of the parent PBrS-b-PMMA di-BCP, we also expect that the orientation of the microdomains can be realized under similar solvent annealing conditions.

Figure 7 shows the SFM height image and the corresponding GISAXS pattern of a thin film of BCP1. The thin film was prepared by spin coating a 1.5 wt % solution in a mixture of toluene and THF (toluene/THF = 7/3 v/v) onto a silicon wafer. The as-spun sample was then annealed under a mixed vapor of hexanes and THF (hexanes/THF = 3/2 v/v) for 3 h. As it can be seen, after solvent annealing, worm-like textures were observed on the top surface [Fig. 7 (a)], and Bragg rod scattering peaks along q_z were seen in the corresponding GISAXS pattern [Fig. 7 (b)], indicating a lamellar microdomain morphology with a perpendicular orientation. Integration of the GISAXS pattern along q_y shows a sharp scattering peak at q_y = 0.17 nm⁻¹, which is close to the position of the peak position from the power spectral density (PSD) profile calculated from the fast Fourier transform of the SFM image (Fig. 8), indicating that the SFM height image is in good agreement with the GISAXS pattern. The GISAXS line profile also shows the second-order peak at q_y = 0.34 nm⁻¹, further confirming the lamellar microdomain morphology in the thin-film sample.

Figure 7. SFM height image (a) and GISAXS pattern (b) of a thin film of BCP1. The thin film was annealed for 3 h under the vapor of hexanes and THF (hexanes/THF = 3/2 v/v). Film thickness is 35 nm. Scale bar is 200 nm. GISAXS incident angle is 0.18°.

Figure 8. PSD profiles of the SFM height images (□) and q_y line profiles at q_z = 0.5 nm⁻¹ of the corresponding GISAXS patterns (○) of the thin film of BCP1. The thin films were annealed for 3 h under the vapor of hexanes and THF (hexanes/THF = 3/2 v/v).

Thin films of BCP2 were also prepared by spin coating a 1.5 wt % solution in a mixture of toluene and THF (toluene/THF = 7/3 v/v) onto silicon wafers. As shown in Figure 9, when the sample was annealed under a mixed vapor of hexanes and THF (hexanes/THF = 3/2 v/v) for 3 h, circular structures on the top surface were observed from the SFM height image [Fig. 9(a)], and the corresponding GISAXS pattern [Fig. 9(c)] shows sharp Bragg rod scattering peaks extending along q_z. These results indicate a hexagonally packed cylindrical microdomain morphology with a well-defined in-plane separation distance, where the cylindrical microdomains are truncated by the surface of the film. It is thus clear that both the lamellar microdomains and the cylindrical microdomains can be aligned perpendicular to the substrate using similar conditions. On the other hand, when only hexanes were used for solvent annealing, worm-like textures were seen from the SFM height image [Fig. 9(b)], and the corresponding GISAXS pattern [Fig. 9(d)] still shows scattering peaks along q_z, although the scattering intensity is weak, which suggests that the cylindrical microdomains are aligned parallel to the substrate. The PSD profiles of the SFM height images are also consistent with the GISAXS line profiles (Fig. 10). It is thus clear that the orientation of microdomains can be conveniently controlled by annealing under different solvent vapors. Such differences may be owing to the different selectivity and evaporation rate of the solvent, as has been demonstrated in thin films of a variety of BCPs.55–59 Further modification of the alkyne-containing P(MMA-r-PgMA) microdomains with azides may generate well-ordered nanoscaled functionalized patterns, which may be used in applications such as biosensors and cell adhesion controlling.

Figure 10. PSD profiles of the SFM height images (□) and q_y line profiles at q_z = 0.5 nm⁻¹ of the corresponding GISAXS patterns (○) of thin films of BCP2. The thin films were annealed for 3 h under the vapor of hexanes and THF (hexanes/THF = 3/2 v/v) (a) and hexanes (b), respectively.

Figure 9. SFM height images (a, b) and GISAXS patterns (c, d) images of thin films of BCP2. The thin films were annealed for 3 h under the vapor of hexanes and THF (hexanes/THF = 3/2 v/v) (a, c) and hexanes (b, d), respectively. Film thickness is 45 nm (a, c) and 36 nm (b, d). Scale bar is 200 nm. GISAXS incident angle is 0.18°.

CONCLUSIONS

In summary, we have successfully synthesized di-BCPs containing pendant terminal alkyne functional groups in one block using ATRP. Microphase separation in bulk di-BCPs was observed by using DSC, SAXS, and TEM, and different microphase separation morphologies were realized by adjusting the volume fraction of both blocks. Well-defined microphase-separated structures were also achieved in thin films. The orientation of microdomains was controlled by annealing under the vapor of different solvents. These results demonstrate a promising route for the fabrication of functionalized nanoscale surface patterns by combining BCP thin films and alkyne/azide click chemistry.

Acknowledgements

This study was supported by the Department of Energy Office of Basic Energy Science under contract no. DE-FG02-96ER45612 (X. Wei, X. Shen, and T. P. Russell) and DE-FG02-04ER46126 (J.-K. Kim and T. P. Russell), the NSF-supported Materials Research Science and Engineering Center at UMass (L. Li, X. Shen) and the NSF-supported Center for Hierarchical Manufacturing at UMass (W. Gu). Use of the National Synchrotron Light Source, Brookhaven National Laboratory, was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DEAC02-98CH10886. Use of the Advanced Light Source is supported by the Director, Office of Science, Office of Basic Energy Sciences, of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231.