Influence of Hyperbranched Polyesters on the Surface Tension of

The influence of hyperbranched polyesters with different functional end groups on the surface tension of mixtures with an oligo(ester diol) was invest...
0 downloads 0 Views 100KB Size
8096

Langmuir 2004, 20, 8096-8102

Influence of Hyperbranched Polyesters on the Surface Tension of Polyols Antje Ziemer,† Mazen Azizi,‡ Dieter Pleul,† Frank Simon,† Stefan Michel,† Mirko Kreitschmann,§ Paul Kierkus,§ Brigitte Voit,† and Karina Grundke*,† Leibniz Institute of Polymer Research Dresden, Hohe Strasse 6, 01069 Dresden, Germany, Aleppo University, Aleppo, Syria, and BASF Schwarzheide GmbH, Schwarzheide, Germany Received January 8, 2004. In Final Form: July 2, 2004 The influence of hyperbranched polyesters with different functional end groups on the surface tension of mixtures with an oligo(ester diol) was investigated. The temperature dependence of the surface tension of the pure components and of the mixtures was measured by a modified Wilhelmy balance technique. The results indicate that the surface tension of the pure hyperbranched polyesters strongly depends on the functionality of the end groups. The functionalization of the hydroxyl end groups by short alkyl chains (methyl, tert-butyl) reduced the surface tension depending on the degree of substitution. The surface tension of the mixtures with the hydroxyl-terminated hyperbranched polyester was slightly increased at higher concentrations of the hyperbranched polymer compared to the surface tension of the pure ester diol. On the other hand, the surface tension of mixtures could be considerably decreased using 1% of hyperbranched polyester polyols partially substituted with short alkyl chains. In that case, the modified hyperbranched polyesters act as surface active agents. On the molecular level, the enrichment of the modified hyperbranched polyester in the surface region was proven by X-ray photoelectron spectroscopy measurements.

1. Introduction Flory introduced hyperbranched polymers (HBPs) from a theoretical point of view in 19521 when he described the intermolecular condensation of ABx-monomers. At that time, HBPs seemed to be of low interest regarding bulk material because of their noncrystallinity, lack of entanglement, and, therefore, limited mechanical strength. Today, the focus of research is concentrated more and more on special polymer architectures and polymer specialties. Worldwide many research groups work on dendritic macromolecules. In 1988, Kim and Webster2 reported hyperbranched polyphenylenes as melt viscosity modifiers for polystyrene which renewed the interest in hyperbranched materials. Since that time a large number of HBPs, for example, polyester, polyamides, polyurethanes, and poly(ester amides), and various applications have been reported.3-9 One of the big advantages of HBPs compared to dendrimers is their one-pot synthesis. Therefore, relatively large amounts of materials can be synthesized with a reasonable effort. Compared to den* To whom correspondence should be addressed: Dr. Karina Grundke. Phone: ++49 351 4658-475. Fax: ++49 351 4658-284. E-mail: [email protected]. † Leibniz Institute of Polymer Research Dresden. ‡ Aleppo University. § BASF Schwarzheide GmbH. (1) Flory, P. J. J. Am. Chem. Soc. 1952, 74, 2718-2723. (2) Kim, Y. H.; Webster, O. W. J. Am. Chem. Soc. 1990, 112, 45924593. (3) Hawker, C. J.; Lee, R.; Fre´chet, J. M. C. J. Am. Chem. Soc. 1991, 113, 4583-4588. (4) Malmstro¨m, E.; Johansson, M.; Hult, A. Macromol. Chem. Phys. 1996, 197, 3199-3207. (5) Schmaljohann, D.; Ha¨ussler, L.; Po¨tschke, P.; Voit, B. I.; Loontjens, T. J. A. Macromol. Chem. Phys. 2000, 201, 49-57. (6) Hong, Y.; Cooper-White, J. J.; Mackay, M. E.; Hawker, C. J.; Malmstro¨m, E.; Rehnberg, N. J. Rheol. 1999, 43, 781-793. (7) Schmaljohann, D.; Po¨tschke, P.; Ha¨ssler, R.; Voit, B. I.; Froehling, P. E.; Mosert, B.; Loontjens, T. J. A. Macromolecules 1999, 32, 63336339. (8) Mulkern, T. J.; Beck Tan, N. C. Polymer 2000, 41, 3193-3203. (9) Dekker, G. H. Paintindia 1999, 49, 117.

drimers, HBPs do not have the perfect structure but they have a high functionality, are soluble in many organic solvents, and because of the globular shape have a lower solution viscosity than linear analogues of the same molar mass. HBPs are built up mostly from AB2 monomers via reaction between the A and B functionalities. The formation of three different structural units (dendritic, terminal, and linear) is possible. In the statistical case, the degree of branching (DB) is 50% as calculated, for example, according to Fre´chet et al.3 The end-group modification of hyperbranched polyesters is also well-known, demonstrating the strong influence of the nature of a large number of end groups on the materials properties.4,5 One of the major applications of HBPs is their use in polymer blends or as additives in linear polymers. Because of the special branched structure, they can act as a rheology modifier. Especially for hyperbranched polyesters, several blend systems have been extensively studied, for example, HBP in combination with polyolefins,6,7 polystyrene,8 alkyl resins,9 or poly(2-hydroxyethyl)methacrylate.10 Different mechanisms are discussed in the literature to explain the reduction of viscosity in mixtures of linear polymers and branched macromolecules, such as HBPs. It is assumed that HBPs disrupt the local entanglement network and reduce the entanglement density of the high-molecularweight, linear polymers as a result of their relatively compact, globular structure.11 It is also assumed that HBPs have the tendency to migrate to the surface to form a lubricating layer leading to a reduction of the viscosity by either cohesive or adhesive failure (slip) at the surface or both.12 In this work, we were interested in the surface properties of hyperbranched polyesters. The surface tension plays a (10) Nunez, C. N.; Chiou, B.-S.; Andrady, A. L.; Khan, S. A. Macromolecules 2000, 33, 1720-1726. (11) Farrington, P. J.; Hawker, C. J.; Frechet, J. M. J.; Mackay, M. E. Macromolecules 1998, 31, 5043-5050. (12) Chan, C. M.; Feng, J. J. Rheol. 1997, 41, 319-333.

10.1021/la049930c CCC: $27.50 © 2004 American Chemical Society Published on Web 08/14/2004

Surface Tension of Polyols

key role in polymer processing and blending.13 Recently, Orlicki et al.14 investigated the role of molecular architecture and end-group functionality on the surface tension of HBPs. By varying the functionality of the end groups, the surface properties of hyperbranched polyetherimides could be tuned over a wide range of solid surface tensions. The surface tension was determined from contact angle measurements on smooth films of the HBPs. Increasing the length of alkane end groups (C, C8, C18) resulted in a decreased solid surface tension (46, 37, 27 mJ/m2). High surface energy materials were obtained with hydroxyterminated hyperbranched polyetherimides. Mackay et al.15 studied the surface properties of hyperbranched polyesters based on 2,2-(bishydroxymethyl)propionic acid with an ethoxylated pentaerytriol core which was modified with alkyl chains. They found that the melt surface tension of hydroxyl-terminated HBPs was high and approached that of water. The surface tension was decreased depending on the degree of substituting a long-chain alkane (C20/22) to the end groups. In our previous work,16 we could show that the solid surface tension of an aromatic hyperbranched polyester terminated with acetoxy end groups (-OCOCH3) is distinctly lower (36 mJ/m2) compared to the same hyperbranched polyester terminated with hydroxyl end groups (46 mJ/m2). Here we report results of the effect of hyperbranched polyesters on the melt surface tension of an oligo(ester diol). This study is aimed at the potential application of HBPs as surface tension modifiers or compatibilizers in polymer blends. The HBPs used were aromatic-aliphatic hyperbranched polyesters based on 4,4-bis(4-hydroxyphenyl)valeric acid. The phenol end groups of the hyperbranched polyester were modified leading to short alkyl chains as end groups (methyl and tert-butyl groups) and different degrees of substitution to investigate the surface tension response to this molecular modification. The results of the surface tension measurements were related to the molecular surface composition of the pure liquid components and of the mixture determined by X-ray photoelectron spectroscopy (XPS). 2. Experimental Section 2.1. Materials and Reaction Procedure. 4,4-Bis(4-hydroxyphenyl)valeric acid (95%) was obtained from Lancaster. Dibutyltin-diacetate was purchased from Fluka, acetyl chloride and pivaline acid chloride (trimethylacetyl chloride; >98%) was from Merck. All solvents were used without any further purification. The oligo(ester diol) based on phthalic acid and mono- and diethylene glycol in a ratio of 1:10 and with a molar mass M h n of 470 g/mol and an OH value of 240 mg KOH/g was obtained from BASF Schwarzheide GmbH. Polymerization (P-OH)5. The monomer 4,4-bis(4-hydroxyphenyl)valeric acid was weighed in a three-necked flask equipped with a stirrer and gas-inlet and -outlet tubes and heated to the reaction temperature of 185 °C. The liquid catalyst was added at the reaction temperature. The temperature was maintained during initial constant nitrogen flow, as well as the subsequent vacuum (about 3 × 10-2 mbar), which was applied in the second reaction period. The product was dissolved in tetrahydrofuran. The polymer was precipitated twice, first into cold water and then in n-hexane, and then it was dried at 40 °C in a vacuum. Yield: 81% (P1-OH). (13) Sammler, R. L.; Dion, R. P.; Carriere, C. J.; Cohen, A. Rheol. Acta 1992, 31, 554-564. (14) Orlicki, J. A.; Viernes, N. O. L.; Moore, J. S. Langmuir 2002, 18, 9990-9995. (15) Mackay, M. E.; Carmezini, G.; Sauer, B. B.; Kampert, W. Langmuir 2001, 17, 1708-1712. (16) Beyerlein, D.; Belge, G.; Eichhorn, K.-J.; Gauglitz, G.; Grundke, K.; Voit, B. I. Macromol. Symp. 2001, 164, 117-132.

Langmuir, Vol. 20, No. 19, 2004 8097 Two samples were prepared: P1-OH, M h n ) 1900 g/mol, M hw) 5100 g/mol, Tg ) 75 °C; P2-OH, M h n ) 900 g/mol, M h w ) 1800 g/mol, Tg ) 58 °C. 1H NMR (DMSO-d6): δ 1.46, 1.52, 1.57, 1.62 (CH3); 1.94 (CH2-COOH); 2.21, 2.28, 2.35, 2.41 (CH2); 6.66, 6.97, 7.18 (CHAr), 9.14, 9.17, 9.20, 9.23, 9.41 (OH), 11.97 (COOH). 13C NMR (DMSO-d6): δ 27.33, 27.12, 26.95, 27.03, 27.22, 27.44 (CH3), 29.91, 29.98, 29.86, 30.07, 30.10, 30.02, 35.78, 35.96, 36.50, 36.29, 36.11, 36.19 (CH2), 43.87, 44.36, 44.85, 43.99, 44.46, 44.94 (Cq), 139.44, 127.8-128.03, 114.78, 114.84, 115.02, 138.44, 139.20, 138.17, 155.09, 155.31, 155.18, 155.41 (CAr), 171.99, 171.89, 171.85, 171.74, 171.71, 172.04, 174.46, 174.63, 174.80 (COOH, COOR). IR cm-1 υ: 3356 (OH); 3060 (CHaromat); 2965, 2875 (CH2); 1724 (CO); 1509; 1205; 1171; 832. Modification of the Polymer (P-C1-51, P-C4-15). For the modified hyperbranched polyesters, the HBP P2-OH was used with a molar mass of M h n ) 900 g/mol and M h w ) 1800 g/mol. At first, 51% of the OH groups of the hyperbranched polyester were functionalized with acetyl chloride leading to methyl groups as end groups (P-C1-51). A second hyperbranched polyester was modified with tert-butyl chain ends using pivaline acid chloride. In this case, the degree of modification was 15% (P-C4-15). All reactions were carried out in a nitrogen atmosphere to avoid hydrolysis of the alkyl acid chloride. A total of 5 g of the polymer P2-OH was dissolved in pyridine. Then the alkyl acid chloride (acetyl chloride or pivaline acid chloride) (ratio: OH groups/alkyl acid chloride ) 1:1) was added dropwise at 0 °C. The temperature was maintained for 1 h. Afterward the mixture was stirred for 24 h at room temperature. The solution was precipitated into cold water, neutralized with HCl, and filtered. The precipitate was washed with aqueous Na2CO3 and dried at 40 °C in a vacuum. P-C1-51: yield 45%, M h n ) 1300 g/mol, M h w ) 2200 g/mol, Tg ) 60 °C. 1H NMR (DMSO-d6/CDCl3): δ 1.49, 1.51, 1.57 (CH3); 1.80 (CH2-COOH); 2.36 (CH2); 3.33 (CH2COOCH3); 6.64, 6.94, 7.12 (CHAr). P-C4-15: yield 55%, M h n ) 6000 g/mol, M h w ) 17 700 g/mol, Tg ) 54 °C. 1H NMR (DMSO-d6/CDCl3): δ 1.11 (tBu); 1.27 (CH3); 1.56, 1.62 (CH2-COOH); 2.28, 2.41 (CH2); 3.27 (CH2COOCH3); 6.67, 6.70, 7.17 (CHAr). The degree of modification cannot be controlled exactly in the synthesis by stoichiometry of the reagents. Blend Preparation. Blends were prepared by simply mixing the hyperbranched polyester with the oligo(ester diol) at room temperature without any use of a solvent. The hyperbranched polyester P-OH is well soluble in the liquid ester diol. When modified HBPs were used, the miscibility was reduced but no indication of phase separation could be observed when 1 wt % HBP was added to the oligo(ester diol). 2.2. Methods. Size Exclusion Chromatography (SEC). The SEC measurements were performed with a modular chromatographic system equipped with a refractive index detector. A combination of Zorbax PSM 60 and 300 columns (Rockland Technologies) was used. The measurements were carried out at room temperature with dimethylacetamide containing 2 vol % water and 3 g/L of LiCl as the mobile phase. The molecular weights were calculated by the use of poly(2-vinylpyridine) standards (Polymer Standards Services). Differential Scanning Calorimetry (DSC). The DSC measurements were performed on a Perkin-Elmer DSC 7, equipped with the software Pyris version 3.51, temperature program -60 to +120 °C, rate of heating 20 K/min, rate of cooling 80 K/min. The glass transition temperatures of the parent hyperbranched polyester P2-OH (Tg ) 58 °C) were nearly unchanged or decreased slightly after modification (P-C1-51, Tg ) 60 °C; P-C4-15, Tg ) 54 °C). Only one glass transition temperature was obtained for the mixture of the OH-terminated HBP and the oligo(ester diol) instead of two for the two single components. This indicates that the mixture is homogeneous and that no phase separation takes place.17 In the case of the modified HBPs, DSC also provided only one glass transition temperature. However, in this case the concentration of about 1% of HBP in the mixture was so small that DSC cannot provide relevant information about the miscibility. (17) Schwarz, A. J. Polym. Sci., Polym. Phys. Ed. 1974, 12, 11951205.

8098

Langmuir, Vol. 20, No. 19, 2004

Ziemer et al. per unit length becomes constant. At 70 and 100 °C, these effects disappeared because of the low viscosity at higher temperatures. The procedure of partially emersing the fiber after immersion is necessary to ensure complete wetting of the fiber by the polymer. This can be seen at lower temperatures where the force per unit length is lower for the immersion mode because of incomplete wetting. The surface tension can be calculated from

F g∆m ) ) γ cos θ p p

Figure 1. Typical plots obtained from a Wilhelmy experiment using the modified hyperbranched polyester (15% C4). Graph A presents the force per unit length versus time at four different temperatures (25, 50, 70, 100 °C). For better graphical clarity, not all data points are shown. Graph B shows the corresponding immersion depths (id’s) during the experiment: (a) approach of the fiber and contact with the liquid surface (id ) 0 mm); (b) immersion of the fiber up to id ) 0.4 mm; (c) constant id for 2 min; (d) emersion of the fiber to id ) 0.2 mm; (e) constant emersion for 5 min (id ) 0.2 mm); (f) complete withdraw until the liquid detached the fiber (id < 0 mm). Surface Tension Measurements. The surface tension of the pure components and the polymer mixtures were determined by the Wilhelmy balance technique using a modified version with thin fibers.18-20 Details of our apparatus were described previously.19,20 It consists of a highly sensitive Sartorius microbalance (sensitivity 1 µg) which is connected to a temperature cell through a glass tube. The temperature can be controlled with an accuracy of (0.5 °C. For each experiment, a thin, carefully cleaned platinum wire consisting of a platinum-iridium alloy (diameter, 52 µm) is attached to the microbalance. Then, approximately 50 mg of the polymeric liquid is filled into a small stainless steel beaker and placed into the temperature cell at room temperature. After introduction of argon gas and heating to the measuring temperature, the polymeric liquid is brought into contact with the fiber by moving a motor driven table with a velocity of 15 µm/s. Typical plots of the Wilhelmy experiments are shown in Figure 1. The fiber is immersed into the polymeric liquid up to an immersion depth of 0.4 mm with a velocity of 2.5 µm/s and held stationary for about 2 min before it is withdrawn to an emersion depth of 0.2 mm and again held stationary for about 5 min. During this procedure, the change in mass ∆m of the liquid lifted above the level of the polymer surface due to capillarity is recorded by the microbalance as force per unit length (cf. Figure 1A). As a result of the forced fiber/fluid motion (sections b, d, and f in Figure 1B), nonnegligible force contributions result from the shear stress exerted by the viscous flow of the liquid on the fiber. Figure 1A shows that these viscosity effects are strongly dependent on the temperature. At 25 °C, additional changes of the force per unit length were observed during immersion and emersion steps. To exclude these hydrodynamic effects in the measurements, the fiber was held stationary at a constant emersion depth until viscous relaxation occurred and the force (18) Sauer, B. B.; Dipaolo, N. V. J. Colloid Interface Sci. 1991, 144, 527-537. (19) Grundke, K.; Uhlmann, P.; Gietzelt, T.; Redlich, B.; Jacobasch, H.-J. Colloids Surf., A 1996, 116, 93-104. (20) Grundke, K.; Michel, S.; Osterhold, M. Prog. Org. Coat. 2000, 39, 101-106.

(1)

where γ is the surface tension, θ is the contact angle, F is the change in the force registered by the balance, p is the perimeter of the fiber, and g is the acceleration due to gravity. Because the contact angle θ is 0 by design of the procedure, the surface tension γ of the polymeric liquid can be determined directly from the force acting on the fiber. Because the buoyancy force can be neglected in measurements with thin fibers (290 eV appears from the π f π* electron transition in the conjugated π-electron system of the phenyl ring. The same component peaks A, C, and D can also be found in the pure modified hyperbranched polyester, indicating that the oligo(ester diol) and the HBP have the same structural units as shown schematically in Figures 2 and 3. The additional component peak B appears from the carbon atom in the β position of the ester carbon [-OC(O)-C-; see Figures 2 and 6b). Because in the modified hyperbranched polyester 15% of all OH end groups were esterified with pivaline acid chloride, the calculated ratio of the component peaks should be [B]/[C]/[D]|calc ) 1:3.16: 1. In the recorded C(1s) spectrum, the ratio is [B]/[C]/ [D]|meas ) 1:3.15:1. Hence, the measured ratio excellently agrees with the calculated ratio. Obviously, the chemical composition of the surface region of the HBP is identical with the bulk composition, showing the expected restrictions in the chain mobility of a hyperbranched structure. The measured C(1s) spectrum of the mixture (Figure 6c) agrees exactly with a calculated C(1s) spectrum of the oligo(ester diol) and the modified hyperbranched polyester P-C4-15 where approximately 43% of all carbon originates from the modified hyperbranched polyester component, while only about 57% of all carbon appears from the oligo(ester diol) component. Hence, the XPS results show that

Figure 6. High-resolution C(1s) XPS spectra of the oligo(ester diol) (a), the modified HBP P-C4-15 (b), and the mixture of the oligo(ester diol) (ode) with 1 wt % P-C4-15 (p-c, c). The specification of the component peaks corresponds with the chemical structures of the components shown in Figures 2 and 3: A, carbon atoms of the phenyl ring; B, carbon atom in the β position of the ester carbon [-O-C(O)-C-]; C, ether and hydroxyl end groups; D, ester groups.

the hyperbranched polyester component whose bulk concentration equals 1 wt % is clearly enriched in the surface region. This result agrees with the conclusion drawn from the surface tension measurements where the surface tension of the mixture nearly equals the surface tension of the modified hyperbranched polyester. 4. Conclusions It was demonstrated that HBPs act as surface modifiers so that the surface properties of other polymers can be tuned by the addition of small amounts of HBPs. The surface tension of hyperbranched polyesters used in this study and their blends with an oligo(ester diol) could be controlled by the molecular composition of the polyesters. Blends containing 10 and 20 wt % of a hydroxyterminated hyperbranched polyester P-OH had a higher surface tension compared to the pure oligo(ester diol). The addition of 1 wt % of a hyperbranched polyester modified

8102

Langmuir, Vol. 20, No. 19, 2004

with short alkyl chains decreased the surface tension of the oligo(ester diol) up to 30%. In this case, the surface tension of the blend with 1 wt % of the modified hyperbranched polyesters equals the surface tension of the pure modified hyperbranched polyesters, indicating that the HBPs are surface active substances which are located in the surface region of the blend. The surface

Ziemer et al.

enrichment of the HBP was directly proven by highresolution XPS measurements. Acknowledgment. The authors would like to thank BASF Schwarzheide GmbH for financial support. LA049930C