Biomacromolecules 2003, 4, 1041-1044
1041
Two-Dimensional/ATR Infrared Correlation Spectroscopic Study on Water Diffusion in a Poly(E-caprolactone) Matrix Yun Peng,† Peiyi Wu,*,† and H. W. Siesler‡ Department of Macromolecular Science and The Key Laboratory of Polymer Engineering Science, Fudan University, 200433 Shanghai, PR China, and Department of Physical Chemistry, University of Essen, D 45117 Essen, Germany Received March 3, 2003; Revised Manuscript Received April 24, 2003
In the present contribution, two-dimensional ATR-FTIR spectroscopy was used to study the diffusion of water in poly(-caprolactone) (PCL). In the spectral region of the ν(OH) stretching vibration of water, four absorption bands (3635, 3560, 3411, and 3220 cm-1) can be identified. The higher wavenumber band pair at 3635 and 3560 cm-1 is assigned to the antisymmetric and symmetric OH stretching vibrations, respectively, of water which is partially hydrogen-bonded to the carbonyl groups of PCL, whereas the lower frequency band pair at 3411 (antisymmetric) and 3220 cm-1 (symmetric) is attributed to the OH stretching vibrations of bulk water which is fully hydrogen-bonded to other water molecules. From the asynchronous map of a 2D correlation analysis of spectra recorded during the diffusion of water into PCL, it was concluded that during the diffusion process the water molecules first penetrate into the free volume (microvoids) of the PCL matrix or are molecularly dispersed in the polymer matrix and then form hydrogen bonds with the CdO groups of the polymer. Introduction In recent years, increasing interest has been devoted to the study the diffusion of water in polymer matrixes. PCL, an attractive biopolymer that has been used in many fields, is a material which easily absorbs water because of the hydrophilic CdO group. The study of water diffusion in PCL will therefore contribute toward a better understanding and exploitation of this biocompatible material. There have been a large number of papers published on the diffusion of water into polymers,1-4 because it was realized that water hydrogen-bonded to the polymer is different from bulk water (e.g., absorbed in the free volume of the polymer) because of the restricted motion of the former species. Therefore, the study of the dynamics of water diffusion in polymers and the structural relations between the polymer matrix and the diffusant are an attractive field of research. There have been several sorption models established for the state of water molecules in polymer networks. In some cases, water was assumed to diffuse into the free volume of the polymer; other researchers have adopted the interaction concept in which water molecules couple strongly with certain hydrophilic groups (e.g.. hydroxyl or amine) to form hydrogen bonds with the polymer. Apart from these different approaches, however, it is well accepted that the properties of water diffusing into a polymer are different from those of pure water. Woo and Piggott1 suggested clustering of water molecules in polymer networks. Moy2 reported that in lower concentra* To whom correspondence should be addressed. Fax: +86-2565640293. E-mail:
[email protected]. † Fudan University. ‡ University of Essen.
tions water molecules interact with specific polymer chain segments or groups, but as the water concentration increases, multilayer sorption water molecules existed as free water. Sammon3 found that there exist four different ν(OH) stretching vibrations of water diffusing into a polymer but that there are differences observed for these ν(OH) stretching vibrations for different polymer materials. According to his hypothesis, the water molecules form four types of hydrogen bonding (weakly H-bonded, moderately weak H-bonded, moderately strong H-bonded, and strongly H-bonded)3 with hydrophilic groups of the polymer, which lead to four absorption bands. However, each type of water would give rise to two bands corresponding to the symmetric and antisymmetric stretching modes of water, thus, complicating the verification of the theory of Sammon. Finney4 has summarized and emphasized the need to clarify the details of this phenomenon: “We need a better understanding of the effect of hydration on structural and dynamic properties in order to control industrially relevant processs”. One approach to fulfill this need is to start with the characterization of the properties of this bonded water. Indeed, much work has been published in this area already using spectroscopic techniques, such as NMR,5-7 FTIR and Raman,8,9 neutron scattering,10-12 and light scattering.13 The main problem tends to be the ability to distinguish bound water from bulk water and to shed more light on the dynamics of water diffusion in polymers. In our study, a novel experimental approach, based on time-resolved two-dimensional (2D) ATR-FTIR spectroscopy has been used to study the diffusion of water in PCL. As experimental technique, we have used ATR/FTIR spectroscopy14-16 which is ideally suited to probe the surface of
10.1021/bm0340624 CCC: $25.00 © 2003 American Chemical Society Published on Web 06/14/2003
1042
Biomacromolecules, Vol. 4, No. 4, 2003
polymer films because the penetration depth of the IR beam into the sample varies generally between about 0.5-10 µm.17 The technique has been used in the study of the diffusion of small molecules into thin polymer films.18-20 Spectra of the polymer surface in contact with the ATR crystal can be collected in real time during the diffusion of water into the polymer. This method is especially useful for in situ analysis of diffusion into polymers, as it overcomes the problems associated with “blot and weigh” immersion techniques that are normally applied for this purpose.21 Generalized two-dimensional (2D) correlation spectroscopy proposed by Noda,22-25 which is an extension of the original 2D correlation spectroscopy, has recently been used to evaluate the IR spectroscopic data of water diffusion in a polymer matrix.26 In the 2D analysis, two kinds of correlation maps, synchronous and asynchronous, are generated based upon a set of dynamic spectra calculated from the fluctuations of the spectroscopic signals during the process under examination.25 The generalized 2D method can handle signals fluctuating as an arbitrary function of time or any other physical variable such as temperature, pressure, or concentration.27-29 Correlation peaks appearing in the synchronous and asynchronous maps represent in-phase and out-of-phase variations, respectively, of band intensities as a consequence of the investigated phenomenon. There are several advantages of the 2D correlation analysis:22, 25 1. It has powerful deconvolution ability for highly overlapped bands. 2. It provides information about inter- and intramolecular interactions by correlating absorption band intensities of different functional groups. 3. The specific sequence of the spectral intensity changes taking place during the measurement can be derived from the analysis of the asynchronous spectra. Time-resolved ATR-FTIR spectra were measured during the process of water diffusion, and synchronous and asynchronous spectra were calculated for the spectral range of 3800-3000 cm-1. The broad ν(OH) absorption of the 1DIR spectra can be effectively deconvolved, and the band positions and relative variation rates are determined by using 2D correlation analysis. On the basis of these data, the diffusion of water into PCL is discussed in some detail in this article. Experimental Section The PCL (purchased from Solvay Co. in Britain) and the solvent tetrahydrofuran (THF) (purchased from Feida Industry and Trade Co. in Shanghai, China) were used without further purification. A thin polymer film was prepared by casting from a clear solution of PCL in THF (2%(m/v)). The solution was cast onto a 75 mm × 25 mm microscope slide. The slide was kept under vacuum at room temperature for about 12 h in order to remove residual solvent. Finally, the film was peeled off the glass support for the diffusion experiments. The time-resolved ATR-FTIR measurements were carried at 26 °C using a Nicolet Nexus Smart ARK FTIR spectrom-
Peng et al.
Figure 1. ATR/FTIR spectra measured during diffusion of water into PCL.
eter equipped with a DTGS detector, an ATR cell with a ZnSe reflection element. For the diffusion experiments, a PCL film with a thickness of 16 µm was sandwiched between the ZnSe reflection element and filter paper and then mounted in the ATR cell. Distilled water was then injected into the filter paper while starting the data acquisition using a macro program. The spectra were collected at a spectral resolution of 4 cm-1 by accumulating 16 scans. The time interval between two successive spectra is 2 min. The measured wavenumber range was 4000-650 cm-1. All of the original spectra were baseline corrected using the Omnic 5.1 software. For the generalized 2D correlation analysis, five spectra measured at intervals of 2 min (wavenumber range 37003000 cm-1) were selected and subjected to the 2D software named 2D Pocha (developed by Daisuke Adachi, Kwansei Gakuin University, Nishinomiya, Japan). Results and Discussions 1D IR Spectra of the Investigated System. The ATR/ FTIR spectra obtained during the diffusion of water into the polymer as a function of time are shown in Figure 1. The ν(OH) stretching band of water is the dominant feature in this spectral region. From these spectra, it is obvious that the intensity of the spectra in the 3700-3000 cm-1 and in the 1800-1500 cm-1 regions increased gradually as a function of time when water diffuses into the polymeric film. Compared to the strong intensity of the ν(OH) absorption band, the δ(OH) band at 1660-1600 cm-1 is relatively weak and varies not very significantly which limits its use for measuring water transport in the PCL film. Hence, the band in the spectral range 3700-3000 cm-1 was selecteded for the subsequent 2D correlation analysis. 2D Correlation Analysis. Figure 2, parts A and 2B, shows the synchronous and asynchronous correlation spectra of PCL in the spectral range 3700-3000 cm-1. Note that only one autopeak was observed at about 3380 cm-1 in the synchronous correlation spectrum (Figure 2A). More information can be derived from the corresponding asynchronous correlation spectrum (Figure 2B). The asynchronous spectrum
Water Diffusion in a Poly(-caprolactone) Matrix
Biomacromolecules, Vol. 4, No. 4, 2003 1043
Figure 3. Infrared spectrum of water in the range 3000-3700 cm-1. (A) pure liquid water; (B) water in PCL matrix.
Figure 2. 2D-IR correlation spectra of the 3700-3000 cm-1 wavenumber region: (A) synchronous; (B) asynchronous contour maps. (In 2D IR correlation spectra, the unshaded and shaded areas in the contour maps represent positive and negative peaks, respectively, and the time-averaged 1D reference spectrum is represented at the left side and at the top of the map.)
is antisymmetric with respect to the diagonal line. In the 2D spectra, an asynchronous cross-peak develops only if the intensities of two spectral features change out of phase (i.e., delayed or accelerated) with each other. Thus, the absence of an asynchronous cross-peak denotes that the two spectral features change synchronously. Two negative cross-peaks (3635/3411 cm-1 and 3560/3411 cm-1) and one positive cross-peak (3411/3220 cm-1) observed in the upper left side of Figure 2B indicate that the broad ν(OH) water band in the spectral region 3700-3000 cm-1 is split into four separate bands located at 3635, 3560, 3411, and 3220 cm-1. The sign of the cross-peaks in an asynchronous spectrum provides additional information about the order of the intensity changes in different bands. According to the rule of Noda,22 the band at 3411 cm-1 varies prior to the band at 3220 cm-1 and the band at 3635 cm-1 changes after 3411 cm-1;
similarly, the band at 3560 cm-1 varies after the band at 3411 cm-1. Therefore, the signs of the 2D-IR asynchronous spectrum suggest that the band pair at 3411 and 3220 cm-1 vary earlier than the other band pair at 3635 and 3560 cm-1. The ATR/FTIR spectrum of pure liquid water (Figure 3) has an intense and broad band envelope that is assigned to the antisymmetric (3400 cm-1, ν3) and symmetric (3205 cm-1, ν1) OH stretching vibrations, respectively, of water which is fully hydrogen-bonded to the environment.30-37 The fact that the ν(OH) band enveloped in the ATR-IR spectra of the water in PCL moves to higher wavenumber (compared to the pure liquid water) indicates that the water hydrogen bond is weakened because of the interaction of water with PCL. Thus, the two lower wavenumber bands derived from the 2D results, 3411 and 3220 cm-1, are assigned to the ν3 and ν1 modes of water molecules in the fully hydrogenbonded environment of bulk water. The two additional absorption bands at approximately 3635 and 3560 cm-1 in the infrared spectra of our studies may be assigned to vibrations of partially hydrogen-bonded water molecules. Partially hydrogen-bonded water has one of its OH groups participating in hydrogen bonding to other water molecules, whereas the second OH group is considered as free or weakly hydrogen-bonded with other groups (in our case, probably with the CdO groups in PCL) and appears at higher wavenumbers. In addition, according to the asynchronous map in Figure 2B, the bands due to the fully hydrogenbonded ν1 + ν3 OH of water (bulk water) vary ahead of the partially hydrogen bonded ν1 + ν3 OH of water (bound water), which means that the intensity changes of water could be caused by the following process: during the process of water uptake in the PCL, the water molecules at first diffuse into free volume (microvoids)26,38 or are molecularly dispersed in the polymer matrix and then form hydrogen bonds with the CdO group of the polymer. Conclusions In the presented study, ATR/FTIR spectra were measured as a function of time to study the diffusion of water in PCL. The application of generalized 2D correlation spectroscopy
1044
Biomacromolecules, Vol. 4, No. 4, 2003
has enabled us to detect spectral variations as a consequence of the water uptake in the polymer. The following conclusions can be achieved from the present study. The splitting of the ν(OH) absorption band in the region of 3700-3000 cm-1 in the 2D-IR spectra elucidates that there are two different states of water in PCL networks, for which the band pair at 3635 and 3560 cm-1 can be assigned to bound water forming partially hydrogen bonds with the polymer, and the other band pair at 3411 and 3220 cm-1 could be attributed to bulk water forming hydrogen-bonds with other water molecules. Furthermore, the following conclusion could be drawn from the asynchronous correlation spectra: water molecules at first penetrate into microvoids of the polymer or are molecularly dispersed in the polymer matrix and then form hydrogen bonds with carbonyl groups of the polymer. Acknowledgment. The study reported in this paper was supported by the National Science of Foundation of China (NSFC) (Nos. 20274010 and 50103003), the “Qimingxing” Project (No. 01QE14011) of Shanghai Municipal Science and Technology Commission, and the “Shuguang” Project (No. 01SG05) of the Shanghai Municipal Education Commission and Shanghai Education Development Foundation. References and Notes (1) Woo, M.; Piggott, M. J. Compos. Technol. Res. 1987, 9, 101. (2) Moy, P.; Karasz, F. E. Polym. Eng. Sci. 1980, 20, 315. (3) Sammon, C.; Mura, C.; Yarwood, J.; Everall, N.; Swart, R.; Hodge, D. J. Phys. Chem. B 1998, 102, 3402. (4) Finney, J. L. Faraday Discuss. 1996, 103, 1 and references therein. (5) Otting, G.; Liepinsh, E.; Wuthrich, K. J. Am. Chem. Soc. 1991, 113, 4363. (6) Emst, J. A.; Chubb, R. T.; Zhou, H. X.; Grogenhorn, A. M.; Clore, G. M. Science 1995, 267, 1813. (7) Denisov, V. P.; Halle, B. Faraday Discuss. 1996, 103, 227 and references therein.
Peng et al. (8) Marechal, Y.; Chamel, A. J. Phys. Chem. 1996, 100, 8551. (9) Crupi, V.; Jannelli, M. P.; Magazu, S.; Maisano, G.; Majolino, D.; Migliardo, P.; Ponterio, R. J. Mol. Struct. 1996, 381, 219. (10) Settles, M.; Doster, W. Faraday Disscuss. 1996, 103, 269. (11) Bryant, R. G. Annu. ReV. Biophys. Biomol. Struct. 1996, 25, 29. (12) Wanderlingh, U.; Giordano, R.; Teixeira, J. J. Mol. Struct. 1993, 296, 271. (13) Tao, N. J.; Lindsay, S. M.; Rupprecht, A. Biopolymers 1989, 28, 1029. (14) Fieldson, G. T.; Barbari, T. A. Polymer 1993, 34, 1146. (15) Marechal, Y.; Chamel, A. J. Phys. Chem. 1996, 100, 8551. (16) Cotugno, S.; Larobina, D.; Mensitieri, G.; Musto, P.; Ragosta, G. Polymer 2001, 42, 6431. (17) Dias, C. R.; Rosa, M. J.; de Pinho, M. N. J. Membr. Sci. 1998, 21, 259. (18) Fieldson, G. T.; Barbari, T. A. AIChE J. 1995, 41, 795. (19) Pereira, M. R.; Yarwood, J. J. Polym. Sci. B. Polym. Phys. 1994, 32, 1881. (20) Scholtter, N. E.; Furan, Y. P. Vib. Spectrosc. 1992, 3, 147. (21) Balik, C. M.; Simendinger, W. H. Polymer 1998, 39, 4723. (22) Noda, I. Appl. Spectrosc. 1993, 47, 1329. (23) Noda, I. Bull. Am. Phys. Soc. 1986, 31, 520. (24) Noda, I. J. Am. Phys. Soc. 1989, 111, 8116. (25) Noda, I. Appl. Spectrosc. 1990, 44, 550. (26) Liu, M. J.; Wu, P. Y.; Ding, Y. F.; et al. Macromolecules 2002, 35, 5500. (27) Noda, I.; Liu, Y.; Ozaki, Y.; Czarnecki, M. A. J. Phys. Chem. 1995, 99, 3068. (28) Noda, I.; Story, G. M.; Marcott, C. Vib. Spectrosc. 1999, 19, 461. (29) Wu, Y. Q.; Murayama, K.; Czarnik-matusewicz, B.; Ozaki, Y. Appl. Spectosc. 2002, 56, 1186. (30) Carey, D. M.; Korenowski, G. M. J. Chem. Phys. 1998, 108, 2669. (31) Green, J. L.; Lacey, A. R.; Sceats, M. G. J. Phys. Chem. 1986, 90, 3958. (32) Brooksby, P. A.; Fawcett, W. R. J. Phys. Chem. A 2000, 104, 8307. (33) Marechal, Y. J. Mol. Struct. 1994, 322, 105. (34) Marechal, Y. J. Phys. Chem. 1993, 97, 2846. (35) Hare, D. E.; Sorensen, C. M. J. Chem. Phys. 1992, 96, 13. (36) Brooksby, P. A.; Fawcett, W. R. J. Phys. Chem. A 2000, 104, 8307. (37) Eisenberg, D.; Kauzmann, W. The Structure and Properties of Water; Clarendon: London, 1969; Chapter 4. (38) Han, H.; Gryte, C.; Ree, M. Polymer 1995, 36, 1663.
BM0340624
