PS-b-PEO block copolymer thin films as structured reservoirs for nanoscale precipitation reactions

Authors


  • Special Issue in honor of Prof. Manfred Stamm of The Leibniz Institute of Polymer Research Dresden, Germany on the occasion of his 60th birthday.

Abstract

Thin films of PS-b-PEO block copolymers were utilized as structured reservoirs for localized nanoscale precipitation reactions. By consecutively immersing the film into solutions of thioacetamide and cadmium chloride, we were able to obtain a monolayer of cadmium sulfide nanostructures on top of the block copolymer film. AFM and grazing incidence small angle X-ray scattering revealed spherical nanostructures (d = 15 nm) corresponding to the dimensions given by the block copolymer film. © 2010 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 48: 1569–1573, 2010

INTRODUCTION

In recent years, thin films of block copolymers have been extensively used as templates for a variety of nanostructures such as nanodots and nanowires.1–7 The reason for this is being the easy preparation of nanometer-sized structure films through simple self-assembly and the convenient domain size control.8–12

To obtain the desired structure a thin block copolymer film is usually prepared first. This is normally done by spin coating of a diluted solution of the block copolymer, followed by an annealing step. A void is then commonly created by the removal of one block. This is done by selectively etching13 or by dissolution of one of the polymer blocks.14, 15 These voids in nanoporous film are then used as a reaction vessel or as a mask to obtain the desired nanostructure.

In a different approach, reconstructed diblock copolymer films were used to create inorganic nanostructures. First, the minor component was drawn to the surface by exposing the film to a solvent of the minor component. Evaporation of, for example, gold on the surface, followed by thermal annealing of the film led to the generation of different nanostructures.5

Despite its common use, sacrificial removal of one block or reconstruction of the diblock copolymer film is not a necessity. Russell and coworkers obtained an array of inorganic nanodots of TiO2/SiO2 with the help of a PS-b-PEO film. The polyethylene oxide (PEO) domains were swollen with water and the film was exposed to TiCl4/SiCl4 vapor to form the desired nanostructure through a simple hydrolyzation reaction at the film's surface.16, 17 This demonstrated the first direct use of one polymer block as a nanometer size, structured reservoir for the formation of nanostructures on surfaces. Although innovative, the reported method has great limitations because of the need of gaseous reactants and its limitation to hydrolysis reactions.

In this article, we demonstrate the use of a PS-b-PEO film as a structured reservoir for utilization in an inorganic precipitation reaction. This method has two distinctive features when compared with the usual template approach. Initially, it constricts the formation of the inorganic nanostructure mainly to the film's surface. Secondly, it shows a pathway to numerous inorganic nanostructures by a versatile precipitation reaction. In contrast to a normal template approach, only one component is first accumulated inside the PEO-block. By subsequently exposing the reservoir to the counterpart, combined with an initialization through a simple increase in temperature, a reaction is induced and the desired inorganic nanostructure is formed. In this article, we apply this approach to PS-b-PEO films with cylindrical PEO domains, which are mostly oriented normal to the surface.

EXPERIMENTAL

In this study, an asymmetric diblock copolymer of polystyrene (PS) and polyethylene oxide (PEO), PS-b-PEO, with a molecular weight of 25,400 g/mol (Mmath image = 19,000 g/mol; Mmath image = 6400) or 26,500 g/mol (Mmath image = 20,000 g/mol; Mmath image = 6500) was used. The polymer was purchased from Polymer Source Inc. and had a PDI of 1.05 and 1.06. To prepare the polymer films, a 2 wt % benzene solution was spin coated onto a glass substrate (26 × 26 mm2) at 500 rpm and was vapor annealed afterward in benzene vapor for 1 h.

The dried film was subsequently put into a 1 mol solution of thioacetamide in distilled water acidified by the addition of sulfuric acid (pH = 1). After 1 h, the film was rinsed with distilled water and dried in air. The obtained film was then put into a 1 mol solution of CdCl2 and placed into a preheated oven. Afterward, the film was again rinsed with distilled water and dried in air.

The thioacetamide, purchased from Aldrich 99+% A.C.S, the cadmium chloride, purchased from Merck-Schuchardt, and all solvents were used without further purification.

The film thickness was determined by X-ray reflectivity measurements on an X-ray diffraction system XRD 3003 TT, Seifert with a wavelength of λ = 0.154 nm.

The AFM images were obtained in tapping mode using a Dimension 3000 scanning probe microscope, Veeco Instruments GmbH or a Multimode scanning probe microscope, Veeco Instruments GmbH. OMCLAC 160 TS-W2, Olympus, Japan cantilevers were used with a resonance frequency of ∼300 kHz and a spring constant of 42 N/m as specified by the manufacturer.

Grazing incidence small angle X-ray scattering (GISAXS) measurements were conducted at the beamline BW4/HASYLAB, DESY, Hamburg, Germany with a sample to detector distance of 2 m, a wavelength λ = 0.138 nm, a beam size of 35 × 17 μm2, and an incident angle αi = 0.55°.

RESULTS AND DISCUSSION

The PS-b-PEO films, obtained by spin coating, were vapor annealed in benzene vapor to obtain cylindrical PEO domains as was reported by Russell and coworkers.11 Indeed, we were able to produce hexagonal arrays of cylindrical domains that have an average diameter of ∼15 nm with a center-to-center distance of ∼26 nm. However, it has to be noted that our approach works equally well for other thin film morphologies such as lamellae normal to the surface. Some examples of structures resulting from such a lamellar arrangement are also investigated in this study.

Our synthetic procedure is illustrated schematically in Figure 1. The obtained PEO domains were loaded with thioacetamide by immersing the copolymer film in an acidic thioacetamide solution and dried in air. Subsequently, the film was put in a solution of cadmium chloride and heated. Thermal treatment induced a hydrolyzation reaction (Scheme 1) of the thioacetamide and S2−-ions were obtained inside the PEO domains. By this means, we created a localized, nanometer size source of S2−-ions inside the polymer film.18–20 AFM studies showed that the integrity and order of the film are not altered by the immersion in aqueous solutions or the elevated temperatures. The stability provided by the insoluble PS matrix is critical for the success of the experiment, as a homopolymer film of PEO would simply dissolve under the experimental conditions.

Figure 1.

Schematic diagram of the preparation steps required to obtain nanostructures on top of the PS-b-PEO thin film, by means of a precipitation reaction. The PS-b-PEO film is first immersed into a thioacetamide solution. After rinsing, the film is immersed into a cadmium chloride solution and heated to 80 °C. The hydrogen sulfide formed rises to the surface of the polymer film and reacts with the Cd2+-ions to form cadmium sulfide. Polyethylene oxide domains normal to the surface are illustrated as one possible structure applicable to this universal method. Such domains can be the result of aligned cylindrical or lamellar microstructures.

Scheme 1.

Hydrolyzation of thioacetamide.

The created hydrogen sulfide then rises to the film/liquid interface and reacts with the Cd2+-ions in solution to create cadmium sulfide (Scheme 2). Because of the low solubility product of cadmium sulfide in aqueous solution, the precipitation takes place at the interface between polymer film and cadmium chloride solution. By this means, we were able to control cadmium sulfide formation and restrict it to the PEO domains. Therefore, the size and the location of the formed cadmium sulfide are predetermined by the structure of the PS-b-PEO film. Using the thermo sensitive thioacetamide as a source for sulfide ions allows us to control the release of sulfur ions and, therefore, the formation of cadmium sulfide by simple temperature variation.

Scheme 2.

Formation of cadmium sulfide.

Figure 2 shows the AFM height image of a ∼290-nm thick PS-b-PEO film at the end of the procedure and the corresponding 3D view. The formation of a dense array of cadmium sulfide on top of the surface is clearly visible. Most cadmium sulfide nanostructures show a dot-like structure corresponding to cylindrical PEO domains normal to the surface. The AFM measurements also revealed strip-like structures corresponding to PEO domains horizontal to the surface. The extracted line profiles show an average height of the dots of ∼2 nm. To further analyze the cadmium sulfide structure, GISAXS was conducted. The q−scan of the scattering pattern is shown in Figure 3. The q−scan was fitted with the Unified Fit Model.21 This model is used for the lateral structural analysis for films containing weakly correlated, polydisperse particles. The method models scattering using different structural levels, which consist of a Guinier and a Porod regime, respectively. This allows the modeling of scattering from primary particles (d ≈ 10 nm) and aggregates (d ≈ 350 nm). The approximation included in this model allows the size and the structural determination with an accuracy of <20%.

Figure 2.

(Top) AFM height image of cadmium sulfide structure on a PS-b-PEO film. (Bottom) 3D view of the AFM height image of cadmium sulfide on a PS-b-PEO film.

Figure 3.

Out-of-plane cut of the GISAXS pattern (empty square) with calculated fit (filled dots). The unified fit resulted in two structural levels confirming the formation of a monolayer (P = 1.8) of sphere-like particles (P = 3.2) with a radius of gyration, Rg, of 7.5 nm. The dots had an average distance (ξ) of 45 nm.

By determining the radius of gyration, Rg, information about mean particle/aggregate dimensions can be obtained. Furthermore, distinctive power law decays yield information on the structure's fractality. This allows us to draw conclusions on the particle and/or aggregate shapes.

From the fit, the formation of particles with an Rg = 7.5 ± 1.5 nm was concluded, in agreement with the diameter of an average PEO domain of the bare polymer film (∼15 nm). Two broad overlapping peaks corresponding to the single and double distance of the cadmium sulfide dots result in an average center-to-center distance of 45 ± 9 nm. The broadness of the peaks is mainly caused by the large-size distribution of the cadmium sulfide and the inhomogeneity in the shape of the particles. The power law exponent P from Porod scattering gives information about the structure of the single scattering object or of the larger aggregate depending on the observed structural level. Spherical particles would yield a power law exponent of 3–4. Thus, the obtained P of 3.2 ± 0.4 points to the formation of mostly dot-structured cadmium sulfide. The second structural level, describing larger structures like aggregates, yields a power law exponent P = 1.8 ± 0.2. Because an exponent smaller than 2 corresponds to a two-dimensional fractal, this supports the assumption of the formation of a monolayer of cadmium sulfide structures on top of the PS-b-PEO film.

In summary, the GISAXS data clearly show that the domain size and spacing of the used copolymer film, and the formed cadmium sulfide structures are in the same order of magnitude. Therefore, we conclude that the formation of cadmium sulfide occurs predominantly on the PEO domains.

To investigate a possible partial incorporation of cadmium sulfide inside the PEO-domain, X-ray reflectivity measurements were conducted. X-ray reflectivity measurements showed three edges of total reflection, relating to the glass substrate, the PS-b-PEO film, and a cadmium sulfide layer. The reflectivity measurement also showed a more pronounced oscillation when compared with the bare polymer film, caused by an increase in electron density inside the polymer film. This is a direct result of the partial incorporation of cadmium sulfide inside the PEO domains.

In summary, the investigation fully supports the proposed formation of a monolayer of mostly “sphere-like” cadmium sulfide dots, which are partially incorporated inside the PEO domain. To further understand the reaction mechanism, investigation of the experimental condition is conducted.

Lowering the temperature to 60° or further led to an absence of cadmium sulfide on the polymer surface. Instead, larger aggregate-like structures were formed. The temperature decrease leads to a deceleration of the speed of hydrolyzation of the thioacetamide.19, 20 As a result, not enough hydrogen sulfide is formed inside the PEO domain. Instead, hydrolysis of the thioacetamide occurs inside the cadmium chloride solution and results in the formation of larger aggregates of cadmium sulfide.

An increase of the pH using acetic acid has the same effect as lowering the temperature, because the pH is also critical for the speed of the thioacetamide hydrolyzation.19, 20

Given that the cadmium sulfide film thickness should be directly proportional to the amount of available thioacetamide film, the effect of the reservoir thickness was further investigated.

Changing the thickness of the PS-b-PEO film from 290 nm to 110 nm resulted in the absence of a large amount of cadmium sulfide dots. This is due to an almost complete diffusion of the thioacetamide out of the PEO domains before its hydrolyzation. In conclusion, this indicates that in the case of the thicker films (∼290 nm), large amounts of the absorbed thioacetamide are not hydrolyzed inside the PEO domain, and thus do not contribute to the formation of cadmium sulfide.

To evaluate the time dependence of the cadmium sulfide dots formation, samples were prepared as usual and heated in cadmium chloride solution for 10, 20, and 30 min, respectively. As seen in the AFM images in Figure 4, after 10 min, the film showed almost no formation of cadmium sulfide on the film's surface, whereas after 20 min, the film showed a large number of cadmium sulfide dots with some voids between them. After 30 min, the PS-b-PEO film sample was completely covered with cadmium sulfide. The time dependence of the cadmium sulfide formation is mainly caused by the thioacetamide hydrolysis. Because the hydrolysis of the thioacetamide is dependent upon temperature, the hydrolysis is slow until the sample has reached the ambient temperature of 80 °C.

Figure 4.

AFM phase images of PS-b-PEO film immersed into thioacetamide solution, immersed into cadmium chloride solution, and heated to 80 °C for 10, 20, and 30 min. After 10 min, the film shows almost no formation of cadmium sulfide, whereas after 20 min, the film shows a large number of cadmium sulfide on the surface. After 30 min, the film is fully covered with cadmium sulfide.

CONCLUSIONS

In conclusion, we have demonstrated the possibility of using a PS-b-PEO block copolymer thin film as a nanostructured reservoir for nanoscale precipitation reactions. Cadmium sulfide could be obtained at the surface of the polymer film with the same sizes and spacing as the block copolymer film we used. Investigation of the reaction conditions indicates that only a hydrolyzation of the thioacetamide inside the PEO domains leads to a formation of cadmium sulfide on top of the polymer film and that a substantial amount of thioacetamide is lost before that, due to diffusion.

This universal method could be applied to a large variety of inorganic precipitation reactions. In combination with the flexibility of the used PS-b-PEO film, this opens the door to a wide range of inorganic nanostructures of different size, spacing, confirmation, and also chemical composition. Its flexibility could make this method a versatile path for use in optoelectronic devices like hybrid solar cells.

Acknowledgements

The authors thank both Uwe Rietzler and Rüdiger Berger for their support during the AFM measurement and Susan Pinnells for proofreading. Beamtime provision by the HASYLAB at DESY is gratefully acknowledged. This research was supported by the Max Planck Society (Institutsübergreifende Forschungsinitiative FRM II), the Korean German IRTG Program (DFG Graduiertenkolleg 1404), and the DFG priority program 1369 (GU771/3).