Tuning solution polymer properties by binary water–ethanol solvent mixtures

Abstract

The solubility of polymers can be significantly altered by the use of solvent mixtures. The solvent composition also effects the self-assembly properties of amphiphilic copolymers. In addition, water–ethanol mixtures are known to exhibit abnormal physicochemical properties due to the presence of hydration shells around the ethanol molecules, while at the same time both solvents have very low toxicity. However, the solution properties of amphiphilic copolymers in water–ethanol mixtures have been scarcely studied. Here we show that the solution polymer properties of amphiphilic copoly(2-oxazoline)s can be significantly altered in binary water–ethanol mixtures resulting in increased solubility, tuneable lower critical solution temperatures as well as polymer–solvent combinations with both a LCST followed by an UCST and improved dispersion stability. Surprisingly, it was found that polymers insoluble in both ethanol and water could be dissolved in water–ethanol mixtures, opening the way to novel formulations for drug delivery or personal care applications. Our results represent a straightforward method for tuning solution polymer properties without the synthetic efforts that are generally required to change the copolymer composition and properties.

---

Introduction

The solution properties of amphiphilic copolymers, such as solubility and aggregation, are of major importance for their use in, e.g., personal care, medical or pharmaceutical applications. In addition, such applications require biocompatible polymers and non-toxic solvents like water–ethanol mixtures. Solution polymer properties in water are well-studied including self-assembly of amphiphilic copolymers into, e.g., micelles^1,2 and vesicles.³ In addition, the lower critical solution temperature (LCST; precipitation upon heating based on the hydrophilicity–hydrophobicity balance) of polymers in water has been studied in detail.^4,5 Solution polymer properties are often optimized for specific applications by changing the (co)polymer composition. However, it would be desirable to tune the solution properties of one single polymer for different applications. Shultz and Flory⁶ as well as Wolf and Blaum⁷ have demonstrated that the solubility of polymers can be affected in an unpredictable way using binary solvent mixtures resulting in better solubility compared to the individual pure solvents. In addition, the aggregation of polymers is also strongly effected by the use of solvent mixtures as was demonstrated by, e.g., Leharne et al.,⁸ Holmqvist and co-workers⁹ as well Yu and Eisenberg.¹⁰ However, surprisingly little is known about solution polymer properties in water–ethanol mixtures despite that it is well-known that such mixtures exhibit interesting abnormal properties due to the presence of hydration shells around the ethanol molecules.^11–13 The presence of such hydration shells can result in solubility maxima for drug molecules in water–ethanol mixtures.^14–16 Furthermore, it was demonstrated that the aggregation of surfactants can be tuned by water–ethanol mixtures.^17–19 Similarly, the self-assembly of amphiphilic polymers could be tuned in water–ethanol mixtures, whereby both the size and the shape of the resulting aggregates depends on the solvent composition.^8,20–22 In addition, the low toxicity of water–ethanol mixtures make them particularly interesting for possible applications in drug delivery and personal care formulations. Nonetheless, up to now the self-assembly of only a few individual polymers has been studied in water–ethanol. To gain more detailed knowledge and understanding of the effects of binary water–ethanol mixtures on solution polymer properties, systematic investigations are required in which the composition of both the polymer and the solvent mixture should be varied.

Here, we report our systematic investigations on the solution polymer properties of amphiphilic copolymers in water–ethanol mixtures. More specifically, the solution properties of quasi-diblock statistical copolymers based on 2-phenyl-2-oxazoline (PhOx; hydrophobic) and 2-methyl-2-oxazoline (MeOx; hydrophilic) or 2-ethyl-2-oxazoline (EtOx; hydrophilic)²³ were studied in solvent mixtures ranging from pure water to pure ethanol (steps of 20 wt.%). This systematic screening allowed the detailed evaluation of the effect of solvent composition on the solution polymer properties including solubility, LCST, self-assembly and dispersion formation.

Experimental

The synthesis and structural characterization of the quasi-diblock statistical copolymers consisting of PhOx and MeOx or EtOx is described elsewhere.²³Ethanol (Biosolve) was used as received and binary solvent mixtures were prepared with deionized water.

Solubility screening

The solubility screening was performed by heating the polymer (5.0 ± 0.2 mg) in the binary solvent mixture (1.0 mL) from 20 to 75 °C with a heating rate of 1.0 °C per minute followed by cooling to 20 °C at a cooling rate of 1 °C per minute after keeping it 10 minutes at 75 °C. During these controlled heating cycles the transmission through the solutions was monitored in the Crystal16 from Avantium Technologies.²⁴ All vials were visually inspected after the heating program to facilitate interpretation of the observed transmission profiles. Polymer–solvent combinations that gave unexpected transmission profiles were rerun and the solution was visually inspected at different stages during the heating run.

Transmission electron microscopy of the micelles

Transmission electron microscopy measurements were performed on a FEI Tecnai 20, type Sphera TEM operating at 200 kV (LaB₆ filament). Images were recorded with a bottom-mounted 1 × 1 k Gatan CCD camera. 200 mesh carbon-coated copper grids for TEM were purchased from SPI. Prior to blotting, the grids were made hydrophilic by surface plasma treatment using a Cressington 208 carbon coater operating at 5 mA for 40 seconds. For sample preparation a droplet was applied to a 200 mesh carbon-coated copper grid and subsequently excess liquid was quickly manually blotted away with filter paper.

Results and discussion

Screening of the solution polymer properties

The investigated statistical copolymer libraries ranged from pure pPhOx (hydrophobic) to pure pMeOx or pEtOx (both hydrophilic) while the total degree of polymerization was kept constant at 100 monomer units. The composition of the statistical copolymers was varied in steps of 10 mol%, and due to the higher reactivity of the MeOx and EtOx compared to the PhOx, the monomers were distributed in a quasi-diblock copolymer architecture.²³ As such, the statistical copolymers will have outer blocks of pure PhOx and MeOx or EtOx, which are connected through a gradient block in which the composition gradually changes from MeOx or EtOx to PhOx. The solubility screening of these statistical copolymer libraries was performed by measuring turbidity of the copolymer samples in water–ethanol mixtures as a function of temperature in the range from 20 to 75 °C. The resulting turbidity curves, together with a visual inspection of the solutions, revealed the solubility for all distinct combinations of the polymer library and water–ethanol mixtures as summarized in Fig. 1. Selected examples of the raw measurement data and their interpretation are given in the ESI†. The solubility phase diagram for the MeOx–PhOx copolymers (Fig. 1, top) reveals that all copolymers with up to 30 mol% of the hydrophobic pPhOx are soluble in all solvent compositions. With increasing mol-fraction of PhOx, the polymers become less soluble in the water-rich solutions, while the copolymers with 70 mol% or more PhOx are insoluble in pure water and only dissolve in ethanol with heating indicating that they exhibit an upper critical solution temperature (UCST).

Fig. 1 Solubility overview for the MeOx–PhOx (top) and EtOx–PhOx (bottom) copolymer libraries in water–ethanol mixtures (5 mg mL⁻¹).

Surprisingly, the copolymers with 80 mol% or more PhOx are soluble at 20 °C with 80 wt.% ethanol and exhibit an UCST in pure ethanol and with 60 wt.% ethanol indicating a solubility maximum, which will be discussed in more detail later. For the solvent mixtures up to 40 wt.% ethanol, the solubility decreases with increasing PhOx mol-fraction as is evident from the transition from completely soluble to micellization, followed by the formation of dispersions and ending with polymers that do not dissolve at all. Furthermore, the effect of ethanol can be clearly seen by the shift of these transitions to higher PhOx-content in solutions with more ethanol. At this point it should be noted, that only those solutions that were translucent could be identified as samples that contained micelles (∼15 nm or larger) during this screening. Nonetheless, a critical micellization solvent composition that shifts to higher ethanol contents with a larger hydrophobic fraction could be identified for the different copolymers. Similar observations were made during the screening of the EtOx–PhOx copolymers (Fig. 1, bottom) including the solubility maximum at high PhOx-content and high ethanol content. However, the more hydrophobic ethyl side chains result in decreased solubility in solutions with high water content and increased solubility in solutions with high ethanol contents. In addition, no clear micellization was observed during the screening of the EtOx-containing copolymers and the EtOx–PhOx copolymers with low PhOx-content revealed a LCST in pure water and 20 wt.% ethanol mixtures.

To summarize, this initial solubility screening of the MeOx–PhOx and EtOx–PhOx copolymers revealed a number of interesting phenomena that were investigated in more detail including the solubility maximum, micellization, dispersion formation and LCST behavior. The effect of solvent composition on these phenomena will be discussed in more detail.

Solubility

The observed solubility maximum was investigated in more detail for the MeOx–PhOx copolymers using a wider temperature range from −15 to 75 °C. The precipitation temperatures (T_p; 50% transmission point in the cooling run), which give an approximation of the UCSTs, were determined as function of wt.% ethanol in the solution for different copolymers (Fig. 2). The T_p for the pPhOx decreased tremendously when only a small amount of water was added to the ethanol solution. More specifically, the T_p dropped from 48.1 to −7.9 °C when only 5 wt.% of water was added.

Fig. 2 Precipitation temperatures for MeOx–PhOx copolymers in water–ethanol mixtures (5 mg mL⁻¹).

When the amount of water was further increased, the polymer became completely soluble within the investigated temperature range for solutions containing 6 wt.% up to 25 wt.% water (corresponding to 94 wt.% and 75 wt.% ethanol in Fig. 2, respectively). The solubility of the pPhOx decreased with the addition of more water, while the polymer became completely insoluble in the investigated temperature range with 45 wt.% water. Increasing the hydrophilicity of the polymers by incorporation of MeOx led to increased solubility and broadening of the soluble regime as expected. The increased solubility with the addition of water is most likely due to the formation of a hydration shell¹⁶ around the carbonyl groups of the polymer that can act as a compatibilizing layer between the polymer and the ethanol. The enhanced solubilization of the polymer was only observed up to 25 wt.% water in ethanol, which can be ascribed to the presence of monomeric water molecules that effectively hydrogen bond to the polymer.¹³ Further increasing the amount of water leads to the formation of water clusters, which increases the polarity of the solvent mixture and thus decreases the polymer solubility. This observed solubility maximum for certain water contents might be of future interest for the preparation of drug or personal care formulations that include hydrophobic polymers.

Micellization

The second observed phenomenon during the initial solubility screening is the existence of a critical micelle solvent composition (cmsc), i.e. at a certain solvent composition translucent solutions were observed for the different amphiphilic statistical copolymers consisting of MeOx and PhOx. With increasing amount of PhOx, a larger ethanol content was required to obtain translucent solutions due to the higher hydrophobic content making the polymer less soluble. At this point it should be noted that the observed translucent solutions demonstrate the existence of such a cmsc, but the determination of the actual cmsc will require smaller steps in solvent composition.

Nonetheless, the observed translucent solutions were investigated with transmission electron microscopy (TEM; Fig. 3). The TEM-images of pMeOx₆₀-stat-PhOx₄₀ in water (Fig. 3A) and pMeOx₄₀-stat-PhOx₆₀ in water with 20 wt.% ethanol (Fig. 3C) revealed the presence of both cylindrical and spherical micelles, whereby the self-assembly of the pMeOx₄₀-stat-PhOx₆₀ also resulted in the formation of merged cylinders, so-called Y-junctions.^25–27 The self-assembly of pMeOx₅₀-stat-PhOx₅₀ in water with 20 wt.% ethanol (Fig. 3B) and pMeOx₃₀-stat-PhOx₇₀ in water with 40 wt.% ethanol (Fig. 3D) resulted in the formation of separate or a mixture of separate and clustered spherical micelles, respectively. The size of the self-assembled structures is increasing from ∼20.5 nm for pMeOx₆₀-stat-PhOx₄₀ to ∼28 nm for pMeOx₃₀-stat-PhOx₇₀ due to the larger fraction of hydrophobic PhOx. However, in-depth interpretation of these results requires further systematic investigations on the effects of both polymer composition and solvent composition on the self-assembly, which will be the focus of future investigations.

Fig. 3 TEM images of the micellar structures obtained from pMeOx₆₀-stat-PhOx₄₀ in water (A), pMeOx₅₀-stat-PhOx₅₀ (B) and pMeOx₄₀-stat-PhOx₆₀ (C) in water with 20 wt.% ethanol as well as pMeOx₃₀-stat-PhOx₇₀ in water with 40 wt.% ethanol (D).

Dispersion stability

In between the non-soluble and soluble regimes in the solubility overview (Fig. 1), an intermediate regime with polymer dispersions was observed. The effect of water–ethanol mixtures on the dispersion stability was investigated in detail for the pMeOx₂₀–PhOx₈₀ copolymer, which is insoluble and not dispersed in pure water. The different polymer–water–ethanol mixtures were heated to 75 °C to prepare the dispersions and, subsequently cooled to 20 °C to study the dispersion stability. The stirring was stopped and the transmission through the samples was followed in time. The resulting transmission curves for different water–ethanol mixtures are displayed in Fig. 3 demonstrating the effect of ethanol on the dispersion stability.

From these curves the 5% transmission points were determined as measure for the stability (Table below Fig. 4). The polymer dispersions in 10 and 15 wt.% ethanol solutions did not reach 0% transmission indicating that no stable dispersions were formed at all. With increasing wt.% of ethanol, the dispersions became more stable and with 35 and 40 wt.% the dispersions were stable for more than 1 week. When the amount of ethanol was further increased to 45 wt.% the stability was decreased, which is most likely due to partial solvation of the copolymer particles making them more sticky leading to coagulation. The increasing stability for polymer dispersions in water–ethanol mixtures was unprecedented, although it has been reported that the stability of aqueous silica suspensions increases with the addition of ethanol.^28,29

Fig. 4 Dispersion stability for pMeOx₂₀–PhOx₈₀ in different water–ethanol mixtures.

LCST behavior

Poly(2-ethyl-2-oxazoline) (co)polymers are known to exhibit a LCST in water, whereby the LCST can be tuned by either changing the polymer length,^30,31 composition,³²salt concentration³⁰ or water–dioxane solvent ratio.³⁰ In addition, the LCST of poly(vinyl methyl ether) in water was reported to increase with the addition of methanol,³³ the LCST of poly(N-isopropylacrylamide) in water first decreased and than increased by the addition of methanol³³ while the LCST of poly(oligoethylene oxide phosphazene)s in water was not affected by the addition of either methanol or ethanol.³⁴ During the course of this study we found that the pEtOx–PhOx copolymers with high EtOx content show a LCST transition in water or in solutions containing 20 wt.% ethanol. The LCST was determined upon both heating and cooling for pEtOx₈₀–PhOx₂₀ and pEtOx₇₀–PhOx₃₀ with different amounts of ethanol (Fig. 5, top).

Fig. 5 Top: LCST as function of wt.% of ethanol for pEtOx₇₀–PhOx₃₀ and pEtOx₈₀–PhOx₂₀. Bottom: transmittance as a function of temperature for pEtOx₅₀–PhOx₅₀ in 40 wt.% ethanol in water demonstrating a LCST transition as well as an UCST.

The LCST increases with increasing ethanol content due to the better solvation of the copolymers. Furthermore, the LCST of the more hydrophilic copolymer (pEtOx₈₀–PhOx₂₀) is higher as expected. The solubility transitions upon heating and cooling are only a few degrees apart demonstrating the good reversibility and the absence of strong hysteresis for the LCST transition. In addition, a remarkable observation was made for the pEtOx₅₀–PhOx₅₀ copolymer in an aqueous solution with 40 wt.% ethanol revealing both a LCST and a subsequent UCST (Fig. 5, bottom), so-called closed-loop coexistence,^35,36 which further demonstrates the abnormal mixing behavior of water–ethanol mixtures. Such closed-loop coexistence for polymer solutions was previously only observed with poly(ethylene oxide) in water.^37,38 The LCST transition proceeds via a two-step process both during heating and cooling. In the first shallower decrease in transmission micellar aggregates are probably formed that fully precipitate during the second sharper decrease in transmission. Upon further heating, the polymer redissolved and the solution became clear again (Fig. 5, bottom). This peculiar solubility behavior, which might be used for special sensing or light-blocking coatings with solutions that become opaque in a narrow temperature range, will be the focus of future investigations.

Conclusions

In summary, we demonstrated that the solution properties of amphiphilic statistical copolymers can be tuned in a wide range by only changing the composition of the water–ethanol mixtures leading to increased solubility, a critical micelle solvent composition, stable dispersions as well as LCST transitions followed by an UCST. In addition, it is believed that the improved polymer solubility in low toxicity water–ethanol mixtures might be beneficial for pharmaceutical and personal care applications of polymers.

Future work will include detailed investigation on the self-assembly of the pMeOx-stat-PhOx copolymers in water–ethanol mixtures as well as the polymer–solvent combinations that exhibit both LCST and UCST behavior. Moreover, the solution polymer properties of a variety of other (co)polymer structures as well as other aqueous binary solvent mixtures will be studied.

Acknowledgements

The authors would like to thank the Dutch Polymer Institute, the Dutch council of scientific research (NWO) and the Fonds der Chemischen Industrie for financial support. Avantium Technologies BV is thanked for the discussions during the course of the investigations. This research has been carried out with the support of the Soft Matter cryo-TEM Research Unit, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology.

References

1. Y.-Y. Won, H. T. Davis and F. S. Bates, Science, 1999, 283, 960.
2. J.-F. Gohy, Adv. Polym. Sci., 2005, 190, 65.
3. D. E. Discher and A. Eisenberg, Science, 2002, 297, 967.
4. G. T. Caneba and L. Shi, Lower Critical Solution Temperature of Polymer–Small Molecule Systems: A Review, Research Signpost, Trivandrum, India, ch. 4, 2002.
5. H. G. Schild, Prog. Polym. Sci., 1992, 17, 163.
6. A. R. Shultz and P. J. Flory, J. Am. Chem. Soc., 1953, 75, 5681.
7. B. A. Wolf and G. Blaum, J. Polym. Sci,, Polym. Phys. Ed., 1975, 13, 1115.
8. J. Armstrong, B. Chowdhry, J. Mitchell, A. Beezer and S. Leharne, J. Phys. Chem., 1996, 100, 1738.
9. P. Holmqvist, P. Alexandridis and B. Lindman, Langmuir, 1997, 13, 2471.
10. Y. Yu and A. Eisenberg, J. Am. Chem. Soc., 1997, 119, 8383.
11. H. S. Frank and M. W. Evans, J. Chem. Phys., 1945, 13, 507.
12. F. Franks, Q. Rev. Chem. Soc., 1966, 20, 1.
13. S. Y. Noskov, G. Lamoureux and B. Roux, J. Phys. Chem. B, 2005, 109, 6705.
14. J. B. Escalera, P. Bustamante and A. Martin, J. Pharm. Pharmacol., 1994, 46, 172.
15. S. Romero, A. Reillo, B. Escalera and P. Bustamante, Chem. Pharm. Bull., 1996, 44, 1061.
16. P. Bustamante, J. Navarro, S. Romero and B. Escalera, J. Pharm. Sci., 2002, 91, 874.
17. A. Goto, M. Takemoto and F. Endo, Bull. Chem. Soc. Jpn., 1985, 58, 247.
18. A. Cipiciani, G. Onori and G. Savelli, Chem. Phys. Lett., 1988, 143, 505.
19. K. Esumi and S. Ogiri, Colloids Surf., A, 1995, 94, 107.
20. W. Zhang, L. Shi, Y. An, L. Gao, K. Wu and R. Ma, Macromolecules, 2004, 37, 2551.
21. W. Zhang, L. Shi, Y. An, L. Gao, K. Wu, R. Ma and B. Zhang, Macromol. Chem. Phys., 2004, 205, 2017.
22. E. Castro, P. Taboada and V. Mosquera, J. Phys. Chem. B, 2006, 110, 13113.
23. R. Hoogenboom, H. M. L. Thijs, M. W. M. Fijten, B. van Lankvelt and U. S. Schubert, J. Polym. Sci., Part A: Polym. Chem., 2007, 45, 416.
24. M. Birhc, S. J. Fussel, P. D. Higginson, N. McDowall and I. Marziano, Org. Process Res. Dev., 2005, 9, 360.
25. S. Jain and F. S. Bates, Science, 2003, 300, 460.
26. N. Dan, K. Shimoni, V. Pata and D. Danino, Langmuir, 2006, 22, 9860.
27. P. Bhargava, J. X. Zheng, R. P. Quirk and S. Z. D. Cheng, J. Polym. Sci., Part B: Polym. Phys., 2006, 44, 3605.
28. J. Ren, S. Song, A. Lopez-Valdivieso, J. Shen and S. Lu, J. Colloid Interface Sci., 2001, 238, 279.
29. A. N. Zhukov, L. I. Zavarovskaya and Y. M. Chernoberezhskii, Colloid J., 2006, 68, 558.
30. P. Lin, C. Clash, E. M. Pearce and T. K. Kwei, J. Polym. Sci., Part B: Polym. Phys., 1988, 26, 603.
31. D. Christova, R. Velichkova, W. Loos, E. J. Goethals and F. Du Prez, Polymer, 2003, 44, 2255.
32. J.-S. Park and K. Kataoka, Macromolecules, 2006, 39, 6622.
33. H. G. Schild, M. Muthukumar and D. A. Tirell, Macromolecules, 1991, 24, 948.
34. S. B. Lee, S.-C. Song, J.-I. Jin and Y. S. Sohn, Polym. Bull., 2000, 45, 389.
35. J. C. Weeler, J. Chem. Phys., 1975, 62, 433.
36. J. C. Lang and R. D. Morgna, J. Chem. Phys., 1980, 73, 5849.
37. G. N. Malcolm and J. S. Rowllinson, Trans. Faraday Soc., 1957, 53, 921.
38. E. E. Dormidontova, Macromolecules, 2002, 35, 987.

---

Footnote

† Electronic supplementary information (ESI) available: Synthetic procedures and explanation of selected transmission profiles obtained during the solubility screening. See DOI: 10.1039/b712771e

---