pubs.acs.org/Langmuir © 2009 American Chemical Society
On the Temperature-Responsive Polymers and Gels Based on N-Propylacrylamides and N-Propylmethacrylamides† Mamoru Kano‡ and Etsuo Kokufuta* Institute of Applied Biochemistry, University of Tsukuba, Tsukuba, Ibaraki 305-8572, Japan. ‡ Temporary guest scientist from Mitsui Chemical Analysis & Consulting Service Inc., Sodegaura, Chiba 299-0265, Japan. Received December 29, 2008. Revised Manuscript Received February 17, 2009 We studied the behavior in water of polymers, microgels, and macrogels based on the following four monomers: N-isopropylacrylamide (NiPA), N-isopropylmethacrylamide (NiPMA), N-n-propylacrylamide (NnPA), and N-npropylmethacrylamide (NnPMA). The thermal phase separation of polymers in water as well as of microgels in the aqueous dispersion was examined by a combination of turbidity measurements and differential scanning calorimetry (DSC). The hydrodynamic radius of microgels and the swelling degree of macrogels (fine cylindrical bulk gels) were also examined as a function of temperature using the dynamic light scattering and the microscopic method, respectively. It was found that all the polymers prepared are water-soluble and clearly exhibit the phase separation on heating. The phase separation temperature varies depending on the constituent monomers and becomes higher in the order of NiPMA > NiPA > NnPMA > NnPA. The endothermic enthalpy from the heating DSC curves increases in the order of NnPMA > NnPA ∼ NiPMA > NiPA. The same trends were observed in the microgels based on NiPA, NiPMA, and NnPA, which were synthesized via chemical cross-linking with N,N0 -methylenebis(acrylamide) (Bis). Although we were unable to synthesize the microgel of NnPMA due to a low water solubility of the monomer, its bulk gel was obtained by γ-ray irradiation to an aqueous poly(NnPMA) solution at a dose of 10 kGy. An irradiation-cross-linked NiPMA gel was also prepared as a counterpart to the Bis-cross-linked gel. We then studied the gel collapses upon heating by use of the chemically cross-linked gels based on NiPA, NiPMA, and NnPA as well as of the irradiation-cross-linked NnPMA and NiPMA gels. All the gels underwent the collapse transition at a certain temperature which is close to or slightly higher than the phase separation temperature of the corresponding polymer solutions or microgel dispersions. These results indicate that in both the linear and cross-linked polymers there is no difference in the thermally induced interactions between the segments as well as between the segment and the solvent, but these interactions are dependent on the structure of the constituent monomers, i.e., whether the R-carbon bears a hydrogen atom or a methyl group and whether the N-propyl group is branched or straight chain. The structure dependence was discussed in terms of amide-amide and amide-water hydrogen bondings as well as of a possible hydrogen bonding of solvent water with the H-C bond of the alkyl groups. Then, water clustering around both the alkyl and the amide groups was considered.
Introduction Many sorts of stimuli-responsive polymers in such forms as linear, grafted, and cross-linked chains have received continuous attention over the past 30 years in both fundamental and applied research.1 Among these, the polymers and gels based on N-isopropylacrylamide (NiPA) were most extensively studied due to their unique temperature-sensitive nature, i.e., an abrupt and reversible phase separation or volume change as the ambient temperature exceeds certain threshold levels, often referred to as a lower critical solution temperature (LCST) for aqueous polymer solutions and as a volume phase transition temperature (Tv) for hydrogels.2 † Part of the Gels and Fibrillar Networks: Molecular and Polymer Gels and Materials with Self-Assembled Fibrillar Networks special issue. *To whom correspondence should be addressed: e-mail kokufuta@sakura. cc.tsukuba.ac.jp; Fax 81-298-53-4605.
(1) (a) Urban, M. W., Ed. Stimuli-Responsive Polymeric Films and Coatings; Oxford University Press: New York, 2005. (b) Galaev, I. Y., Mattiasson, B., Eds. Smart Polymers for Bioseparation and Bioprocessing; Taylor & Francis Publishers: London, 2002. (2) (a) Chiantore, O.; Guaita, M.; Trossareffi, L. Makromol. Chem. 1979, 180, 969. (b) Hirokawa, Y.; Tanaka, T. J. Chem. Phys. 1984, 81, 6379. (c) Pelton, R. H.; Chibante, P. Colloids Surf. 1986, 20, 247. (d) Otake, K.; Inomata, H.; Konno, M.; Saito, S. Macromolecules 1990, 23, 283. (e) Schild, H. G.; Tirrell, D. A. J. Phys. Chem. 1990, 94, 4352. (f) Kokufuta, E.; Suzuki, H.; Sakamoto, D. Langmuir 1997, 13, 267. (g) Suzuki, H.; Kokufuta, E. Colloids Surf., A 1999, 147, 233. (h) Katsumoto, Y.; Tanaka, T.; Sato, H.; Ozaki, Y. J. Phys. Chem. A 2002, 106, 3429. (i) Kita, R.; Wiegand, S. Macromolecules 2005, 38, 4554. (j) Cheng, H.; Shen, L.; Wu, C. Macromolecules 2006, 39, 2325. Note that there are so many works on PNiPA, from which the above references were cited for clarifying the purpose of this study..
Langmuir 2009, 25(15), 8649–8655
Several of the N-substituted alkylacrylamide and methacrylamide monomers have been known to provide temperatureresponsive polymers that undergo a phase separation in water at different levels of LCST. For example, aqueous solutions of poly(N-n-propylacrylamide) (PNnPA) and poly(N-npropylmethacrylamide) (PNnPMA) exhibit the LCST ∼24 C (ref 3) and ∼28 C (ref 4), respectively. Moreover, the LCST of poly(N-isopropylmethacrylamide) (PNiPMA)5 in water is ca. 43 C, the value of which is higher by about 10 C than that of poly(NiPA) (PNiPA).2 Then, a simple but “big” question naturally arises as to why the LCST of PNiPMA or PNnPMA, both of which have an R-CH3 group on the monomer unit, is higher than that of the corresponding polymer (PNiPA or PNnPA) of N-propylacrylamides. One might suggest that a clue to answer this question lies in recent studies2h,6-9 on the phase separation (3) (a) Inomata, H.; Goto, S.; Saito, S. Macromolecules 1990, 23, 283. (b) Ito, D.; Kubota, K. Polym. J. 1999, 31, 254. (4) Ito, S. Kobunshi Ronbunshu 1989, 46, 437. (5) (a) Kubota, K.; Hamano, K.; Kuwahara, N.; Fujishige, S.; Ando, I. Polym. J. 1990, 22, 1051. (b) Tiktopulo, E. I.; Uversky, V. N.; Lushchik, V. B.; Klenin, S. I.; Bychkova, V. E.; Ptitsyn, O. B. Macromolecules 1995, 28, 7519. (c) Netopilik, M.; Bohdanecky, M.; Chytry, V.; Ulbrich, K. Macromol. Rapid Commun. 1997, 18, 107. (6) Maeda, Y.; Nakamura, T.; Ikeda, I. Macromolecules 2001, 34, 1391. (7) Maeda, Y.; Nakamura, T.; Ikeda, I. Macromolecules 2001, 34, 8246. (8) Ramon, O.; Kesselman, E.; Berkovici, R.; Cohen, Y.; Paz, Y. J. Polym. Sci., Part B: Polym. Phys. 2001, 39, 1665. (9) Paz, Y.; Kesselman, E.; Fahoum, L.; Portnaya, I.; Ramon, O. J. Polym. Sci., Part B: Polym. Phys. 2004, 42, 33.
Published on Web 03/26/2009
DOI: 10.1021/la804286j
8649
Article
of these polymers in water using Fourier transform infrared (FTIR)6,7 or attenuated total reflectance (ATR)-FTIR spectroscopy.2h,8,9 For example, ATR-FTIR measurements of PNiPA in water8,9 showed that the amide I (mostly the CdO stretch) peak in the subtraction spectra is deconvoluted into three sub-bands assignable to CdO 3 3 3 HN, CdO 3 3 3 H2O, and non-hydrogenbonded CdO, and also the amide II (mostly N-H deformation) is into the sub-bands of N-H 3 3 3 OdC, N-H 3 3 3 OH2, and nonhydrogen-bonded N-H (note that a broken line denotes a hydrogen bond). According to Paz et al.,9 there are four temperature ranges in which an increase in temperature leads to the following characteristics: (1) below 31 C, a moderate decrease in the amide-water bonds (CdO 3 3 3 H2O and N-H 3 3 3 ΟH2), a moderate increase in the free amide groups (CdO and N-H), and a little change in the population of the CdO 3 3 3 HN bonds; (2) from 32 to 35 C passing through the LCST, an abrupt increase of the CdO 3 3 3 NH bonds and the free amide groups which is accompanied by an abrupt decrease in the amide-water bonds; (3) at 36-39 C, an increase in the relative amount of the amide-water bonds; and (4) above 40 C, a moderate increase in the relative population of the free amide, coupling with a moderate decrease in the CdO 3 3 3 HN bonds as well as in the amide-water bond populations. Taking these into account, we would pay attention to the molar fraction ( fm) of the CdO 3 3 3 HN bonds at temperatures near the LCST for each set of the acrylamide and methacrylamide polymers. For this, the following data have been reported: fm = 0.13 for PNiPA6 and 0.42 for PNiPMA7 as well as fm = 0.30 for PNnPA6 and 0.40 for PNnPMA.7 One glance at this comparison suggests to us that the R-CH3 group raises the LCST due to an increase in the amount of the CdO 3 3 3 HN bonds. However, the fm value is larger in the order of PNiPMA > PNnPMA > PNnPA > PNiPA. This is different from the fact that the LCST becomes higher in the order of PNiPMA > PNiPA > PNnPMA > PNnPA. Next, let us look at the temperature effect on the hydrophobic groups; for example, the C-H stretching bands at 28403000 cm-1. It has been reported that the changes of the IR bands in this region with temperature provides information about the hydration of alkyl groups.6 For all of the four polymers, however, the IR bands assignable to the alkyl groups undergo an abrupt red shift as the temperature rises around the LCSTs.6-9 From the ATR-FTIR or FTIR studies, therefore, there is little or no evidence on which the R-CH3 group of PNiPMA or PNnPMA solely leads to an increase in the LCST of the corresponding acrylamide polymers (PNiPA and PNnPA). Although the red shift would be attributed to dehydration of the hydrophobic groups (i.e., breakage of H2O 3 3 3 H-C interactions), a redshifting of the bands assignable to the -CH2- and CH3- groups on the NiPA monomer unit could originate from an increase in their ordering or packing density (see p 41 of ref 9). In addition, there is a claim that it is not easy to discuss directly the effects of hydration using the IR spectra (see p 3434 of ref 2h). As mentioned above, the structural effect of constituent monomer units on the thermal phase separation has not yet been clear, at least in the polymer/water system to which we paid attention. It is therefore of importance to collect more “basic” data on the thermal behavior of the polymers and gels which are obtained from the two sets of the N-propylacrylamide and (10) (a) Inomata, H.; Goto, S.; Saito, S. Macromolecules 1990, 23, 283. (b) Kokufuta, E.; Tanaka, T.; Ito, S.; Hirasa, O.; Fujishige, S.; Yamauchi, A. Phase Transitions 1933, 44, 217. (c) Jin, M. R.; Wang, Y. X.; Zong, X.; Wang, S. C. Polymer 1995, 36, 221. (d) Norisuye, T.; Kida, Y.; Masui, N.; Tran-Cong-Miyata, Q.; Maekawa, Y.; Yoshida, M.; Shibayama, M. Macromolecules 2003, 36, 6202. (e) Takekawa, M.; Kokufuta, E. Colloid Polym. Sci. 2009, 287, 323.
8650 DOI: 10.1021/la804286j
Kano and Kokufuta
methacrylamide monomers. The temperature dependence of volume change in gels provides a macroscopic view of the behavior of the constituent polymers, so that with the use of the gel we may further explore how the R-CH3 and N-propyl groups affect the thermally induced interactions. Several studies have been performed on NnPA gels,10 but there are a very few reports on NiPMA-based microgels11 and macrogels.12,13 In ref 13, the effect of the R-CH3 group on the volume phase transition was discussed by use of ultrasmall-angle X-ray scattering data obtained at different temperatures (5 C interval); however, the temperature dependence of the swelling degree12 was not considered as well. As far as we know, there is no report on NnPMA gels, even though in 1997 a patent14 hinted the potential use of “aqueous gel” of NnPMA. Here we systematically studied the thermal behavior in water of the polymers, microgels, and macrogels (i.e., fine cylindrical bulk gels) based on the NiPA, NiPMA, NnPA, or NnPMA monomer to promote a discussion on the thermal phase separation and volume transition in terms of the dependence of constituent monomer structure. A gel based on NnPMA was prepared by γ-ray irradiation to an aqueous solution of the polymer (PNnPMA) because this monomer exhibited a very low solubility in water. The experiments were performed by a combination of turbidity measurements and differential scanning calorimetry (DSC). Also performed were the size measurements using the dynamic light scattering (DLS) method for microgels and the microscopic method for macrogels.
Experimental Section Preparation of Monomers. All the monomers, other than PNiPA (kindly supplied by Kojin Chemical Co., Tokyo, Japan), were synthesized from the corresponding acryloyl chlorides and propylamines according to the following chemical reaction: H2 CdCðR1 ÞCOOCl þ H2 NR2 þ ðC2 H5 Þ3 Nf H2 CdCðR1 Þ-COðNHR2 Þ þ ðC2 H5 Þ3 N 3HCl where R1 = CH3 and R2 = CH(CH3)2 for NiPMA, R1 = H and R2 = (CH2)2CH3 for NnPA, and R1 = CH3 and R2 = (CH2)2CH3 for NnPMA. The desired acryloyl chloride was coupled with the corresponding amine in a benzene solution containing triethylamine (dehydrochlorinator). The coupling reaction was performed by slowly dropping the acryloyl chloride into the benzene solution at ca. 5 C for 3 h under stirring; then the mixture was left overnight in a refrigerator. The resulting two monomers (NnPA and NnPMA) were purified by vacuum distillation, and the others (NiPA and NiPMA) were recrystallized three times from a benzene/hexane mixture. The purity and chemical structure of each monomer were confirmed by different methods: elemental analysis, IR spectroscopy, and 1H NMR spectroscopy. The NnPA and NnPMA monomers obtained were a liquid at room temperature: bp = 94 C (3 mmHg) for NnPA and 96 C (1 mmHg) for NnPMA. The NnPMA monomer has a very low solubility in water (5-95 C), although its polymer (PNnPMA) was water-soluble at temperatures lower than the LCST. The other two monomers (NiPA and NiPMA) were a solid: mp = 64.5 ( 0.5 C for NiPA and 87.8 ( 0.3 C for NiPMA. :: (11) (a) Duracher, D.; Elaissari, A.; Pichot, C. J. Polym. Sci., Part A: Polym. Phys. 1999, 37, 1823. (b) Guillermo, A.; Addad, J. P. C.; Bazile, J. P.; Duracher, D.; :: Elaissari, A.; Pichot, C. J. Polym. Sci., Part B: Polym. Phys. 2000, 38, 889. (12) Fomenko, A.; Pospısil, H.; Sedlakova, Z.; Plestil, J.; Ilavsky, M. Phys. Chem. Chem. Phys. 2002, 4, 4360. (13) Tirumala, V. R.; Ilavsky, J.; Ilavsky, M. J. Chem. Phys. 2006, 124, 234911. (14) “Hardenable composition, aqueous gel and applications”. US Patent 5624998, , issued on April 29, 1997.
Langmuir 2009, 25(15), 8649–8655
Kano and Kokufuta Preparation of Polymers. The preparation of PNiPMA and PNnPMA was performed by free radical polymerization using methanol as a solvent and R,R’-azobis(isobutyronitrile) as an initiator. The monomer solution was filled in glass ampules, frozen with liquid nitrogen,and flame-sealed under vacuum. The polymerization was performed in a water bath at 60 C for 12 h. The resulting reaction mixture was slowly poured into n-hexane to precipitate the polymer, which was then separated by filtration and dried under vacuum. Each crude polymer was reprecipitated three times from benzene with n-hexane and then dried under vacuum. Further purification was carried out by dialyzing an aqueous polymer solution against distilled water for a week. We then employed a cellulose ester dialysis membrane (Spectra/ Por CE) having a molecular weight cutoff (MWCO) of 100 000. The dialyzed solution was lyophilized and finally dried under reduced pressure at 50 C for 3 days. The weight-average molecular weight (M w) was determined by static light scattering (SLS): 4.20 105 for PNiPMA and 6.02 105 for PNnPMA. The other two polymers (PNiPA with M w = 5.53 105 and PNnPA with M w = 3.61 105) were the stocks, both of which had been prepared in our laboratory and characterized by SLS.10e Preparation of Microgels. The microgels based on NiPA, NiPMA, or NnPA were prepared via an aqueous redox polymerization using ammonium persulfate (APS) as an initiator. To obtain the gel particles with diameters in the range of several hundreds of nanometers, the monomer was cross-linked with N,N0 -methylenebis(acrylamide) (Bis) in the presence of sodium dodecylbenzenesulfonate (NaDBS) as a surfactant and at temperatures above the LCST (e.g., 60 C for NiPA). All the chemicals, other than the monomers, were commercially obtained from Wako Pure Chemical Co., Osaka, Japan. The preparation procedures are the same as in our previous studies (e.g., see ref 2g). The following is the outline: A well-degassed pregel solution (monomer ∼0.078 mol, Bis ∼0.000 78 mol, NaDBS ∼0.0036 mol, water ∼200 g) was placed in the usual separable flask (500 mL) equipped with a cooler and a magnetic stirrer. Nitrogen gas was continuously supplied above the surface of the solution for 1 h before the reaction to remove oxygen well. The polymerization reaction was initiated by adding an aqueous O2-free solution (10 mL) containing 50 mg of APS, allowed to continue several hours, and terminated by blowing oxygen through the reactor. After that, residual monomers and NaDBS were removed from the resulting reaction mixture by a dialysis method. Further purification was carried out by passing the dialyzed solution through a mixed bed of anion and cation exchange resins. The purified microgel dispersion were then lyophilized for several days. Preparation of Macrogels. An aqueous solution (9.5 mL) containing the monomer (ca. 7 mmol), Bis (0.091 mmol), and N,N,N0 ,N0 -tetramethylethylenediamine (TMED) (94 μL) was prepared by dissolving these chemicals in distilled water, degassed with N2 for 10 min in an ice bath, and finally mixed with 0.5 mL of an aqueous O2-free solution including 10 mg of APS. After that, 3 mL of the above solution was quickly transferred into a test tube into which glass capillaries with an inner diameter of 0.3 mm had previously been inserted. The polymerization was carried out under an N2 atmosphere in an ice bath. After the gelation was completed, the gels were taken out of the capillaries and thoroughly washed with pure water. All the samples were cut into cylinders of ∼2 mm length and stored at 3 C before use. A gel consisting of the NnPMA monomer unit was prepared by γ-ray irradiation at a dose of 10 kGy because of a low water solubility of the monomer. An aqueous solution (3 mL) of PNnPMA whose concentration (in unit mole base) is equivalent to that of each monomer used in the chemical cross-linking was filled in a glass ampule with inserted glass capillaries (the same as above) and degassed well under a vacuum before flamesealing. The gelation was carried out in an ice bath at a dose rate of 1.56 kGy/h for 6.41 h using γ-rays from a 60Co source. Langmuir 2009, 25(15), 8649–8655
Article As a counterpart to the Bis-cross-linked NiPMA gel, a gel sample was also prepared from PNiPMA by the γ-ray irradiation at the same condition as above. Measurements. A temperature (Tc) at which a phase separation takes place in an aqueous polymer solution or in an aqueous microgel dispersion was determined by the measurement of transmittance at 500 nm using a Hitachi spectrophotometer (model U-2001). The temperature of the cell was regulated to (0.1 C by a circulating water bath. DSC measurements were performed using a Seiko DSC 6100 calorimeter (Seiko Instruments Inc., Chiba, Japan) with a scanning rate of 0.30 C/min. The polymer solutions or microgel dispersions (65.00 ( 0.02 mg) with different concentrations were exactly weighted by a PerkinElmer AD-6 Ultra Microbalance and subjected to the measurement. The same amount (within the error range of (0.02 mg) of the solvent (pure water) was used as the reference. DLS measurements for microgels were carried out at a scattering angle (θ) of 90 using an Otsuka DLS 7000 apparatus (Osaka, Japan) equipped with a 10 mW He-Ne laser as the light source and with a water-thermostated cell holder. The average hydrodynamic radius (ÆRhæ) of a gel particle was estimated as a function of temperature by analyzing the autocorrelation functions with the CONTIN program (for more details, see ref 15 and references therein). The diameter of macrogels (cylindrical in shape) was measured as a function of temperature by use of our own setup.16 A section of the gel sample was inserted into a water-jacketed microcell together with distilled water. The gel was then observed using a microscope with a calibrated scale, from which the images were recorded digitally on videotape as a function of time. The temperature was controlled within a range of (0.05 C with water circulating around the measuring cell (for more details, see ref 16).
Results and Discussion Polymer/Water Systems. Figure 1 shows the aspects of phase separation observed in the four polymer/water systems by measurements of transmittance (T%) as a function of temperature. An abrupt decrease in T% of each polymer solution on heating indicates that all of the polymers in water undergo a thermally induced phase separation. The Tc becomes higher in the order of PNiPMA > PNiPA > PNnPMA > PNnPA. As mentioned in the Introduction, these results have been understood in terms of a “coil-to-globular” transition of the polymers under investigation. This transition temperature was often referred to as LCST; however, we hereafter expressed as Tc because we deal with not only the polymer but also the microand macrogel. Also, we note here that the temperature change of T% indicates the overall aggregation process of the polymer chains due to both intrapolymer and interpolymer interactions; therefore, the Tc is not solely due to the association among fully collapsed globular chains. In other words, there is a large possibility that several chains under “not” fully collapsed states aggregate with each other. This would be supported from a recent study2j on a very dilute water solution of PNiPA by use of DLS. Thus, we must consider this matter when looking at the results of DSC measurements. We studied the thermal phase separations by DSC. As can be seen from Figure 2, the four polymer/water systems exhibit the endothermic transition upon heating. The onset and peak temperatures as well as the calorimetric enthalpy (ΔH as a peak (15) Kokufuta, E.; Ogawa, K.; Doi, R.; Farinato, R. S. J. Phys. Chem. B 2007, 111, 8634. (16) Kokufuta, E.; Suzuki, H.; Yoshida, R.; Yamada, K.; Hirata, M.; Kaneko, F. Langmuir 1998, 14, 788.
DOI: 10.1021/la804286j
8651
Article
Kano and Kokufuta
Figure 1. Temperature changes of transmittance (T) for aqueous solutions (0.02 mg/mL) of four polymers shown in the figure. Figure 3. Polymer concentration dependence of endothermic enthalpy (Δ H) obtained from the heating DSC measurements. The kind of polymers is shown in each figure. Table 1. Results for Transmittance and DSC Measurements of Aqueous Polymer Solutions Tc (C) sample
Tca
Tc1b
Tc2b
ΔHc (kJ/mol)
PNiPA PNiPMA PNnPA PNnPMA
31.2 41.2 23.2 27.2
30.9 41.8 23.0 26.9
32.1 42.0 24.5 28.0
5.01 5.53 5.67 7.80
a Determined from the onset of the temperature vs T% curves. Shows the averaged values which were determined from the onset (Tc1) and peak (Tc2) of DSC curves at different concentrations. c Obtained from Figure 3. b
Figure 2. Typical heating DSC thermograms for aqueous solutions of four polymers shown in the figure. Polymer concentrations in unit mole per gram of sample solution are 2.10 10-4 for PNiPA, 1.81 10-4 for PNiPMA, 1.71 10-4 for PNnPA, and 1.70 10-4 for PNnPMA.
area) were obtained for each sample by considering the effect of polymer concentrations. Figure 3 shows the change of ΔH with the moles of polymer repeating units per gram of the sample solution. A good linear line passing through the origin (correlation coefficient >0.98) can be obtained at the concentrations below 0.6 mmol (∼68 mg) per gram of the aqueous PNiPA or PNnPA solution and below 0.4 mmol (∼51 mg) per gram of the aqueous PNiPMA or PNnPMA solution. Although it is not clear whether or not this difference depends on the structure of the constituent monomer units, the slope of the linear range allows us to estimate the ΔH value which is independent of the polymer concentration. We did not measure the calorimetric change on cooling because this process is dominated by the dissolution of “aggregated” polymers (e.g., see ref 2j), but not solely by the conformation transition from the globular to the coiled state of the polymer. Summarized in Table 1 are the results of the DSC and T% measurements. A difference of the Tc values from both methods falls within 1.1 C at the minimum and 1.5 C at the maximum. There are differences of 3-4 C in the Tc values which have been reported using the “same sort” of temperature-responsible polymers in water: for example, in the case of PNiPA, Tc = 32.4-35.5 C on the endothermic maximum of DCS curves and 32.2-35.7 on the onset of T% changes, using the polymer samples having M w ∼ 1.4 104-5.3 105 (see ref 2e); 8652 DOI: 10.1021/la804286j
ca. 33 C by turbidity measurements;6 and 30.9-35.2 C by visual observation and 31.7-33.9 C on the onset of heating DCS curves (see ref 2d). Also, in the ΔH values for the PNiPA/ water system, there are some differences among these previous studies: 4.6-7.1 kJ/mol (ref 2e), 5.8 kJ/mol (ref 6), and 4.8-6.1 kJ/mol (ref 2d). In ref 2e the difference of Tc as well as of ΔH was discussed in terms of the effect of the chain length on the transition. However, we believe that the differences in the Tc and ΔH data are dependent on the nature of a phase separation of this sort. In other words, neither Tc nor ΔH (obtained here and also previously) is an intrinsic characteristic of the coilto-globular transition of thermo-sensitive polymers; rather, both quantities display the aspect of an aggregation process in which all the polymer chains do not need to become a completely collapsed (globular) state. Even when taking these into account, however, we may draw the following conclusions from the results in Table 1: (a) the substitution of the R-hydrogen on the NiPA or NnPA monomer unit by a CH3 group leads to an increase of its Tc and ΔH, and (b) the branching of the N-npropyl group on the NnPA or NnPMA unit raises the Tc but lowers the ΔH. These results cannot be understood by considering the temperature changes in the amount of the CdO 3 3 3 HN bonds (compare the fm values as mentioned in the Introduction). In addition, it seems that the free rotation of R-methyl and/or N-propyl groups is not the sole case because a “configuration” change of several monomer units in the main chain has been suggested in the case of PNiPA (see ref 2h). There has been a reasonable (or plausible) explanation for why both Tc and ΔH values of PNnPA are different from those Langmuir 2009, 25(15), 8649–8655
Kano and Kokufuta
Article
Figure 5. Microgel concentration dependence of endothermic enthalpy (ΔH) obtained from the heating DSC measurements. The kind of microgels is shown in each figure. Table 2. Results for Transmittance and DSC Measurements on Aqueous Dispersions of Bis-Cross-Linked Microgels Based on Three Monomers Tca (C)
Figure 4. Typical heating DSC thermograms for aqueous dispersions of three microgels shown in the figure. The concentrations in unit mole per gram of sample dispersion are 2.41 10-4 for NiPA microgel, 2.00 10-4 for NiPMA microgel, and 2.52 10-4 for NnPA microgel. Note that the amount of Bis in the microgel networks is neglected in the calculation of the concentration. Our preliminary experiments have showed that an error from this approximation is less than 3%.
of PNiPA.6 Then, the authors assumed that a difference in the solvent accessible area between alkyl groups of amino acid residues in globular proteins17 would be applicable to both polymers in water. In their next paper7 dealing with PNnPMA and PNiPMA, however, the differences between PNnPA and PNnPMA, as well as between PNiPA and PNiPMA, were not discussed in terms of the solvent-accessible area. Now it becomes clear that the systematic work on the gels of these polymers is of importance to look at the effects of the R-methyl group and the branching of the N-n-propyl group. Microgel Systems. We prepared microgels from the NnPA, NiPA, or NiPMA monomer by cross-linking with Bis. Their aqueous dispersions were then subjected to the T% and DSC measurements as made in the polymer/water systems. Figure 4 shows typical DSC thermograms obtained in the microgel systems. Each endothermic peak upon heating becomes wider as compared with the case of the corresponding polymers; however, the concentration dependence of ΔH shows a good linear line (see Figure 5). The onset and peak temperatures (denoted as Tc1 and Tc2, respectively) were determinable by the DSC, although there was a slight fall in the precision due to the use of an interpolation technique. Table 2 shows the results of the T% and DSC measurements. The phase separation upon heating has been reported in aqueous dispersions of the microgels on which we are focusing: Tc ∼ 32 C for NiPA microgel2c,18 and Tc ∼ 44 C for NiPMA microgel.11 Although we were unable to find the data for the other two microgels, the Tc values in Table 2 were little different from those reported previously. Let us now compare the results in Tables 1 and 2. Then, we can see the following important facts: (a) there is little difference in the Tc values between the polymer and microgel systems, and (b) the monomer species dependence of ΔH of the microgels shows the same trend as that of the polymers, but (c) the ΔH of each microgel is about 75% that of the corresponding polymer. Prior to discussing the above results, we should note that microgel particles have close to a spherical shape but that (17) Rose, G. R.; Geselowitz, A. R.; Leser, G. J.; Lee, R. H.; Zehfus, M. H. Science 1985, 229, 834. (18) Pelton, R. Adv. Colloid Interface Sci. 2000, 85, 1.
Langmuir 2009, 25(15), 8649–8655
monomer
Tc1
Tc1
Tc2
ΔHb (kJ/mol)
NiPA NiPMA NnPA
31.4 41.1 23.5
31.2 41.6 23.2
32.6 44.8 25.1
3.83 4.19 4.24
a
Determined in the same way as in Table 1. b Obtained from Figure 5.
attached dangling chains cover their surface (e.g., see Figure 4 of ref 15); therefore, the microgel lies between the polymer and the macrogel. There are two types of interactions, i.e., “intra”and “inter”-particle interactions. The former becomes dominant in the swelling-shrinking change of a gel particle, and the latter is in the association among the particles. Taking these into account, from result (a) we may suggest that the Tc of each microgel is mainly determined by the thermally induced “interparticle” interaction because of little difference in Tc between the polymer and microgel systems. In contrast, with a consideration of the “intraparticle” interaction within a microgel or its aggregate, we may understand result (c). This is mainly because the cross-linking with Bis could reduce the number of segmental contact points in the polymer chain. It has been reported that the ΔH of Bis-cross-linked NiPA macrogels is lower than that of PNiPA,2d although we did not perform the DCS of macrogels due to a lack in the reproducibility in our experimental procedures. Next we pay attention to result (b), from which it may be suggested that the effects of the R-methyl and the N-propyl group on the phase separation in the polymer/water system can be taken as the interactions based on the segment. In our discussion on the gels (cross-linked polymers), this should be in preference to the coil-to-globular transition in a single polymer chain. To make more clear this point, we studied the temperature dependence of particle size using the three microgel systems. In the DLS measurements at the minimum concentration (ca. 10 mg/mL) used in the DSC, the analysis of intensityintensity time correlation function by the CONTIN showed a bimodal or trimodal pattern at temperatures close to the Tc in Table 1 or 2, indicating that the gel particles aggregate with each other. We thus performed the DLS on a dilute microgel dispersion (0.1 mg/mL) in the following ways: At first, the hydrodynamic radius (ÆRhæ0) of a gel particle was determined at a temperature that is lower by ca. 5 C than its Tc; after that, the radii (ÆRhæT) were measured as a function of temperature until the observation of the aggregation. Figure 6 shows the plots of ÆRhæT/ÆRhæ0 vs temperature. It is found that the gel collapse due to intraparticle interactions begins at the temperature close to the Tc of each microgel. This indicates that there is little difference in the temperatures at which the thermally induced DOI: 10.1021/la804286j
8653
Article
Figure 6. Temperature changes of normalized equilibrium hydrodynamic radius (ÆRhæT/ÆRhæ0) for aqueous dispersions (0.01 mg/mL) of microgels. The values of ÆRhæ0 are 96 nm at 19 C for NnPA microgel, 132 nm at 26 C for NiPA microgel, and 106 nm at 36 C for NiPMA microgel. Each upward arrow in the figure shows the temperature at which a bimodal pattern was observed by the CONTIN analysis; therefore, the plotted data were calculated from the smallest ÆRhæT.
intra- and interparticle interactions occur in aqueous microgels based on the same monomer. Macrogel Systems. We prepared a NnPMA macrogel using γ-rays to provide a full comparison between the polymer and gel. The diameter (dT) of a fine gel cylinder was microscopically determined as a function of temperature. The swelling degree (dT/d0) was then calculated by normalizing with a gel diameter (d0) in a fully collapsed state. Figure 7 shows the plots of dT/d0 vs temperature for the gels obtained from all the four monomers under investigation. Also shown as a control is the result for an NiPMA gel obtained by γ-ray irradiation. An abrupt volume collapse upon heating was observed at a temperature (Tv) close to, or slightly higher than, the Tc in both Tables 1 and 2. This is a strong evidence for our claim that the thermally induced interactions between segments as well as between segment and water is affected by the structure of the constituent monomers. One might care for whether each swelling curve falls in the category of a continuous or discontinuous volume-phase transition in the gel. Our experiments showed the latter for all the gels studied (see broken lines in Figure 7 which were determined in a similar manner as reported in ref 16). In contrast to our NiPMA gel, a continuous transition has been reported by Fomenko et al.,12 who prepared Bis-cross-linked NiPMA gels having a different amount of sodium methacrylate in an “ethanol/water” (8/0.5 by volume) mixture. Interestingly, a pure NiPMA gel in their preparations undergoes an “abrupt” volume change at a temperature which agrees well with the Tv of our gel, but there is a certain difference in the swelling degree between both gels. For such differences, we note that an increase in the amount of cross-linker (e.g., Bis) alters a discontinuous transition in the gel to a continuous one, accompanied by an overall decrease in the gel volume. Also note that the radiation crosslinking of an aqueous polymer solution using γ-rays affects slightly both the Tc and gel volume but extremely the shrinking kinetics through an alteration in the static inhomogeneities frozen in the gel.10e The inhomogeneities of this sort are directly related to the phase equilibrium properties of the gel during the gelation process (see the Introduction of ref 10c). For the gels based on the same monomer but prepared in different ways, therefore, the discussion on whether the transition is continuous 8654 DOI: 10.1021/la804286j
Kano and Kokufuta
Figure 7. Temperature changes of normalized equilibrium diameter (d/d0) for bulk gels. “CC” and “RC” in parentheses denote chemical cross-linking with Bis and radiation cross-linking with γ-rays, respectively. The values of Tv indicated by broken lines are 23.7 C for NnPA gel (CC), 28.4 C for NnPMA gel (RC), 33.9 for NiPA gel (CC), 44.5 C for NiPMA gel (RC), and 45.4 C for NiPMA gel (CC).
or discontinuous is not of the essence, at least in the this paper. This is the reason why we used the radiation cross-linked NiPMA gel as a counterpart to the Bis-cross-linked gel. We must consider several factors influencing the volume of gels at or around Tv; e.g., the cross-linking degree as mentioned above. Taking this into account, let us look at the changes of dT/d0 around which each gel undergoes an abrupt volume collapse upon heating, for example, the dT/d0 change at temperature of 44-46 C for the NiPMA gel. The magnitude of such a change seemed to become larger in the order of NnPA > NnPMA > NiPA > NiPMA, suggesting that the differences of R1 (H or CH3) and R2 (n-propyl or isopropyl) on the constituent monomers -CH2-CR1(CONHR2)- are a key factor which determines not only the transition temperature but also the volume at the transition.
General Discussion and Conclusions For the understanding of the phase separation as well as the volume phase transition reported here, it would be reasonable to consider the segment-based interactions which are separable in the following three types: water/amide (W/A), amide/amide (A/A), and water/hydrocarbon (W/Hc). Here, both the W/A and A/A types of interactions are based on hydrogen bonding. The W/Hc type means the interactions of water with all the C-H bonds of the segment which are a combination of either -CH2-CH= or -CH2-C(CH3)= against either of CH3(CH2)2- or (CH3)2CH-. Ab initio calculations for a system consisting of the proton donor such as fluorinated hydrocarbons and the proton acceptor such as H2O and CH3OH have shown that the CH 3 3 3 O interaction behaves very much like a conventional OH 3 3 3 O bond.19 At a more macroscopic level, we have to focus on the hydrophobic hydration of nonpolar groups in water, through which water molecules trend to form shell-like configurations.20 Thus, the W/Hc interactions should be taken at both microscopic (chemical bonding) and macroscopic (water structuring) levels. (19) Gu, Y.; Kar, T.; Scheiner, S. J. Am. Chem. Soc. 1999, 121, 9411. (20) (a) Dixit, S.; Crain, J.; Poon, W. C. K.; Finney, J. L.; Soper, A. K. Nature (London) 2002, 416, 829. (b) Bowron, D. T.; Finney, J. L. J. Phys. Chem. B 2007, 23, 9838.
Langmuir 2009, 25(15), 8649–8655
Kano and Kokufuta
Now let us consider a heating process at temperatures below a Tc or Tv value, during which the W/A and W/Hc interactions would be fully eliminated in a “moderate” amount, but not all, of the segments. When this is the case, the transition is triggered by minimizing both W/A and W/Hc interactions among the segments. This minimizing process would cause a “suitable” degree of the A/A interaction, accompanied by the formation of the free amide.21 At the same time, the W/Hc interactions would alter due to a structural change in clustered water molecules within the hydration shells. As a result, the segments trend to associate each other,22 causing the phase separation and volume phase transition. Then, the discussion is centered on how the structural differences on the R-carbon and N-propyl group of the monomer unit affect these interactions, in particular in the minimizing process of the W/A and W/Hc interactions. Our experimental results may be summarized as in Figure 8, when we consider the dependencies of Tc and Tv on the monomer structure. It is found that the substitution of the R hydrogen with a methyl group raises Tc and Tv as well as ΔH, while the branch of the N-n-propyl group leads to an increase of both Tc and Tv but a decrease of ΔH. The W/A and W/Hc interactions originated from the hydrogen bonding are weakened by heating. Hence, the results in Figure 8 mean that the substituted and/or branched amide monomer units form “hydrated” segments with a high thermal stability against their association at the Tc or Tv, so as to increase both thresholds. Then, assuming that the ΔH of the transition would be consumed to break a hydrated structure, one considered that the branch of the N-n-propyl group decreases the ΔH due to a decrease in the solvent accessible area of alkyl chains.6 However, this discussion was made relying on the data17 of amino acid residues in proteins. Therefore, we hesitate over which data to choose when attempting to discuss the effect of the substitution on the ΔH. It cannot simply be said that the substitution of the R hydrogen with a nonpolar methyl group enhances the hydrophobic effect because amides (RC(dO)NHR0 in our case) have the electron-withdrawing nature of the carbonyl group where the lone pair of electrons on the nitrogen is delocalized by resonance, thus forming a partial double bond with the carbonyl carbon and putting a negative charge on the oxygen.23 Since this nature varies depending on the kinds of R and R0 , the A/W interaction should be affected by whether the R-carbon bears a hydrogen atom or a methyl group and whether the N-propyl group is branched or straight chain. Very recently, Bowron and Finney20b reported the three-dimensional spatial density functions (3DSDF) showing the preferred hydration regions of amphiphile molecules (0.04 mole fraction solution of tert-butanol in water) as a function of temperature. From their results (Figure 13 of ref 20b), we can see three different images at each temperature (25, 80, or 106 C) and at different distances from the central carbon: nearest-neighboring small shell around the polar OH group at distances 3-4 A˚, first-neighboring nonpolar hydration shell of the alcohol molecules at 4-6 A˚, and second-neighboring hydration shell of the alcohol molecules at 6.5-9.5 A˚. Then, it is imagined that the clusters formed in (21) Note that a part of the resulting free amide could form the CdO 3 3 3 HN bonds, but its considerable amount is remaining even at temperatures higher than the Tc (e.g., see Figure 6 of ref 9). As far as we know, there is no explanation for this fact. However, it is reasonable to consider that these -CONH- groups are surrounded by nonpolar alkyl groups, thereby not accessible with both water and the other free amide. (22) A recent study (ref 20b) using isotope substitution neutron scattering difference measurements on a dilute aqueous solution of (CH3)3COH has suggested that it could be misleading to think of the aggregation of aqueous amphiphiles solely in terms of a hydrophobic driving force. (23) For example, see: Kemnitz, C. R.; Loewen, M. J. J. Am. Chem. Soc. 2007, 2521.
Langmuir 2009, 25(15), 8649–8655
Article
Figure 8. Systematic comparison of the experimental results. Tc denotes the average of phase separation temperatures for each polymer, and Tv is that of volume-phase transition temperatures for each of both micro- and macrogels. ΔH shows transition enthalpies with the subscripts p for polymers and g for microgels.
amide/water systems are more complicated than that of (CH3)3COH, because the polar C(dO)NH group is between two nonpolar tails (R and R0 ), the kind of which affects the situation of the nearest-neighboring shell around the C(dO)NH group and also of the other two shells. Without relying on such 3DSDF data, on the other hand, one might say that the reason why PNnPMA has the largest ΔH of the four is due to formation of very stable hydration shells around the segment because (i) the aqueous solubility of NnPMA (monomer) is very low and (ii) such a compound could be stabilized by increased clustering of solvent water molecules. These consequences are, however, available for the other three because all of them have a good water solubility. A systematic comparison of our data (Figure 8) and a discussion by considering the newest studies on the hydration20b of amphiphiles in water as well as on the electronic structure23 of amide bonds indicate a way in which what we need in future to really make a good understanding of the molecular mechanism for the phase separation of poly(N-alkylacrylamide)s and poly(N-alkylmethacrylamide)s and the volume-phase transition of their hydrogels. Such calculations as ab initio and in silico methods for model compounds based on the structure CH3-CH (R1)(CONHR2) would provide a good information about localized and delocalized electron states over the molecule. In addition, when the association-dissociation behavior of such model compounds (amphiphilic solutes) in water would be studied using the techniques as used in ref 20, we could have a goal for discussion. In conclusion, it has become apparent that all four monomers with a structural formula of CH2dC(R1)(CONHR2), where R1 is H or CH3 and R2 is n-propyl or isopropyl, provide the temperature-responsive polymers and gels. The phase separation of an aqueous polymer solution shows the same aspect as the volume phase transition of a hydrogel when both preparations are based on the same monomer. However, a difference in either or both of R1 and R2 draws a different picture in both the phase separation and the volume transition. This knowledge would be helpful for further developments in the studies of water-soluble or accessible materials based on various N-substituted acrylamide derivatives in both fundamental and applied fields. Acknowledgment. This work was supported in part by Grants-in-Aid for Scientific Research to E.K. from the Ministry of Education of Japan (No. 09875232) and the Japan Society for the Promotion Science (No. 20550183) as well as from University of Tsukuba. DOI: 10.1021/la804286j
8655
