Solvent vapor induced morphology transition in thin film of cylinder forming diblock copolymer

Applied Surface Science 257 (2011) 8093–8101

Contents lists available at ScienceDirect

Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc

Solvent vapor induced morphology transition in thin film of cylinder forming diblock copolymer Yuhu Li a , Haiying Huang a , Tianbai He a,∗ , Yumei Gong b,∗∗ a b

State Key Laboratory of Polymer Physics and Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, PR China School of Chemical and Material, Dalian Polytechnic University, Dalian 116034, PR China

a r t i c l e

i n f o

Article history: Received 19 February 2011 Received in revised form 24 April 2011 Accepted 24 April 2011 Available online 30 April 2011 Keywords: Controlled solvent vapor pressure annealing Diblock copolymers Self-assembly

a b s t r a c t The morphology formation and transition of thin film of a cylinder-forming polystyrene–block– poly(methyl methacrylate) (PS–b–PMMA) diblock copolymer annealed under 1,1,2-trichloroethane (TriCE), toluene (Tol), and their binary mixed solvent vapors is investigated by using optical microscopy (OM) and transmission electronic microscopy (TEM). By modulating the annealing solvent vapor pressure and the preferential affinities, a detailed morphology evolution with increasing the vapor pressure and a series of morphologies depending on the preferential affinities have been observed. A phase diagram by plotting the morphologies as a function of the annealing solvent vapor pressure and its preferential affinity is subsequently constructed. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Block copolymers, which consisting of two different polymer molecules linked end-to-end by a covalent bond, have become the subject of considerable attention for their ability to self-assemble into complex periodic structures with a characteristic length scale of tens of nanometers [1–7]. The key in the use of block copolymers is the control over the orientation and lateral ordering of the morphologies in thin films [8]. Generally, the order in thin films of block copolymers is usually induced by the way of thermal annealing and solvent annealing. For thermal annealing, there usually only exists a small window between the Tg and the degradation temperature of the involved blocks which are usually susceptible to thermal degradation [9]. For these systems, solvent annealing provides an attractive alternative to thermal annealing. Since sufficient mobility of the block chains is easily induced at room temperature without the danger of degradation in solvent annealing [10–14], the approach of solvent annealing has been used more and more widely in recent years [15–20]. However, the details of solvent annealing remain rather unclear since the relevant experimental parameters governing the resultant block copolymer morphologies (nature of the solvent, the relative solvent vapor pressure, the annealing time, etc.) are usually complex and difficult to control. Among these factors, the nature of

∗ Corresponding author. Tel.: +86 431 85262123; fax: +86 431 85262126. ∗∗ Corresponding author. Tel.: +86 411 86323736; fax: +86 411 86323736. E-mail addresses: [email protected] (T. He), [email protected] (Y. Gong). 0169-4332/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2011.04.113

the solvent [11,17,21–23] and the relative solvent vapor pressure [9,10,13,17,24–26] are crucial. The nature of the solvent influences the degree of the swelling of each block and accordingly has severe effects on the resulting block copolymer morphologies [17,22,23]. It has been shown that block copolymer thin films annealed under different selective solvents and nonselective solvents result in different morphologies and orientations [11,17,22,23]. On the other hand, once a solvent is selected, the solvent vapor pressure is the key factor to control the resultant morphologies. A lot of methods have been used in reported works to control the solvent vapor pressures, such as the use of the flow of N2 [10,13,24,25], vary the amount of solvent in reservoir [17], close the lid of dish more or less tightly [9], and change the ratio between the surface area of the solvent and the empty volume of the annealing chamber [26]. More recently, Lu et al. [27,28] reported a novel controlled solvent vapor pressure annealing (C-SVA) method which can control the solvent vapor pressure easily and precisely, and have created homogeneous and reproducible fullerene nanorods in thin films successfully. In the present work, the morphology transition of a cylinderforming PS–b–PMMA diblock copolymer thin films annealed under Tri-CE, Tol, and their binary mixed solvent vapors with different vapor pressures by the way of C-SVA is systematically investigated. In our previous work [29], the morphology transition of a similar PS–b–PMMA diblock copolymer at different block copolymer concentration cast from 1,1,2,2-tetrachloroethane has been studied. With the solution becoming concentrated, disordered spheres, inverted spheres, inverted cylinders, inverted hexagonally perforated lamella, lamella and finally normal cylinder in turn have been obtained by freeze-drying method. However, because of the

8094

Y. Li et al. / Applied Surface Science 257 (2011) 8093–8101

limitations of the freeze-drying method, the copolymer concentration increases rapidly as the amount of solvent decreases, especially at relative higher concentrations, it is difficult to trap the morphology of the diblock copolymer in detail. Here, the C-SVA is an effective method in which the solvent vapor pressure can be controlled precisely and stably (from saturated pressure (P0 ) to close to zero) [27,28], the morphology transition at higher block copolymer concentration can right be easily implemented. Meanwhile, both Tri-CE and toluene has similar volatility and opposite preferential affinity for the PS and PMMA blocks, which makes it is easy to modulate the preferential affinity of the binary mixed solvent vapors simply by changing the mixed volume fraction of two solvents, TriCE and toluene. By precisely controlling solvent vapor pressure and solvent vapor preferential affinity, the evolution of morphologies of solvent vapor annealed PS–b–PMMA thin films has been understood in detail, which will offer a good reference in modulating the morphology of block copolymer thin films. 2. Experimental 2.1. Materials An asymmetric diblock copolymer used in this work PS1411 –b– PMMA540 (the numbers in subscript refer to the number-average degree of polymerization of each block) with polydispersity index 1.10 was purchased from Polymer Source Inc. and used as received. In bulk state, the block copolymer adopts hexagonally packed PMMA cylinders with a center-to-center distance (the period L0 ) about 65 nm embedded in a PS matrix [30]. 2.2. Sample preparation Block copolymer films were prepared by dissolving the sample in 1,1,2-trichloroethane (Tri-CE) to yield 10 mg/ml solution and spin-coating the solution onto carbon-coated mica at 2000 rpm for 30 s. The thickness of the films is approximately 70 nm as measured by transmission electronic microscopy (TEM) cross-sectional observation. For the experiment of controlled solvent vapor pressure annealing (C-SVA), the fresh spin-coated films were suspended at different height in a long glass tube (6 cm in diameter and 160 cm in height) containing Tri-CE, Tol, and their binary mixtures at the bottom respectively, as shown in Scheme 1. The characteristics of the annealing solvents are listed in Table 1, and the calculated polymer–solvent interaction parameters () for different pairs of polymers and solvents at room temperature are shown in Table 2. Upon achieving an equilibrium state through the diffusion of solvent vapor within the tube, a gradient distribution of solvent vapor pressure along the tube is constructed, from saturated pressure (P0 ) at the bottom to close to zero at the top. Note that the performance should be done carefully so as to avoid interrupting too much to the vapor diffusion equilibrium in the tube. After being exposed to the solvent vapor for 24 h, the films were removed from the glass tube as quickly as possible and air-dried at room temperature to make the residual solvent in the film volatize instantaneously and accordingly to preserve the morphology formed during the

Scheme 1. Schematic diagram of experimental setup used for C-SVA. L0 is the effective length given by distance from the up edge of the setup to the surface of the annealing solvent at the bottom of the tube. The actual value of L0 in this work is 160 cm so as to achieve a stable and highly resolved gradient of solvent vapor pressures P. L is the distance from the up edge of the setup to the specimen position. Solvent vapor pressure is given by P = (L/L0 )P0 . This scheme is from Ref. [27].

Table 2 Polymer–solvent interaction parameters () calculated from different pairs of polymers and solvents.

Tri-CE Tol

PS

PMMA

0.448 0.344

0.385 0.374

Calculated from P–S = VS (ıS − ıP )2 /RT + 0.34 [31,32], where P represents polymer, S represents solvent, VS is the molar volume of solvent, R is the gas constant, T is the temperature, and ıS and ıP are the solubility parameters of the solvent and polymer respectively.

annealing process in the film. The period 24 h of the film exposure to solvent vapor is sufficient to make the film form a morphology matching a certain circumstance of the solvent vapor. And the morphology did not change with further prolonging the annealing time within the time scale of experiment ∼48 h. 2.3. Instruments The morphologies of the films were examined by optical microscopy (OM) and transmission electronic microscopy (TEM). The surface topographies of films were investigated by optical microscopy (OM). OM experiments were carried out by using a Carl Zeiss A1m microscope equipped with a CCD camera. A

Table 1 Characteristics of the solvents.

Tri-CE Tol a b

Solubility parametera ı (J/cm3 )1/2

Molar volumeb V (cm3 /mol)

Boiling temperatureb Tb (◦ C)

Vapor pressureb (kPa)

20.3 18.3

92.5 106.3

114 111

4.80 (30 ◦ C) 4.00 (26 ◦ C)

Obtained from polymer handbook 4th edition [31]. Obtained from solvents handbook 4th edition [52].

Y. Li et al. / Applied Surface Science 257 (2011) 8093–8101

transmission illumination is used and hence the thicker or higher domains appear dark and the thinner or lower domains appear bright. TEM experiments were performed on a JEOL 1011 TEM with an accelerating voltage of 100 kV in the bright-field mode. For plane-view TEM observation, the films and their carbon support were floated off onto a pool of distilled water and then picked up with copper grids. To enhance the contrast between the PS and PMMA phases, the specimens were stained with RuO4 for several hours prior to observation. For cross-sectional TEM experiments, some portions of the floated film were collected onto a piece of cured epoxy resin and dried. After staining with RuO4 , these epoxy pieces were coated with a thin layer of carbon, and then these epoxy pieces (epoxy + carbon layer + polymer film + carbon layer) were embedded in epoxy resin and subsequently heated to 35, 45, and 55 ◦ C, respectively for 12 h. The ultrathin sections with approximately 60 nm in thickness were microtomed using a LEICA Ultracut R microtome and a glass knife at room temperature and collected onto the carbon-coated copper grids. Finally, the specimens of ultrathin sections were stained with RuO4 for several hours again prior to observation. Since RuO4 selectively react with the phenyl of the PS block, the PS nanodomains appear dark and the PMMA nanodomains appear bright in the TEM micrographs. 3. Results According to the Flory–Huggins theory [31,32], polymer and solvent are completely miscible over the entire composition range when the obtained value  < 0.5. Hence, based on the characteristic

8095

of the annealing solvents listed in Table 1 and the calculated value of P–S shown in Table 2, both Tri-CE and Tol with similar volatility and accordingly similar vapor pressure used in this work should be good solvents for both PS and PMMA blocks but having opposite preferential affinity for each block, Tri-CE has preferential affinity for the minority PMMA block and Tol has preferential affinity for the majority PS block. Therefore, the binary mixtures of the two solvents are good for both two blocks and the preferential affinity of the binary mixed solvent vapors can be modulated by changing the mixed volume fraction of the two solvents. We describe the influence of the annealing solvent vapor pressure and its preferential affinity on the film morphology formation and transition as follows. 3.1. Influence of the solvent vapor pressure Upon annealing under Tri-CE vapor with different vapor pressures for 24 h, the surface of the block copolymer films spin-coated from Tri-CE exhibits different patterns. The surface of the films annealed under Tri-CE vapor with the vapor pressures lower than ∼0.81 P0 maintains flat (the micrographs are shown in Fig. 1). The surface of the films annealed under Tri-CE vapor with the vapor pressures higher than ∼0.81 P0 , however, forms terraced patterns (the micrographs are shown in Figs. 2 and 3). Fig. 1 shows the TEM micrographs of the films annealed under Tri-CE vapors for 24 h with the pressures lower than ∼0.81 P0 . Fig. 1a displays the morphology of the film unannealed under the solvent vapor, that is, the intrinsic morphology formed during the spin-coating process. It exhibits disordered dark PS nanodomains

Fig. 1. TEM micrographs of the films spin-coated from Tri-CE followed by annealing under Tri-CE vapor for 24 h with the vapor pressures lower than ∼0.81 P0 : (a) the unannealed film, (b) ∼0.69 P0 , (c) ∼0.75 P0 , (d and d ) ∼0.81 P0 . The black and the white arrows in (c) denote the dark PS and the white PMMA nanodomains respectively. The micrographs in d and d are the plane-view and the corresponding cross-sectional TEM micrographs, respectively.

8096

Y. Li et al. / Applied Surface Science 257 (2011) 8093–8101

Fig. 2. Optical micrographs of the films spin-coated from Tri-CE followed by annealing under Tri-CE vapor for 24 h with the vapor pressures higher than ∼0.88 P0 : (a) ∼0.88 P0 , (b) ∼0.94 P0 . The letters T0 and T1 in both two micrographs indicate the lower terrace and the higher terrace respectively.

dispersing in a bright PMMA matrix. When the film is annealed under Tri-CE vapor with the pressure ∼0.69 P0 for 24 h the intrinsic morphology is maintained, as shown in Fig. 1b. Obviously, the morphology shown in Fig. 1b is very similar to that in Fig. 1a. As the Tri-CE vapor pressure increases to ∼0.75 P0 , the intrinsic morphology has been destroyed, the TEM micrograph of the film is shown in Fig. 1c, in which some dark circular and striped PS nanodomains (marked by black arrows in Fig. 1c) and bright striped PMMA nanodomains (marked by white arrow in Fig. 1c) are observed. Increasing the Tri-CE vapor pressure to ∼0.81 P0 , there are no lateral morphologies are detected at the surface of the film, as shown in the plane-view TEM micrograph in Fig. 1d. And the cross-sectional TEM micrograph of this film in Fig. 1d in which the two black layers (marked by C1 and C2) sandwiching the film PS–b–PMMA with 80 ± 2 nm thick (marked by AB) are carbon coatings introduced during the sample preparation also shows no any microphase separated morphologies. Fig. 1d also shows that there is a thin layer of PMMA (marked by A1) sandwiched by the

black carbon layer C2 at bottom of the film and the homogeneous block copolymer AB. Combined these two micrographs as shown in Fig. 1d and d , it is expected that a homogeneous morphology forms in the film annealed under Tri-CE vapor with the pressure ∼0.81 P0 for 24 h. Then the morphology shown in Fig. 1c can be defined a transitional morphology, suggesting a transitional morphology from a microphase separated morphology formed during the spin-coating process to a homogeneous one. As the Tri-CE vapor pressure increases to ∼0.88 P0 and ∼0.94 P0 , there form terraced patterns at the surface of the films, as shown in the OM micrographs in Fig. 2a and b respectively, in which the bright and the dark domains correspond to the lower terrace T0 and the higher terrace T1 , respectively. A remarkable characteristic of the two OM micrographs shown in Fig. 2 is that the fraction of the T1 area annealed under ∼0.94 P0 (Fig. 2b) is obviously larger than that annealed under ∼0.88 P0 (Fig. 2a). The thicknesses determined from cross-sectional TEM micrographs of the lower terrace T0 (35 ± 2 nm, data not shown) and the higher terrace T1 (103 ± 5 nm, shown

Fig. 3. TEM micrographs of the films spin-coated from Tri-CE followed by annealing under Tri-CE vapor for 24 h with the vapor pressures higher than ∼0.88 P0 : (a and a ) the plane-view and the cross-sectional TEM micrographs within T1 of the film annealed under ∼0.88 P0 , (b and b ) the plane-view and the cross-sectional TEM micrographs within T1 of the film annealed under ∼0.94 P0 . The black and the white arrows in (a) denote the dispersed circular and striped PS nanodomains.

Y. Li et al. / Applied Surface Science 257 (2011) 8093–8101

below) of the two films obtained from annealing under ∼0.88 P0 and ∼0.94 P0 are similar to each other. However, different morphologies are observed within the lower terrace T0 and the higher terrace T1 . Within the lower terrace T0 there is no distinct microphase separated morphology is observed (data not shown) indicating that a layer of block copolymer brush is formed [33–36]. Within the higher terrace T1 , however, a distinct microphase separated morphology is found (Fig. 3). Fig. 3a shows the plane-view TEM micrograph of T1 of the film annealed under Tri-CE vapor with the pressure ∼0.88 P0 for 24 h, which presents a morphology of dark circular and striped PS nanodomains (marked by black and white arrows respectively) dispersing in a PMMA matrix. Nevertheless, the cross-sectional TEM micrograph of this film (Fig. 3a ) displays that the bright PMMA (marked by A1) forms the dispersed nanodomains surrounded by PS nanodomains (marked by B1 and B2). Combining these two micrographs in Fig. 3a and a , the morphology obtained within T1 in the film annealed under Tri-CE vapor with the pressure ∼0.88 P0 for 24 h is expected to be a perforated lamellar morphology, and the lamella of the minority PMMA block are perforated by striped and circular channels of the majority PS block [8,30,37,38]. Since the dominating PS channels are striped, the morphology formed within T1 in the film annealed under the solvent vapor pressure ∼0.88 P0 for 24 h (Fig. 3a and a ) is defined as perforated lamella with striped PS channels. Meanwhile, Fig. 3b shows the plane-view TEM micrographs of T1 in the film annealed under the solvent vapor pressure ∼0.94 P0 for 24 h in which the dispersed nanodomians are PS. The cross-sectional TEM micrograph of the film (Fig. 3b ) displays that the bright dispersed nanodomains (marked by A1) surrounded by dark PS nanodomians (marked by B1and B2) are PMMA, indicating that the lamella of the minority PMMA block are perforated by circular channels of the majority PS block. 3.2. Influence of the solvent vapor preferential affinity In this work, the preferential affinity of the annealing solvent vapor is modulated by changing the mixed volume fraction of two solvents, Tri-CE and Tol, due to both two solvents with similar vapor pressures are good solvents for both PS and PMMA blocks but having opposite preferential affinity for each block, Tri-CE has preferential affinity for the minority PMMA block and Tol has preferential affinity for the majority PS block. An expression  (S-PMMA − S-PS ) [30], which can be modulated by controlling the volume fraction of the two solvents Tri-CE/Tol (VTri-CE /VTol ), is used to represent the preferential affinity of a solvent vapor for a certain block. That is,  < 0, =0, and >0 indicate the solvent vapor is preferential affinity for the minority PMMA block, neutral for each block, and preferential affinity for the majority PS block, respectively. The characteristics of the solvent vapors used in this work are listed in Table 3. According to the value of , with decreasing the volume ratio of Tri-CE/Tol (VTri-CE /VTol ) from 100/0 to 0/100, the preferential affinity of the binary mixed solvent vapors of the two solvents for the minority PMMA block decreases. Fig. 4 shows the optical micrographs of the final morphologies of the films spin-coated from Tri-CE followed by annealing under Tri-CE, Tol, and their binary mixed vapors with the pressure ∼0.94 P0 for 24 h. Obviously, different surface patterns form depending on the preferential affinity of the annealing solvent vapors in the films with similar thickness. Terraced surface pattern forms in the films annealed under ∼0.94 P0 solvent vapor with  < 0, Fig. 4a shows

8097

a typical OM micrograph of the pattern at  ∼ −0.063. There is no any terraced pattern forms in the films annealed under ∼0.94 P0 solvent vapor with  ∼ 0 (Fig. 4b). The terraced surface pattern forms again in the films annealed under ∼0.94 P0 solvent vapor with  > 0, and Fig. 4c shows a typical OM micrograph of the pattern at  ∼ 0.030. It should be noted that the morphologies (see Fig. 7) within the lower terrace and the higher terrace shown in Fig. 4c are different from those (see Fig. 5) in Fig. 4a. Therefore, the lower terrace and the higher terrace are marked by T0 and T1 respectively in Fig. 4a, but the lower terrace and the higher terrace are marked by TL and TH respectively in Fig. 4c. Then the detail morphologies of the lower terraces and the higher terraces are detected by TEM. TEM observation suggests that the lower terrace T0 of all the films annealed under solvent vapor with  < 0 form a layer of block copolymer brush (date not shown). Within the higher terrace T1 , however, distinct microphase separated morphologies are formed (Fig. 5). As mentioned above when the film annealed under ∼0.94 P0 solvent vapor with  ∼ −0.063 (pure Tri-CE) for 24 h, a morphology of hexagonally perforated lamella with circular PS channels is obtained (Fig. 3b and b ). When the film annealed under ∼0.94 P0 solvent vapor with  ∼ −0.054, a morphology consisting of binary circular and striped PS nanodomains dispersing in a PMMA matrix (Fig. 5a) forms. Fig. 5a shows the cross-sectional TEM micrograph of the film, in which the bright PMMA (marked by A1) surrounded by dark PS (B1and B2) forms the dispersed nanodomains. Combining the two micrographs in Fig. 5a and a , we expect the morphology in the film annealed under ∼0.94 P0 solvent vapor with  ∼ −0.054 is a perforated lamellar phase [33], and lamella of the minority PMMA block is perforated by circular and striped channels of the PS block. As the preferential affinity of the solvent vapor  increases further to ∼−0.037, the planeview and cross-sectional TEM micrographs are shown in Fig. 5b (the dispersed nanodomians are PS) and Fig. 5b (the dispersed nanodomians are PMMA), which clearly show the dominating morphology is lamella of the minority PMMA block perforated by striped channels of the PS block. As  further increases to ∼−0.020, the plane-view and cross-sectional TEM micrographs are shown in Fig. 5c (the dispersed nanodomians are PS) and Fig. 5c (the dispersed nanodomians are PMMA), which suggest that the whole terrace T1 forms a morphology of lamella of the minority PMMA block perforated by striped channels of the PS block. When the preferential affinity of the solvent vapor increases to  ∼ 0 (a neutral solvent vapor), a disordered morphology is found in the film annealed under ∼0.94 P0 for 24 h (Fig. 6). It exhibits disordered bright PMMA nanodomains dispersing in a dark PS matrix. As the preferential affinity increases further to  > 0, terraced surface patterns form again in the film annealed under ∼0.94 P0 for 24 h (Fig. 4c), whereas the thickness of the lower terrace TL (70 ± 2 nm) and the higher terrace TH (130 ± 5 nm) determined from the cross-sectional TEM micrographs (data not shown) is different from that of the lower terrace T0 (35 ± 2 nm) and the higher terrace T1 (103 ± 5 nm) in the films annealed under ∼0.94 P0 solvent vapor with the preferential affinity  ∼ −0.063 (Fig. 4a). Fig. 7 shows the TEM micrographs of the lower terrace TL and the higher terrace TH of the film annealed under ∼0.94 P0 solvent vapor with the preferential affinity  > 0. The TEM micrograph of the film annealed under ∼0.94 P0 solvent vapor with the preferential affinity  ∼ 0.021 is shown in Fig. 7a. Obviously, both terraces TL and TH exhibit similar ordered morphology of whole-parallel PMMA cylinders. With the solvent preferential affinity for PS block increasing

Table 3 Preferential affinity of the binary mixed solvent vapors. VTri-CE /VTol

100/0

90/10

70/30

50/50

30/70

10/90

0/100



−0.063

−0.054

−0.037

−0.020

0

0.021

0.030

8098

Y. Li et al. / Applied Surface Science 257 (2011) 8093–8101

Fig. 4. Optical micrographs of the films spin-coated from Tri-CE followed by annealing under Tri-CE, Tol, and their binary mixed vapors for 24 h with the vapor pressure ∼0.94 P0 . (a)  < 0, (b)  ∼ 0, (c)  > 0. The letters T0 and T1 in (a), TL and TH in (c) indicate the lower terraces and the higher terraces respectively.

Fig. 5. TEM micrographs of the films spin-coated from Tri-CE followed by annealing under the solvent vapors with preferential affinity for the minority PMMA block  < 0 with the vapor pressure ∼0.94 P0 for 24 h. The preferential affinities of the solvents are (a and a ) ∼−0.054, (b and b ) ∼−0.037, (c and c ) ∼−0.020. The parts a , b , and c show the cross-sectional TEM micrographs corresponding to plane-view micrographs in a, b, and c, respectively.

further to  ∼ 0.030 (pure Tol), a morphology of hybrid PMMA cylinders and spheres forms within both TL and TH as shown in Fig. 7b. 4. Discussion

Fig. 6. TEM micrograph of the film spin-coated from Tri-CE followed by annealing under the neutral solvent  ∼ 0 with the vapor pressure ∼0.94 P0 for 24 h.

It is well known that when block copolymer thin film is exposed to a good solvent vapor, the polymer chains in each block will be swollen and hence mobile [10–14,30,39]. In our case as the above description, with the Tri-CE vapor pressure increasing from 0 (unannealing) to ∼0.94 P0 , the degree of the swelling and accordingly the mobility of PS and PMMA chains increase. The results (Figs. 1–3) indicate that there exist three stages of the morphology evolution during the process of increasing the annealing solvent vapor pressures: maintain the intrinsic morphology formed during the spin-coating process under lower pressures (lower than ∼0.69 P0 ), destroy the intrinsic morphology under moderate pressures (∼0.75 P0 to ∼0.81 P0 ), and build new morphologies under higher pressures (higher than ∼0.88 P0 ). When the annealing solvent vapor pressures are lower than ∼0.69 P0 , the mobility of PS and PMMA chains is too low to destroy the intrinsic morphology (disordered dark PS nanodomains dispersing in a bright PMMA matrix [40] (Fig. 1a and b)). With the annealing solvent vapor pressure increas-

Y. Li et al. / Applied Surface Science 257 (2011) 8093–8101

8099

Fig. 7. TEM micrographs of the films spin-coated from Tri-CE followed by annealing under the solvent vapors with preferential affinity for the majority block PS ( > 0) with the vapor pressure ∼0.94 P0 for 24 h. The preferential affinities of the solvents are (a) ∼0.021, (b) ∼0.030. The letters TL and TH in both two micrographs indicate the lower terrace and the higher terrace respectively.

ing to ∼0.75 P0 , the mobility of PS and PMMA chains can destroy the intrinsic morphology and some deformed nanodomains (Fig. 1c) are observed. As further increases the annealing solvent vapor pressure to ∼0.81 P0 , the mobility of PS and PMMA chains lead the film form a homogeneous morphology (Fig. 1d and d ). We only observe a homogeneous morphology here. It is expected that the mobility of PS and PMMA chains should be responsible for the interesting morphology formation in the film annealed under solvent vapor with pressure ∼0.81 P0 . Although the mobility of PS and PMMA chains can destroy the intrinsic morphology, it is not strong enough to help the film form a terraced surface under this moderate vapor pressure as ∼0.81 P0 . Moreover, the film thickness used here is only 80 ± 2 nm determined from the cross-sectional TEM micrograph in Fig. 1d , which will constraint the polymer chains to rearrange to form an ordered morphology despite there forms a thin layer of PMMA at the bottom of the film due to the carbon coated substrate prefers slightly to the PMMA block. Thus, the homogeneous morphology (Fig. 1d and d ) forms in the film annealed under ∼0.81 P0 solvent vapor. With the annealing solvent vapor pressure increasing to higher than ∼0.88 P0 , the polymer chains of each block have sufficient mobility not only to destroy the intrinsic morphology but also to overcome the constraints of the film and hence to form terraced surface (Fig. 2a and b). As a result, the morphologies of perforated lamella with striped PS channels and hexagonally perforated lamella with circular PS channels (Fig. 3a and a , b and b ) are observed in the films annealed under Tri-CE vapor with the pressures ∼0.88 P0 and ∼0.94 P0 , respectively. It should be noted that the morphologies of the cylinder-forming PS–b–PMMA films annealed under Tri-CE vapor with vapor pressures ∼0.88 P0 and ∼0.94 P0 are perforated lamella. Recently, the observation of the morphologies of hexagonally perforated lamella in thin films of cylinder-forming and lamella-forming block copolymers have been reported by some other groups [8,10,24,26,38,41–47]. In thin films of a cylinder-forming block copolymer, the morphology of hexagonally perforated lamella is likely due to the surface-induced reorganization [10,24,44]. For the thin films of lamella-forming block copolymer, the morphological transition from the lamella to the hexagonally perforated lamella maybe due to the high humidity [8]. In the present work, however, the formation of the morphology of perforated lamella can be explained as follows. In general, the solvent with different preferential affinity for a certain block will influence the degree of the swelling of the polymer chains in each block, and hence result in different effective

volume fractions of each block [30,39]. It means that the existence of solvent will change the effective composition of the block copolymer and such temporary composition is related to the amount of the solvent. The solvent used here Tri-CE is good for both PS and PMMA blocks but has preferential affinity for the minority PMMA block. In such case, the solvent molecules will diffuse into PMMA block easier and lead to higher degree of the swelling of PMMA block than that of PS block. In other words, Tri-CE results in selective increase of the effective volume fraction of PMMA, namely higher than that in bulk 27%. Thus, the curvature of the interface between PS and PMMA nanodomains is different from that in bulk with hexagonally packed PMMA cylinders embedded in a PS matrix. When the effective volume fraction of the PMMA block is high enough and makes it form a perforated layer, the morphologies of perforated lamella (Fig. 3a and a , b and b ) become to form. Moreover, our results indicate that different morphologies of perforated lamella can be observed by controlling the annealing solvent vapor pressure and accordingly the effective volume fraction of the PMMA block precisely. When the film annealed under ∼0.94 P0 solvent vapor, the effective volume fraction of the PMMA block is high enough and the classical morphology of perforated lamella with circular PS channels (Fig. 3b and b ) is observed. However, when the vapor pressure decreases to ∼0.88 P0 , the effective volume fraction of the PMMA block correspondingly decreases due to the amount of solvent molecules decreases in the films. Therefore, there may not have enough PMMA chains to form perforated lamella with circular channels, they can only form perforated lamella with striped channels, and hence the morphology of perforated lamella with striped PS channels (Fig. 3a and a ) forms. In a word, this kind of morphology is achieved by controlling the degree of the swelling and the mobility of the polymer chains in each block, which can be implemented by precisely controlling the annealing vapor pressures of a solvent with preferential affinity for the minority PMMA block. On the other hand, when the film is exposed to Tri-CE, Tol, and their binary mixed solvents with different preferential affinities for each block under ∼0.94 P0 solvent vapor, the solvent molecules will diffuse easier into the block to which the solvent prefer and lead to higher degree of the swelling of this block than that of the other one. For instance, when the preferential affinity of a solvent  < 0 (the solvent vapor is preferential affinity for the minority PMMA block as the case of  ∼ −0.063, −0.054, −0.037, and −0.020), the solvent molecules will diffuse easier into the minority PMMA block and lead to higher degree of the swelling of PMMA block than that of PS block. In such case, the morphologies of perforated lamella

8100

Y. Li et al. / Applied Surface Science 257 (2011) 8093–8101

are observed and the perforating PS channels change from circular to striped (Figs. 3b, b and 5) as the preferential affinity of the solvents for the minority PMMA block decreases. On the contrary, when the preferential affinity of the solvent  > 0 (the solvent vapor is preferential affinity for the majority PS block as the case of  ∼ 0.021 and ∼0.030), the solvent molecules will diffuse easier into the majority PS block and lead to higher degree of the swelling of PS block than that of PMMA block. Hence the effective volume fraction of PS is higher than that in bulk. Therefore, there form the morphologies of whole-parallel PMMA cylinders ( ∼ 0.021, Fig. 7a) and even hybrid PMMA cylinders and spheres ( ∼ 0.030, Fig. 7b). When the preferential affinity of the solvent  ∼ 0 (the solvent vapor is neutral to each block), the solvent molecules may diffuse equal into both PS and PMMA blocks and lead to the degree of the swelling of both two blocks equally, then a disordered morphology (Fig. 6) is observed. This can be explained by using an expression of the effective Flory–Huggins interaction parameter   (eff ) eff ∼ϕ(AB + ) = ϕ(AB + A-S − B-S ) [9,48–50], where ϕ is the volume concentration of copolymer in a solvent and  is the absolute difference between A-solvent and B-solvent interaction parameters A-S and B-S . Obviously, the value of eff N of a block copolymer in a neutral solvent ( ∼ 0) is smaller than that in a preferential affinity solvent ( > 0), which results in the lower degree of microphase separation [51]. Therefore, a disordered morphology (Fig. 6) can be observed in the film annealed under a neutral solvent vapor with the vapor pressure ∼0.94 P0 . This explanation can be confirmed by the ensuing data (Fig. 8). As the volume concentration of copolymer in the solvent ϕ increases, namely, the solvent vapor pressure decreases to ∼0.88 P0 , the morphology of parallel PMMA cylinders is obtained (see Fig. 8). Based on the observed morphologies of PS–b–PMMA annealed under different solvent vapors, a phase diagram by plotting the morphologies as a function of the annealing solvent vapor pressure and the solvent vapor preferential affinity is constructed, as shown in Fig. 8. The phase diagram clearly shows that there exist three stages of the morphology evolution during the process of increasing the annealing solvent vapor pressure for all solvents: First, maintain the intrinsic morphology formed during the spincoating process under lower solvent vapor pressures (green region,

detailed description can be recalled to Fig. 1b). Second, destroy the intrinsic morphology under moderate solvent vapor pressures (yellow region, detailed description can be recalled to Fig. 1c and d. In this region, the transitional morphologies from the intrinsic morphology (Fig. 1a) to parallel PMMA cylinders (Fig. 7a) or hybrid PMMA cylinders and PMMA spheres (Fig. 7b) marked by 夽 formed at the preferential affinity  = 0.010–0.030 and the vapor pressure 0.81 P0 –0.88 P0 are shown in Supporting Information in Fig. S1). Third, build new morphologies under higher solvent vapor pressures (gray region, detailed description can be recalled to Figs. 3–7). This finding suggests that a higher vapor pressure which can induce enough mobility of the block copolymer chains should be used to build new morphologies during the process of solvent vapor annealing. Moreover, when the solvent vapor pressure is high enough, the morphology is also strongly influenced by the preferential affinity of the solvent vapors. For instance, when the films annealed under ∼0.94 P0 solvent vapor, the morphology evolves from hexagonally perforated lamella with circular PS channels (Fig. 3b and b ), perforated lamella with striped PS channels (Fig. 5b and b , c and c ), through parallel PMMA cylinders (Fig. 7a), to the morphology of hybrid PMMA cylinders and PMMA spheres (Fig. 7b) as the preferential affinity of the solvents for the minority PMMA block  decreases from ∼−0.063 to ∼0.030. In the phase diagram, the morphology of perforated lamella with striped PS channels obtained both at  ∼ −0.063, ∼0.88 P0 (Fig. 3a and a ) and  ∼ −0.020, ∼0.94 P0 (Fig. 5c and c ), indicating that exposure a film of PS–b–PMMA with PMMA cylinder forming composition under a solvent vapor with stronger preferential affinity for the minority PMMA block  ∼ −0.063 at lower solvent vapor pressure ∼0.88 P0 and under a solvent vapor with weaker preferential affinity for the minority PMMA block  ∼ −0.020 at higher solvent vapor pressure ∼0.94 P0 have equal effect on the morphology formation. It means that increasing of the annealing vapor pressure of a solvent with preferential affinity for the minority block can offset the decreasing of the solvent preferential affinity for the minority block on a morphology formation. Our results suggest that both of the annealing solvent vapor pressure and solvent vapor preferential affinity which can control the degree of the swelling and hence the mobility of the block copolymer chains in each block are expected to be the key factors in controlling the complex morphology formation and transition. Moreover, combining the results in this work and our previous work [29], one can conclude that the morphologies of a cylinder-forming block copolymer will undergo inverted spheres, inverted cylinders, inverted hexagonal perforated lamella, lamella, perforated lamella, and finally normal cylinders with increasing the concentration of the copolymer in a solvent which is preferential affinity for the cylinder-forming block.

5. Conclusions

Fig. 8. Phase diagram plotted as a function of solvent vapor pressure and preferential affinity (). () hexagonally perforated lamella with circular PS channels; () perforated lamella with striped PS channels; (♦) disordered morphology; () parallel PMMA cylinders; () morphology of hybrid PMMA cylinders and PMMA spheres; (夽) transitional morphologies; () homogeneous morphology. The green region denotes the stage of maintaining the intrinsic morphology under lower vapor pressures, the yellow region denotes the stage of destroying the intrinsic morphology under moderate vapor pressures, and the gray region denotes the stage of building new morphologies under higher solvent vapor pressures, respectively. All lines and regions are drawn to guide the eye. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

The morphology formation and transition of PS–b–PMMA diblock copolymer thin films (∼70 nm in thickness) annealed under solvent vapor with different vapor pressures and preferential affinities have been investigated. The solvent vapors used are Tri-CE, Tol, and their binary mixed solvent vapors. Tri-CE and Tol, with similar volatility, are good solvents for both PS and PMMA blocks but having opposite preferential affinity for each block. With increasing the annealing solvent vapor pressures, there exist three stages of the morphology evolution: maintain the intrinsic morphology formed during the spin-coating process under lower pressures, destroy the intrinsic morphology under moderate pressures, and build new morphologies under higher pressures. And with the preferential affinity of the solvent vapors for the minority PMMA block decreasing, the morphology transition from hexagonally

Y. Li et al. / Applied Surface Science 257 (2011) 8093–8101

perforated lamella with circular PS channels, perforated lamella with striped PS channels, parallel PMMA cylinders, to hybrid PMMA cylinders and spheres are observed. Based on the observation, a phase diagram as a function of the annealing solvent vapor pressures and their preferential affinities is constructed. The results indicate that both the annealing solvent vapor pressure and its preferential affinity can influence the degree of the swelling and the mobility of the polymer chains in each block and hence play a crucial role in the complex morphologies formation. Acknowledgments We are grateful to Ms. Guifen Sun for technical help with the microtomy. This work is supported by National Natural Science Foundation of China (20774095 and 21074135) and subsidized by National Basic Research Program of China (2005CB623806) and Liaoning Province Foundation for Doctors (910502). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.apsusc.2011.04.113. References

[15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35]

[36] [37] [38]

[1] I.W. Hamley, The Physics of Block Copolymers, Oxford University Press, New York, 1998. [2] V. Abetz, P.F.W. Simon, Adv. Polym. Sci. 189 (2005) 125. [3] M.J. Fasolka, A.M. Mayes, Annu. Rev. Mater. Res. 31 (2001) 323. [4] F.S. Bates, G.H. Fredrickson, Annu. Rev. Phys. Chem. 41 (1990) 525. [5] J.N.L. Albert, T.H. Epps, Mater. Today 13 (2010) 24. [6] T.P. Lodge, Macromol. Chem. Phys. 204 (2003) 265. [7] T. Pietsch, E. Metwalli, S.V. Roth, R. Gebhard, N. Cindy, P. Müller-Buschbaum, A. Fahmi, Macromol. Chem. Phys. 210 (2009) 864. [8] J. Bang, B.J. Kim, G.E. Stein, T.P. Russell, X. Li, J. Wang, E.J. Kramer, C.J. Hawker, Macromolecules 40 (2007) 7019. [9] W.V. Zoelen, T. Asumaa, J. Ruokolainen, O. Ikkala, G.T. Brinke, Macromolecules 41 (2008) 3199. [10] A. Knoll, R. Magerle, G. Krausch, J. Chem. Phys. 120 (2004) 1105. [11] K.A. Cavicchi, K.J. Berthiaume, T.P. Russell, Polymer 46 (2005) 11635. [12] S.H. Kim, M.J. Minser, T. Xu, M. Kimura, T.P. Russell, Adv. Mater. 16 (2004) 226. [13] U. Zettl, A. Knoll, L. Tsarkova, Langmuir 26 (2010) 6610. [14] Z. Di, D. Posselt, D.M. Smilgies, C.M. Papadakis, Macromolecules 43 (2010) 418.

[39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52]

8101

Y. Chen, H. Huang, Z. Hu, T. He, Langmuir 20 (2004) 3805. Y. Chen, Z. Wang, Y. Gong, H. Huang, T. He, J. Phys. Chem. B 110 (2006) 1647. R. Guo, H. Huang, Y. Chen, Y. Gong, B. Du, T. He, Macromolecules 41 (2008) 890. R. Guo, H. Huang, B. Du, T. He, J. Phys. Chem. B 113 (2009) 2712. C. Tang, J. Bang, G.E. Stein, G.H. Fredrickson, M. Sprung, J. Wang, E.J. Kramer, C.J. Hawker, Macromolecules 41 (2008) 4328. S.-H. Tung, T. Xu, Macromolecules 42 (2009) 5761. H. Elbs, C. Drummer, V. Abetz, G. Krausch, Macromolecules 35 (2002) 5570. J. Peng, Y. Xuan, H. Wang, Y. Yang, B. Li, Y. Han, J. Chem. Phys. 120 (2004) 11163. Y. Xuan, J. Peng, L. Cui, H. Wang, B. Li, Y. Han, Macromolecules 37 (2004) 7301. A. Knoll, A. Horvat, K.S. Lyakhova, G. Krausch, G.J.A. Sevink, A.V. Zvelindovsky, R. Magerle, Phys. Rev. Lett. 89 (2002) 035501. K.A. Cavicchi, T.P. Russell, Macromolecules 40 (2007) 1181. Y.S. Jung, C.A. Ross, Adv. Mater. 21 (2009) 2540. G. Lu, L. Li, X. Yang, Small 4 (2008) 601. G. Lu, L. Li, X. Yang, Macromolecules 41 (2008) 2062. Y. Gong, H. Huang, Z. Hu, Y. Chen, D. Chen, Z. Wang, T. He, Macromolecules 39 (2006) 3369. Y. Li, H. Huang, T. He, Y. Gong, J. Phys. Chem. B 114 (2010) 1264. D.W. Van Krevelen, Properties of Polymers, Elsevier Scientific Publishing, Company, Amsterdam–Oxford–New York, 1976. J. Brandrup, E.H. Immergut, E.A. Grulke, Polymer Handbook, 4th ed., John Wiley & Sons, New York, 1999. R. Limary, P.F. Green, Langmuir 15 (1999) 5617. R. Limary, P.F. Green, Macromolecules 32 (1999) 8167. Y. Liu, W. Zhao, X. Zheng, A. King, A. Singh, M.H. Rafailovich, J. Sokolov, K.H. Dai, E.J. Kramer, S.A. Schwarz, O. Gebizlioglu, S.K. Sinha, Macromolecules 27 (1994) 4000. J.P. Spatz, M. MÖller, M. Noeske, R.J. Behm, M. Pietralla, Macromolecules 30 (1997) 3874. L.H. Radzilowski, B.L. Carvalho, E.L. Thomas, J. Polym. Sci. Part B: Polym. Phys. 34 (1996) 3081. I. Park, S. Park, H.W. Park, T. Chang, H. Yang, C.Y. Ryu, Macromolecules 39 (2006) 315. Y. Funaki, K. Kumano, T. Nakao, H. Jinnai, H. Yoshida, K. Kimishima, K. Tsutsumi, Y. Hirokawa, T. Hashimoto, Polymer 40 (1999) 7147. S. Gourianova, J. Fuhrmann, Macromolecules 37 (2004) 1825. L. Tsarkova, A. Knoll, G. Krausch, R. Magerle, Macromolecules 39 (2006) 3608. A. Sperschneider, F. Schacher, M. Gawenda, L. Tsarkova, A.H.E. Müller, M. Ulbricht, G. Krausch, J. Köhler, Small 3 (2007) 1056. S. Ludwigs, K. Schmidt, G. Krausch, Macromolecules 38 (2005) 2376. S. Ludwigs, A. Böker, A. Voronow, N. Rehse, R. Magerle, G. Krausch, Nat. Mater. 2 (2003) 744. W. Sriprom, M. James, S. Perrier, C. Neto, Macromolecules 42 (2009) 3138. Y. Yang, F. Qiu, H. Zhang, Y. Yang, Polymer 47 (2006) 2205. P. Chen, H. Liang, A.-C. Shi, Macromolecules 40 (2007) 7329. Y. Li, X. Wang, I.C. Sanchez, K.P. Johnston, P.F. Green, J. Phys. Chem. B 111 (2007) 16. B. Nandan, M.K. Vyas, M. BÖhme, M. Stamm, Macromolecules 43 (2010) 2463. W.A. Phillip, M.A. Hillmyer, E.L. Cussler, Macromolecules 43 (2010) 7763. L. Tsarkova, Progr. Collid Polym. Sci. 137 (2010) 1. N. Cheng, Solvents Handbook, Chemical Industry Press, Bei Jing, 2008.