The Crystallization of Confined Polymers and ... - ACS Publications


María Teresa Casas , Rose Mary Michell , Iwona Blaszczyk-Lezak , Jordi ... Jon Maiz , Wei Zhao , Yu Gu , Jimmy Lawrence , Arantxa Arbe , Angel AlegrÃ...
0 downloads 0 Views 571KB Size


Article pubs.acs.org/Macromolecules

The Crystallization of Confined Polymers and Block Copolymers Infiltrated Within Alumina Nanotube Templates Rose Mary Michell,† Arnaldo T. Lorenzo,† Alejandro J. Müller,*,† Ming-Champ Lin,‡ Hsin-Lung Chen,‡ Iwona Blaszczyk-Lezak,§ Jaime Martín,§ and Carmen Mijangos§ †

Grupo de Polímeros USB, Departamento de Ciencia de los Materiales, Universidad Simón Bolívar, Apartado 89000, Caracas 1080-A, Venezuela ‡ Department of Chemical Engineering and Frontier Center of Fundamental and Applied Sciences of Matters, National Tsing Hua University, Hsin-Chu 30013, Taiwan § Instituto de Ciencia y Tecnología de Polímeros, CSIC, Juan de la Cierva, 3, 28006 Madrid, Spain ABSTRACT: Porous anodic aluminum oxide (AAO) templates with 20, 35, and 60 nm cylindrical pores were prepared and subsequently infiltrated with poly(ethylene oxide), PEO, polyethylene, PE, and polyethylene-block-polystyrene diblock copolymers, PE-b-PS, of various compositions. The crystallization of the nanocylinders within the AAO templates was studied by differential scanning calorimetry (DSC) and wide angle X-ray diffraction (WAXD). A clear change from heterogeneous to homogeneous bulk nucleation was detected in the PEO before and after infiltration within a 20 nm template, respectively. The homogeneously nucleated nanocylinders needed extreme supercooling in order to crystallize, and their low crystallization temperature was successfully correlated with the volume of the crystallizing phase. 2D-WAXD measurements demonstrated that the PEO chains within the crystals formed inside the nanocylinders, preferentially orient perpendicular to the cylinder axis. This chain orientation is probably due to the easier crystal growth propagation along the cylinder length (200 μm for the templates employed to infiltrate PEO). In the case of the infiltrated PE, although its crystallization temperature was also lowered by confinement within a large number of nanocylinders, its value is still at least 75 °C above Tg; therefore, its nucleation is probably originated at the surface of the nanocylinders and cannot be considered homogeneous bulk nucleation. Strongly segregated semicrystalline PE-b-PS diblock copolymers were infiltrated into AAO templates for the first time, thereby creating a nanostructured hybrid material where the PE phase experienced double confinement for some compositions (i.e, the phasesegregated confinement within a vitreous PS matrix and the physical confinement within the nanopores). Fractionated crystallization was observed for the PE block microdomains within both neat copolymers and infiltrated ones. However, in the case of the infiltrated block copolymers, the dominant crystallization was always that produced at maximum supercooling, as expected for the extreme confinement environment. In the cases of PE blocks that self-assemble to form 42 nm diameter cylinders and 24 nm spheres within the neat block copolymers, the PE phase crystallization was impossible to detect once they were infiltrated into the 35 nm templates, as opposed to the 60 nm case where the lowest crystallization and melting temperatures were recorded. The crystallization and melting temperatures decreased as the volume of the crystallizing phase decreased (i.e., with a decrease in template pore volume, for neat PE, or PE phase volume within neat of infiltrated copolymers). The nucleation of the confined PE block microdomains is probably originated at their surface.



INTRODUCTION Nanostructured materials have attracted much scientific and technological attention.1−7 In a semicrystalline polymer, its crystallization process and final morphology determine the end properties of the material and its application.2,5 The confinement of the polymer within nanometer environments (e.g., ultrathin films, block copolymer nanophases and more recently, inorganic templates) can dramatically altered both the nucleation and growth processes.8−15 Linear diblock copolymers in the strong segregation limit undergo phase segregation to form microdomains such as spheres, cylinders, gyroids, and lamellae. If these microdomain © 2012 American Chemical Society

morphologies are formed by crystalline blocks, their crystallization process can be effectively confined within the microdomains with the degree of confinement governed by their morphology.10 Another way to prepare 1D polymer nanostructures is through the use of porous anodic aluminum oxide (AAO) templates and subsequent infiltration of the polymer into the nanocavities of such templates. Templates of this type are prepared Received: October 17, 2011 Revised: November 24, 2011 Published: January 31, 2012 1517

dx.doi.org/10.1021/ma202327f | Macromolecules 2012, 45, 1517−1528

Macromolecules

Article

block copolymers in AAO templates has an influence on the morphology of the microdomains.9,38−55 The changes in the morphology depends on the relationship between the size of the microdomain and the pore diameter. For example, a nearly symmetric PMMA-b-PS diblock copolymer was studied by Sun et al.38 and they found that the lamellar morphology was transformed to concentric nanotubes in pores bigger than the microdomain size. Xiang et al.40 studied the morphology of PSb-PB copolymers infiltrated in an AAO template. In this case, two different compositions were studied corresponding to lamellae and cylinders. The cylinders were aligned with the pores, while the lamellae arranged into concentric cylinders when the pore size was larger than the microdomains,40 but when the pore diameter was similar to the MDs size, then a completely different morphology was obtained, since the cylinders transformed into helicoidal or toroidal forms.42−51 In this paper, we study the crystallization phenomena of crystallizable polymers and copolymers within AAO nanopores. For this purpose, three systems were selected: a PEO homopolymer, a hydrogenated polybutadiene or model polyethylene (PE) and several anionically polymerized linear PE-b-PS diblock copolymers. The study of the crystallization of block copolymers into AAO templates is presented here for the first time. The different confined environments generated for the crystallizable PE (i.e., AAO nanopores) and PE blocks within the copolymers upon infiltration into AAO templates are compared. We also analyze the low temperature crystallization of the infiltrated PEO and PE and consider whether the crystals originate from homogeneous bulk nuclei or surface nuclei.

electrochemically from aluminum metal. They are highly versatile with respect to the diameter and length of the obtained 1D structure. AAO templates are characterized by uniform tailored pores with long-range ordered hexagonal symmetry in arrangement. The porous AAO templates were first synthesized by Masuda and Fukuda,16 and further developed for the fabrication of 1D polymer nanostructures by means of nanomolding processes.17−21 As any molding process, it consists first of the infiltration of a polymeric melt fluid into nanocavities with a well-defined shape, then the polymer is solidified within the cavities, and finally the molded polymeric material is removed (when a free polymeric nanostructure is required). The use of polymer filled inorganic templates has caused a large impact because of the versatility of these novel hybrid systems.1,9 In this case the microdomains are formed by the material filling the pores of the template. The length and the diameter can be adjusted through the synthesis of the template. Different polymers have been infiltrated in alumina templates, but the crystallization within the nanopores has been studied only recently.4,5,22−30 Woo et al.27 infiltrated polyethylene in alumina templates. They found a clear dependence of the crystallization temperature (Tc) on the pore diameter. When the diameter was smaller, Tc was lower. They also performed isothermal crystallization studies and fitted the results with the Avrami equation. They found values of the Avrami index of around 1.7.27 Similar results have been found for infiltrated isotactic polypropylene (PP),4 poly(vinylidene fluorideco-trifluoroethylene),24 and syndiotactic polystyrene (sPS).23 The authors claimed that they have found homogeneous nucleation of the infiltrated materials. However, the nucleation at the pore surface was not ruled out and in some cases not even considered. Lutkenhaus et al.24 found that infiltration of poly(vinylidene fluoride-co-trifluoroethylene) within AAO pores induces the formation of a β polymorphism.24 Similarly, Wu et al.23,29 have shown that sPS changes its crystal modification from β in the bulk to mostly α upon infiltration.23,29 Polyaniline (PANI) also changes its unit cell from orthorhombic to pseudo-orthorhombic after infiltration,30 while in PVDF a change from α to γ form has been reported.22 In the case of block copolymers, when the crystallizable block forms cylinders, it has been shown that the chain orientation is a function of the supercooling employed during crystallization. At moderate supercoolings, the chain orientation within the cylinders is usually perpendicular to the cylinder axis.31−37 In the templates a preferential orientation of the crystals has also been found. Shin et al.5 studied the crystal orientation within the nanopores. They found that the chain in the crystals are perpendicular to the pore axis.5 The same orientation of the crystals was found in sPS28,29 and poly(vinylidene difluoride) (PVDF).26 Steinhart et al.25 studied the crystal orientation of the polyvinylidene fluoride (PVDF) infiltrated in an AAO template. They obtained two different situations, if the nanopores are independent and isolated no preferential orientation was found. On the other hand, if the nanopores are connected through a film on the top of the template they found a preferential orientation, the crystal grew in the fastest direction and the chain was perpendicular to the pore axis.25 Garcia-Gutierrez et al., also found a perpendicular orientation of crystalline PDVF chains within the pores.22 The study of block copolymers infiltrated within AAO templates has been oriented to the study of morphological changes, but so far, not to the crystallization of such systems. The infiltration of



EXPERIMENTAL PART

Materials. A poly(ethylene oxide) sample was purchased from Polymer Source Inc. with a number-average molecular weight of 41 000 g/mol and a polydispersity of 1.05. The synthesis of poly(1,4-butadiene)-block-polystyrene diblock copolymers (PB-b-PS) and poly(1,4-butadiene) (PB) homopolymer was reported previously.56 It was performed by sequential anionic polymerization of butadiene and styrene in benzene at 60 °C for butadiene (3 h) and 40 °C for styrene (5 h) using sec-BuLi as initiator. The reaction was terminated using isopropanol followed by precipitation in the same solvent. These block copolymers were later hydrogenated (employing a Wilkinson catalyst) in order to obtain PEb-PS linear diblock copolymers. Table 1 lists the copolymers employed

Table 1. Molecular Characteristics of the Polymers and Copolymers Employed sample

% 1,2-units

M̅ w/M̅ n

PEO41 PE25 E79S2141 E53S4751 E26S74105 E11S89244

11.0 12.6 11.3 11.3 11.3

1.05 1.01 1.02 1.04 1.05 1.02

with the notation AxBym, where the subscripts denote the mass fraction in weight percent and the superscript gives the number-averaged molecular weight Mn in kg/mol of the entire block copolymer. The synthesis process for all the diblock copolymers was designed in order to obtain a PE block within the block copolymers with the same molecular weight and approximately the same content of 1,2 units as in the PB precursor (see Table 1). After hydrogenation the PE component contains ethyl branches randomly distributed along the backbone of the chains, hence the PE component can be considered as a model ethylene/α-olefin copolymer. 1518

dx.doi.org/10.1021/ma202327f | Macromolecules 2012, 45, 1517−1528

Macromolecules

Article

Figure 1. SEM micrographs of the surfaces of AAO templates with pores of 20 nm (a, b), 35 nm (c, d), and 60 nm (e, f). Templates Preparation. Ordered AAO templates have been prepared by two-step electrochemical anodization of aluminum as described elsewhere.16,57 First, ultrapure (99.999%) aluminum foils (Advent Research Materials, England) were cleaned and degreased by sonication into solvents with different polarity (acetone, isopropanol, deionized water and ethanol). Foils were then electropolished during 4 min in a solution of perchloric acid/ethanol (1/3) below 10 °C in a constant voltage of 20 V. After that, the first anodization was achieved using sulfuric acid under 19 V and at 1 °C, in order to obtain 20 nm diameter pores; and oxalic acid as electrolyte under 40 V and 1 °C for the 35 and 60 nm diameter pores. Both anodizations were performed during 24 h. Then, the first anodic layer was removed into chromic and phosphoric acid solution. Next, a second anodization was carried out under the same conditions as the first one and templates with 20 nm diameter pores (the one anodized in sulfuric acid) and 35 nm diameter pores (the one anodized in oxalic acid) were obtained. Subsequent processing under phosphoric acid (5 wt % at 30 °C) was performed to widen the pores of some of the templates from 35 to 60 nm in diameter. Pore lengths of the templates used were 200 μm for the templates with 20 nm pores, and 100 μm for the templates with 35 and 60 nm pores, respectively. Infiltration of Polymer Melts. Infiltrating PE-b-PS Copolymers and PE Homopolymer. The polymer was in the molten state at a temperature well above the melting point (Tm) during infiltration. The infiltration of the PE-b-PS diblock copolymers into the alumina templates with 35 and 60 nm pores was carried out by placing the solid materials onto the AAO at 170 °C and then annealed at 180 °C under a nitrogen atmosphere. The high temperature was needed for complete infiltration and the nitrogen atmosphere was employed to prevent degradation. For the infiltration of the copolymer E11S89244, the annealing time was extended to 10 h. In the case of the PE homopolymer the temperature of infiltration was 150 °C. After the infiltration process, the samples were quenched in an ice−water mixture.19 Infiltrating PEO Homopolymer. A PEO macroscopic film was placed onto the AAO template and annealed at 110 °C under N2 atmosphere

for 60 min. The residual PEO located outside of the pores was mechanically removed using a blade in order to induce an independent crystallization of the PEO confined within each nanopore. Scanning Electron Microscopy (SEM) and Transmission Electron Microscopy (TEM). In order to obtain a through-hole membrane to see nanorods by SEM/TEM microscopy, the aluminum substrate and the alumina barrier layer were removed with a mixture of HCl, CuCl2, and H2O, and with 10 wt % H3PO4 at 30 °C, respectively.58 The morphology of the AAO templates was characterized by SEM/TEM (Philips XL-30ESEM and FESEM Hitachi model SU8000 with TE Detector). Differential Scanning Calorimetry (DSC). For the infiltrated and bulk PEO, sample masses of 21.6 mg (the weight includes the alumina template and the polymer, the mass of the polymer was 0.6 mg) and 2 mg were employed, respectively. The crystallization and melting temperatures of bulk and confined PEO were evaluated by DSC (Perkin-Elmer DSC-8500). Two cooling−heating cycles were applied to the sample (10 °C/min). In the case of the infiltrated PE and PE-b-PS AAO templates, sample mass was 25 mg (these samples were of the alumina template plus infiltrated polymer, the exact mass of infiltrated polymer is unknown but it has been estimated by TGA to be of the order of 2% with respect to the total mass, i.e., 0.5 mg approx.). DSC standard runs were performed on a Pyris 1 Perkin-Elmer Instrument calibrated with tin and indium under an ultra high purity nitrogen atmosphere. In order to enhance the sensitivity of the experiment we employed a high cooling and heating rate of 30 °C/min. The results are compared with the PE-b-PS neat samples at the same scan rates and employing a mass of 2 mg. Wide Angle X-ray Diffraction (WAXD). The WAXD measurements were performed at station BL01C at the NSRRC using an imaging plate as the detector. The X-ray beam with the wavelength of 0.954 Å was collimated into the beam size of 0.5 mm ×3 mm by two slits separated by 1.1 mm. With a sample-to-detector distance of 1519

dx.doi.org/10.1021/ma202327f | Macromolecules 2012, 45, 1517−1528

Macromolecules

Article

409.2 mm, the diffraction patterns were collected over the q range of 1 to 25 nm−1.

in view of their large size. They have successfully employed a correlation analysis based on the crystallization temperatures of individual droplets for consecutive cooling runs to differentiate homogeneous from heterogeneous nucleation (see refs 59−63). Carvalho and Dalnoki-Veress62 have recently demonstrated that isolated PEO nanodroplets can experience homogeneous nucleation in 3, 2, or 1 dimensions (i.e., bulk, surface or edge nucleation) depending on the substrate on which they are deposited. The supercooling needed to crystallize the nanodroplets was much larger for the homogeneous bulk nucleation case than for the surface nucleation case. We will follow the terms they have employed in their work. Bulk homogeneous nucleation occurs when the polymer chains aggregate to form homogeneous nuclei that originate within the volume (or bulk) of the micro or nanodomain. This is the process that requires the largest possible supercooling. The presence of a surface during the solidification of the liquid polymer may cause some chain orientation that lowers the free energy for the formation of a homogeneous nuclei; hence, its crystallization can occur at lower supercoolings.62 Homogeneous bulk nucleation occurs at the maximum possible supercooling available to the material or just before vitrification, for the specific microdomain volume under consideration. The glass transition temperature of PEO is approximately −67 °C.64 Table 2 provides a list of low temperature peak crystallization transitions reported for different types of confined PEO samples. The lowest temperature reported so far for the crystallization of PEO is −42 °C and it was observed for PEO nanospheres formed by a PEO block within PS-b-PEOb-PCL linear triblock terpolymers.65 It is important to consider that the crystallization temperature after homogeneous bulk nucleation depends on the volume of the isolated crystallizable phase involved. A correlation between droplet volume and the crystallization temperature of homogeneously nucleated PEO was calculated by Müller et al. from a large number of literature data.8 The dependence of homogeneous nucleation temperature of PEO on volume had been previously reported by Massa and Dalnoki-Veress60 and by Chen et al.66 Such dependence is expected since the probability of nucleation and the nucleation rate depend on sample volume. The correlation found by Müller et al.8 in terms of the volume of the crystallizing phase was



RESULTS AND DISCUSSION AAO Templates. Figure 1 shows SEM micrographs of the well developed structure of the neat alumina templates employed to infiltrate the polymer samples. The AAO templates used had an average diameter of 20, 35, and 60 nm. The pore lengths of the templates used was about 200 μm for the 20 nm template and 100 μm for the other two. Confined Crystallization of PEO within the AOO Templates with 20 nm Pores. Neat PEO41 displays a sharp crystallization exotherm, as shown in Figure 2a, with a

Figure 2. (a) DSC cooling curves from the melt at 10 °C/min for neat PEO41 and the same sample infiltrated within a 20 nm AAO template (PEO41/20 nm). (b) Subsequent heating runs after the cooling shown in part a.

peak Tc value of 37 °C. This is the typical nonisothermal crystallization behavior of heterogeneously nucleated PEO. On the other hand, the PEO41 that was infiltrated within the nanocylinders of 20 nm diameter only crystallizes at a Tc value of −30 °C. This is a consequence of the confinement experienced by the PEO, since it has been subdivided into isolated nanocylinders. Similar results have been found for infiltrated PVDF,26 sPS,28,29 PP4 and linear PE.5,27 It is clear that when the polymer is isolated into nanocylinders the crystallization temperature is much lower than that exhibited by the same polymer in the bulk. The number of active heterogeneities in bulk PEO is of the order of 105 heterogeneities/cm3.8 However, the number of nanocylinders per unit volume in the 20 nm template is several orders of magnitude larger (approximately 10 orders of magnitude larger, i.e., 1015 nanocylinders/cm3). Therefore, statistically speaking, the PEO nanocylinders within the AAO template are virtually heterogeneity free and must undergo either homogeneous nucleation or a heterogeneous surface nucleation on the AAO inner cylinder walls. However, homogeneous nucleation can only occur at larger supercoolings when compared with a typical heterogeneous nucleation, hence the low temperature crystallization observed. In our case, it is difficult to ascertain whether the PEO nucleation inside the AAO template is homogeneous or heterogeneous.8 One way to distinguish between heterogeneous and homogeneous nucleation is to determine the scaling of the time constant, associated with the nucleation event, (i.e., the particle diameter in the case of droplets) a technique that was successfully employed by Massa et al.59 In their case, the nucleation of dewetted droplets was directly observed in an optical microscope

Tc = − 41.8 + 2.89 log(vd)

(1)

where Tc is the peak crystallization temperature during a DSC cooling scan in °C and vd the volume of the PEO phase in nm. The prepared templates for PEO infiltration had a height of 200 μm. Hence the volume of one PEO cylinder (if one considers that the infiltration was 100% efficient and the entire nanopore was filled with PEO) is of the order of 6.3 × 107 nm3. Equation 1 would then predict an experimental peak crystallization temperature during a cooling DSC run of −19.3 °C if the nucleation is homogeneous. The experimental Tc value obtained in Figure 2 for the confined PEO cylinders within the AAO template was −30 °C. Taking into consideration the experimental uncertainty in eq 1 (which can easily amount to 10 °C when experimental data dispersion is considered; see ref 8) then the nucleation process for PEO cylinders inside the AAO template is most probably homogeneous bulk in nature and not originated on the surface of the AAO nanopores. In order to reveal the orientation of the PEO crystals within the AAO cylindrical nanopores, the samples were examined by 2-D wide-angle X-ray diffraction (WAXD). Three patterns were 1520

dx.doi.org/10.1021/ma202327f | Macromolecules 2012, 45, 1517−1528

Macromolecules

Article

Table 2. Low Temperature Crystallization Temperatures Reported for Confined PEO Micro- or Nanodomains material 11

EO55B45 /PB EO55B4511/PB B24I56EO2067 B11I70EO19120 B17I57EO26130 B19I39EO42135 E24I57EO1964 E11I71EO18123 E18I57EO25133 E19I40EO41138 E24EP57EO1969 E11EP71EO18123 E18EP57EO25133 E19EP40EO41138 S81EO1918.5 S63EO16C2124 EO17-b-B8325.4 PEO droplets PEO droplets PEO droplets EO60B4012.5/PB EO60B4012.5/PB HBd85EO154,22 HBd76EO246,44 S74EO26133.3 B16S68EO16210 B29S40EO31168 B16S40EO47143 B19S35EO46217 E17S67EO16211 E29S40EO31170 E38S16EO4677 E16S40EO44144 E19S34EO47143 E19S35EO46219 B81EO1934 E82EO1835 B89EO11102 E89EO11105 PEO/PHB PEO droplets a

PEO content (wt %)

Tc peaka (°C)

crystallization conditions (°C/min)

morphology

particle diameter (nm)

reference

13 17 24 19 26 42 19 18 25 41 19 18 18 41 19 16 17 100 100 100 26 43 15 24 26 16 31 44 46 16 31 46 44 47 46 19 18 11 11 20 100

−33 −33 −23.9 −22.2 −21.0/16.1 −25.0/19.8/37.5 −26.4 −25.4 −21.1/26.8 −24.0/23.6/37.6 −27 −25.4 −21.1/26.8 −24.0/23.6/37.6 −40 −42 −25 −23.3 −23.9 −20.5 −35 −29 −31.4 −26.4/-7.3 −27 −33.8 −27.1/37.9 −24.4/39.9 −26.4/37.2 −30.9 −26.0/42.0 −27.4/39.1 −18.2/46.2 −29.1/42.4 −25.2/40.6 −21.1/-8.2/46.1 −34.3/20.2/45.7 −26.8/-11.5 −24.3/0.5 −27/-13 −4.5/6/12

5 5 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 5 5 5 5 5 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 1

spheres spheres --------------spheres droplets droplets droplets spheres cylinders --cylinders spheres cylinders lamellar lamellar spheres cylinders lamellar ---spheres spheres spheres spheres -droplets

-----------------83 83 120 8.8 8.4 --13.3 ----------13 ± 1.3 21 ± 1.0 19 ± 1.4 31 ± 1 -different sizes

66, 67 66, 67 68−70 69, 70 69−71 70 70 70 70 70 69−71 69, 70 70 70 65,69 65 72 73 73 73 74 74 75 75 33 76 76 76 76 76 76 76 76 76 76 77 77 77 77 78 62

In some cases multiple crystallization peaks were found. All the temperatures are reported.

collected for the sample by rotating the specimen to allow for the incident X-ray to pass through the three principal axes of the specimen. That is, the PEO infiltrated template was collected by having the incident X-ray traveling along the x̂, ŷ, and ẑ(where these three directions are orthogonal to each other), respectively, as schematically shown in Figure 3. The x̂−ŷ plane is corresponding to film surface and the ẑ direction is parallel to the film normal. (i.e., cylinder axis of AAO channels). Figure 4 shows the 2D WAXD patterns of the PEO infiltrated AAO template with 20 nm cylindrical pores. When the X-ray is traveling along the x̂ and ŷ, similar patterns can be observed, as shown in Figure 4a and Figure 4b, respectively. The first diffraction arcs (q = 13.7 nm−1) were found to locate on both the meridian and equator, which was associated with (120) diffraction of PEO crystals. On the other hand, the second quadrant (q = 16.1 nm−1) was attributed to the overlap of (13̅ 2), (112), (032), (12̅ 4), (204)̅ , (21̅ 2) and (004) diffractions.

Figure 3. Schematic illustrations of the geometry employed to collect the WAXD data on the PEO infiltrated template indicating different incident directions of the X-ray beam.

In the azimuthal scans (obtained by scanning the intensity of a specific diffraction around the azimuthal angle (ψ) starting from 1521

dx.doi.org/10.1021/ma202327f | Macromolecules 2012, 45, 1517−1528

Macromolecules

Article

Figure 4. 2D WAXD patterns of PEO infiltrated AAO templates with 20 nm diameter cylindrical pores, collected by having the incident X-ray beam traveling along (a) x̂, (b) ŷ, and (c) ẑ.

Table 3. Domain Size “l” and Spacing “d” of the PE-b-PS Diblock Copolymers As Determined by TEM and SAXS Measurements at Room Temperature,82 with the Degree of Crystallinity Obtained by DSC also Listed

the vertical direction of the pattern) of the first and second diffractions, the maximum intensity of the first diffraction was found to locate at ψ = 2, 91, 182, and 273°, whereas that of second diffraction was at ψ = 35, 64, 115, 148, 213, 248, 295, and 327°, which was consistent with the reported [120] uniaxial pattern of PEO crystal.35 Moreover, the intensity of the (120) diffraction on the meridian was stronger than that on the equator, which can also be elucidated by the PEO [120] uniaxial pattern.79 On the basis of the fact that [120] uniaxial pattern appeared in both of the WAXD patterns when the X-ray is traveling along x̂ (Figure 4a) and ŷ (Figure 4b), and the ring-like pattern was obtained by having the X-ray beam traveling along ẑ (see Figure 4c), we conclude that the crystalline stems of the PEO crystals were perpendicular to the cylindrical interface of the AAO pores but randomly oriented with respect to the x̂ − ŷ plane. Such an orientation of PEO crystals was in accord with the previous results while the PEO was the minority constituent in the cylinder-forming diblock copolymer.33,35 Confined Crystallization of PE within PE-b-PS Diblock Copolymers. The materials employed for infiltration have been the subject of previous studies (refs 56, 80−82) Table 3 shows a summary of the characterization results obtained previously by SAXS and TEM. These PE-b-PS diblock copolymers are strongly segregated in the melt and the crystallization of the PE block occurs within the phase segregated self-assembled micro or nanodomains indicated in Table 3. These phase segregated structures will be referred to as microdomains (MDs). Figure 5 presents the DSC cooling scans from the melt and subsequent heating runs for the neat PE and PE-b-PS diblock copolymer samples at a scanning rate of 30 °C/min. The results are consistent in general terms with those reported previously at 10 °C/min (see ref56). It is important to note that the molecular weight (Mn ≈ 26 kg/mol) and the 1,2 units content (% 1,2-units ≈ 11%) of the PE block within the diblock copolymers as well as in the neat PE were kept constant in order to eliminate the influence of these parameters on the thermal and morphological behavior of the diblock copolymers. The crystallization of PE25 displays a behavior similar to ethylene/α-olefin copolymers.83,84 There is a sharp main crystallization process (labeled “α”), where the majority of the PE25 crystallizes (at Tc = 85.1 °C, see Table 4) and a much smaller exotherm at 52.1 °C. Similar small exotherms have been commonly reported in a temperature range of 40−60 °C for branched polyethylenes in general, and for linear low density polyethylene in particular. They are caused by the crystallization of short methylene sequence lengths belonging to

TEM

SAXS

sample

l (nm)

d (nm)

PE25 E79S2141

-

-

E53S4751 E26S74105 E11S89244

16a ± 3 21b ± 5 12b ± 2

62 ± 3 126c ± 29 105c ± 13

d (nm) 65 99 108

morphology bulk PS cylinders in a PE matrix lamellar PE cylinders PE spheres

Xc (%)d

Xc (%)e

35 26

27 23

23 18 15

20 15 11

a

Thickness of the PE lamellae. bRadius of the microdomain. Interdomain distance. dObtained by DSC at 10 °C/min, data from ref 82 eObtained by DSC at 30 °C/min in this work.

c

chains whose longer methylene sequences crystallize at higher temperatures (in the exotherm α). It is well-known that such small exotherms exhibited by branched polyethylenes at 40− 60 °C are due to intramolecular fractionation during crystallization produced by the distribution of branches along the chain.84 Therefore, this exotherm has been labeled “IF” to indicate its origin. In Table 4, the dominant crystallization peak temperatures for each sample have been reported in bold. Exotherm α in Figure 5 is related to the crystallization of PE chains starting from heterogeneous nuclei. In the diblock copolymers case, the Tc values for the α exotherm are shifted to lower values because of the topological restrictions imposed by the covalently bonded glassy PS phase. The observation of the α exotherm is still possible because of the percolation of the crystallizing phase, usually present through defects in the microdomain morphologies (e.g., lamellae and cylinders). In the cases of E26S74105 (PE cylinders within a glassy PS matrix) and E11S89244 (PE spheres in a PS matrix) diblock copolymers, multiple crystallization exotherms are observed (more easily appreciated in the close up shown in Figure 5b). This is caused by a fractionated crystallization phenomenon that is produced as a consequence of confinement of the crystallizable phase into isolated MDs and by the existence of different types of nucleation events within such MDs8,69,85 Before each crystallization exotherm a specific nucleation event must occur (e.g., heterogeneous or homogeneous). The heterogeneity type present in different MDs may also vary. The fractionated behavior of the crystallization process is due to the much larger number of MDs as compared to the number of heterogeneities available to produce nucleation. According to 1522

dx.doi.org/10.1021/ma202327f | Macromolecules 2012, 45, 1517−1528

Macromolecules

Article

Fractionated crystallization will be present if there are MDs with different kinds of heterogeneities inside each population. In the present case, there could be a fraction of MDs with highly active heterogeneities (those usually present in bulk PE) that can crystallize at low supercoolings, like in the Tc range of the α exotherm. Then, a second type of heterogeneity may be present in a different fraction of MDs, that can crystallize at lower temperatures in exotherm β (for instance that produced by the remains of the Wilkinson catalyst employed for hydrogenation purposes). Finally, a clean fraction of MDs, that does not contain any heterogeneity at all, must undergo homogeneus nucleation (either surface or homogeneous bulk nucleation). The higher the heterogeneity nucleating efficiency, the lower the supercooling needed by that specific fraction of MDs to crystallize. However, the interconnection (or percolation) level in the MDs could allow the spread of secondary nucleation through out the crystallizable component.69,86 In Figure 5, the IF exotherm characteristic of branched polyethylenes may well be overlapping with the β exotherm (that originates on the crystallization provoked by the nucleation of a less efficient type of heterogeneity), hence we have labeled some exotherms as IF + β Figure 5. The exotherms that most probably represent the crystallization from clean or heterogeneity free MDs have been labeled γ in Figure 5. The only block copolymer where MDs percolation is highly unlikely is E11S89244, since it forms 24 nm PE spheres within a PS matrix. In fact, the number of clean droplets dominate the crystallization behavior because the most prominent crystallization event is that of exotherm γ at 46.9 °C, the lowest peak crystallization temperature for all the diblock copolymers examined. The corresponding nucleation process for this lowest temperature exotherm could be either homogeneous bulk nucleation or surface nucleation. Table 5 presents the low temperature crystallization of PE droplets or PE MDs within block copolymers reported in the literature. The value of 46.9 °C is one of the lowest ever reported for the crystallization of confined PE phases, although even lower temperatures can be achieved by infiltration within AAO templates as will presented below. We exclude the possibility that fractionated crystallization could be a result of a wide range of microdomain sizes (like in polyblends) since our samples were prepared by anionic polimerization and a narrow distribution of microdomains sizes was obtained. In the case of PE spheres, TEM measurements (whose experimental error due to staining could be specially large in semicrystalline samples) yielded dv/dn values (average volume diameter/average number diameter) lower than 1.5 (see ref 56). Figure 5 also shows the DSC heating scans. The neat PE25 sample exhibits a peak melting temperature (Tm) of only 98.7 °C because of the influence of its short chain branching

Figure 5. (a) DSC cooling curves from the melt at 30 °C/min. (b) Zoom of the E26S74105 and E11S89244and (c) subsequent heating runs after the cooling shown in part a. Arrows indicate the position of the Tg corresponding to the PS block.

Figure 6. TEM images of nanocylinders obtained by AAO template removal. (a) E53S4751 copolymer and (b) PE homopolymer.

an estimation made by measuring TEM images, there are ∼1012 MDs/cm3 and ∼1015 MDs/cm3 for E26S74105 and E11S89244 respectively. In contrast, the heterogeneities concentration usually present in an equivalent bulk sample of polyethylene is only ∼109/cm3.53

Table 4. Peak Crystallization and Melting Temperatures Obtained from Figures 5, 7 and 8 uninfiltrated

60 nm template

Tc sample 25

E100 E79S2141 E53S4751 E26S74105 E11S89244 a

α 85.1 73.9 73.8 72.4 66.9

IF 52.1 51.9 53.3 ---

35 nm template

Tc β + IF ---56.9 56.9

γ ---49.9 46.9

Tm 97.5/87.5 92.8 92.3 91.8 85.8

a

Tc

α

β

γ

Tm

γ

Tm

84.0 77.0 ----

60.0 57.0 61.0 ---

51.5 48.5 54.5 44.5 --

96.4 90.4 91.9 86.8 --

45.2 48.2 NA ---

88.4 90.1 NA ---

Shoulder. NA = not available (the sample was not prepared). 1523

dx.doi.org/10.1021/ma202327f | Macromolecules 2012, 45, 1517−1528

Macromolecules

Article

Table 5. Low Temperature Crystallization Temperatures Reported for Confined PE Droplets or PE MDs Tc peaka (°C)

crystallization conditions

-------------

85−87 79−81 80−82 82−84 53.5/61.5/70/100 74.8

slow cooling slow cooling slow cooling slow cooling 2 °C/min 1 °C/min

droplets droplets droplets droplets droplets droplets