pubs.acs.org/Langmuir © 2010 American Chemical Society
Resins with “Nano-Raisins” S. Sinha-Ray,† Y. Zhang,† D. Placke,† C. M. Megaridis,† and A. L. Yarin*,†,‡ †
Department of Mechanical and Industrial Engineering, University of Illinois at Chicago, 842 W. Taylor Street, Chicago, Illinois 60607-7022, and ‡Center for Smart Interfaces, Technische Universit€ at Darmstadt, Petersenstrasse 32, 64287 Darmstadt, Germany Received January 28, 2010
Thermosensitive hydrogels are materials which globally shrink/swell in water when the surrounding temperature crosses the lower critical solution temperature (LCST). We demonstrate here a novel class of cross-linked polymeric materials, which do not shrink/swell in water globally, but nevertheless reveal a hydrogel-like, stimuli-responsive behavior. In particular, they demonstate a positive thermosensitive release of the embedded fluorescent dye significantly modulated when temperature crosses the LCST. Using staining with copper, transmission electron microscopy and energy dispersive X-ray analysis, we show that this effect is associated with nanogel “raisins” dispersed in such materials (e.g., polymer nanofibers). Shrinkage of individual nanogel “raisins” at elevated temperatures increases nanoporosity via increased exposure of the existing nanopores to water, or formation of new nanopores/nanocracks in the overstretched polymer matrix in the vicinity of shrinking nanogel “raisins”. As a result, the release rate of the embedded dye from the nanofibers increases at elevated temperatures. We suggest that similar functional materials with embedded nanogel “raisins” will find applications in nanofluidics and as drug carriers for controlled drug release.
1. Introduction Polymeric materials that are responsive to external stimuli are considered to be promising candidates for micro- and nanofluidics and as new drug carriers.1 In this class of materials, stimuliresponsive hydrogels have attracted the most attention due to their adjustable-on-demand properties.2 These materials might be used to eliminate many undesirable characteristics of traditional drug carriers and introduce new, beneficial traits, e.g., capability of pulsatile, site-specific and/or externally triggered release.3 Hydrogels are polymer networks which can take up water causing them to swell dramatically. On the other hand, they can expel water when shrinking. If hydrogel swelling/shrinkage in water can also be controlled by an external stimulus, such as temperature or pH, the term smart hydrogel is adopted. Poly(N-isopropylacrylamide) (PNIPAM) is one of the most extensively studied synthetic polymers forming smart hydrogels. When temperature is increased above its lower critical solution temperature (LCST) of about 32 °C, PNIPAM undergoes a sharp transition from a hydrophilic to a hydrophobic state.4 Typically, a hydrogel containing PNIPAM is in a swollen state in the presence of water at room temperature but shrinks and expels water when temperature rises above the LCST. This behavior is attributed to the disruption of hydrogen bonding between water molecules and the amide side groups of PNIPAM macromolecules. The process is fully reversible, so the hydrogel swells and takes water again when temperature drops below the LCST. In ref 5, PNIPAMbased hydrogel nanofibers were first electrospun6-8 from a blend of high-molecular weight (Mw = 79-203 kDa) PNIPAM, *To whom correspondence should be addressed. E-mail:
[email protected]. Telephone: (312) 996-3472. Fax: (312) 413-0447.
(1) Schmaljohann, D. Adv. Drug Delivery Rev. 2006, 58, 1655. (2) Hoare, T. R.; Kohane, D. S. Polymer 2008, 49, 1993. (3) Sinha-Ray, S.; Yarin, A. L. J. Appl. Phys. 2010, 107, 0294903. (4) Schild, H. G. Prog. Polym. Sci. 1992, 17, 163. (5) Chen, H.; Hsieh, Y.-L. J. Polym. Sci., A: Polym. Chem 2004, 42, 6331. (6) Reneker, D. H.; Yarin, A. L.; Zussman, E.; Xu, H. Adv. Appl. Mech. 2007, 41, 43. (7) Reneker, D. H.; Yarin, A. L. Polymer 2008, 49, 2387. (8) Agarwal, S.; Greiner, A.; Wendorff, J. H. Adv. Funct. Mater. 2009, 19, 2863.
Langmuir 2010, 26(12), 10243–10249
poly(vinyl alcohol) (PVA) and poly(acrylic acid) (PAA). At the second step, the polymers were thermally cross-linked to form a hydrogel. When raising the temperature from 25 to 70 °C, the crosslinked hydrogel fiber mats shrank to one-fourth of their original size within a few seconds, indicating global shrinkage/swelling behavior that is commonly observed for PNIPAM-based hydrogels.5 Moreover, recently developed poly(methyl methacrylate)/N-isopropylacrylamide (PMMA-NIPAM) copolymers, which are practically water-insoluble and possess tunable LCST up to 52 °C, demonstrated shrinkage ratio in water in the range 70-90%.9 Electrospinning is a unique straightforward process to produce polymer nanofibers of the order of several hundreds of nanometers in diameter.6-8 Among many possible applications of electrospun nanofibers that are being investigated is their use as drug delivery systems. A possible advantage of nanofibers as drug carriers is that the smaller the dimensions of the drug carrier, the better the drug can be absorbed by the human body. Also, nanofiber mats containing, for example, camptothecin (an anticancer drug) are very flexible and could be set into spaces near a brain tumor without damaging the surrounding healthy tissue. Modifying the release rate from PNIPAM-containing nanofibers on demand is not only a goal but also a tool to probe the intrinsic physical mechanisms of the process. In particular, the theoretical model of ref 10 assumed that PNIPAM is distributed as nanogel islands in a polymer matrix. At T > LCST, these nanogel islands shrink and stretch the surrounding polymer matrix, which creates new nanopores/nanocracks in it. As a result, drug release rate increases, even though PNIPAM shrinks. The elucidation of the intrinsic physical mechanism of PNIPAM swelling/shrinkage in nanofibers submerged in water has immense practical consequences as well. Recently it was shown11,12 that solid-state diffusion can hardly be the mechanism responsible (9) Zhang, Y.; Yarin, A. L. J. Mater. Chem. 2009, 19, 4732. (10) Yarin, A. L. Math. Modell. Nat. Phenom. 2008, 3, 1. (11) Srikar, R.; Yarin, A. L.; Megaridis, C. M.; Bazilevsky, A. V.; Kelley, E. Langmuir 2008, 24, 965. (12) Gandhi, M.; Srikar, R.; Yarin, A. L.; Megaridis, C. M.; Gemeinhart, R. A. Mol. Pharm. 2009, 6, 641.
Published on Web 02/12/2010
DOI: 10.1021/la1004177
10243
Article
Sinha-Ray et al.
for dye/drug release from nanofibers, while dye/drug desorption from nanopore surfaces plays the main role. This poses a crucial question whether PNIPAM “raisins” can indeed be distributed in nanofibers and create new nanopores without global shrinkage/ swelling of nanofiber mats?
2. Experimental Methods and Setup 2.1. Materials. Poly(acrylic acid), PAA (Mw = 450 kDa), Poly(vinyl alcohol), PVA, (Mw = 9 - 10 kDa), poly(N-isopropylacrylamide) (PNIPAM, Mw = 20-25 kDa) and dimethylformamide (DMF) solvent were obtained from Sigma-Aldrich and used without further purification. The fluorescent dye, Rhodamine 610 chloride (Mw=479 Da), was obtained from Exciton. 2.2. Electrospinning. The polymer solutions used for electrospinning were prepared by dissolving the fluorescent dye Rhodamine 610 in DMF and subsequently adding the required amounts of all three polymers together. The dye concentration was adjusted to 0.1 wt %. The solutions were prepared in 20 mL glass vials and wrapped with aluminum foil to prevent photobleaching. The electrospinning experiments were carried out using a standard single-nozzle setup.6,11,12 2.3. Annealing of Nanofiber Mats. The annealing steps were carried out in a furnace (Barnstedt Thermolyte 47900). The nanofiber mats were placed on a ceramic tile and kept at 150 °C for 2.5 h, which were the optimal conditions found given the restrictions imposed on fiber stability in water and thermodegradation of the polymers. During the annealing, an inert atmosphere was maintained inside the furnace by a constant flow of nitrogen gas. The gas flow was kept between 3 and 4 L/min at all times. It is emphasized that the Rhodamine dye should easily survive the annealing at 150 °C, as its melting point is around 205 °C with decomposition starting around 300 °C. The present experiments suggest that the fluorescence intensity of Rhodamine 610 chloride is not significantly diminished when the dye has been subjected to elevated temperatures. After keeping a solution of the dye in a closed vessel at 90 °C for 1 h and subsequent cooling, we found that the fluorescence decreased by 1 to 3%. We suspect that the incorporation of the dye into nanofibers to a certain extent protects the dye, reducing the loss of fluorescence even further. Generally, embedding the dye in a matrix is known to enhance the thermal stability of Rhodamine 610. Overall, we are confident that the dye fluorescence is only negligibly affected by being subjected to elevated temperatures. It is emphasized that the annealing of nanofibers at 150 °C for 2.5 h dramatically diminishes residual solvent traces in the solidified fiber mats and their possible effect on dye release. 2.4. Stability in Water. Special experiments were undertaken to elucidate stability in water of the nanofiber mats. All three polymers incorporated in our nanofibers, namely PAA, PVA, and PNIPAM, are water-soluble. To prevent excessive solubility in water, our nanofibers were internally cross-linked by annealing. The mechanism of the thermally induced crosslinking has already been described.5 It involves the following cross-linking reactions: dehydration, esterification and imidation. Additional experiments were designed to evaluate stability of our nanofiber mats in water. For this purpose, samples of 1-2 mg of the annealed (cross-linked) nanofibers were immersed in water for up to 4 days at temperatures of 25, 40, and 55 °C. After 4 days, the fibers were put on glass slides and allowed to dry at ambient conditions for at least 24 h. Subsequently, they were reweighed to determine the weight loss and analyzed by SEM to evaluate the structural changes caused by water. 2.5. Wettability of Nanofiber Mats below and above the LCST. An additional characterization step of PAA/PVA/PNIPAM nanofiber mats dealt with water wettability at temperatures below and above the LCST. High resolution, digital photo images were employed to measure contact angles of sessile water drops on the electrospun nanofiber mats at different temperatures (below and above the LCST). 10244 DOI: 10.1021/la1004177
Figure 1. Cross-linked electrospun nanofibers: (a) fibers with low PNIPAM concentration, (b) fibers with low PNIPAM concentration after immersion in water for 4 days, (c) fibers with medium PNIPAM concentration, (d) fibers with medium PNIPAM concentration after immersion in water for 4 days, (e) fibers with high PNIPAM concentration, and (f) fibers with high PNIPAM concentration after immersion in water for 4 days. Table 1. Shrinkage Ratio (%) of Two Samples without Dye When Water Temperature Increased from about 25 °C to 51.2 or 53.1 °Ca sample
51.2 oC
53.1 oC
1 5.06 5.06 2 32.31 38.46 a Sample 1 is a sheet of nanofibers with medium PNIPAM content. Sample 2 consists of fibers produced from a poly(methyl methacrylate/ N-isopropyl acrylamide) copolymer CP1-1 as in ref 9, and demonstrates strong shrinkage/swelling; it is used here for comparison. The experiment is similar to the one in ref 9, which demonstrated strong shrinkage/ swelling of poly(methyl methacrylate/N-isopropyl acrylamide) and poly(methyl methacrylate/N-isopropyl acrylamide/acrylic acid) copolymers. The error of this area measurement method is (2%.
2.6. Dye Release Experiments. The release experiments were carried out by cutting the annealed nanofiber mats into small sheets of 1-2 mg weight and immersing them in 2 mL of deionized water in a glass vial. After a given time, water was removed and replaced with fresh water, and fluorescence of the removed supernatant was measured. This was done with a SpectraMax spectrofluorometer (Molecular Devices) using an Langmuir 2010, 26(12), 10243–10249
Sinha-Ray et al.
Article
Table 2. Area of the Samples with Dye with Low, Medium, and High PNIPAM Concentration during the Swelling Experimentsa 25 oC
sample
40 oC
55 oC
low PNIPAM concentration 20.1 19.8 20.0 medium PNIPAM concentration 16.1 16.1 15.6 high PNIPAM concentration 15.6 15.3 15.6 a The areas (in mm2) were measured using the program ImageJ. Nanofiber samples containing dye floating in water were maintained at different temperatures: 25, 40, and 55 °C. The experiment is similar to the one in ref 9, which demonstrated strong shrinkage/swelling of poly(methyl methacrylate/N-isopropyl acrylamide) and poly(methyl methacrylate/N-isopropyl acrylamide/acrylic acid) copolymers. The error of this measurement is (2%.
Table 3. Contact Angles (deg) upon Drop Deposition Measured at Different Temperatures sample
SCA θ at 20 oC
SCA θ at 53 oC
sample 1 of Table 1 sample 2 of Table 1
57.6 116.4
116.7 129.1
excitation wavelength of 540 nm and measuring emission at 610 nm. The measured fluorescence was compared to a calibration curve and thereby the amount of dye released from nanofiber mats into water was quantified. In all these experiments the so-called “sink conditions” were strictly fulfilled, i.e., the release was not saturating the solution at any point. Indeed, the solubility of Rhodamines is typically of the order of 0.1 wt %, while in the release experiments in this work the dye concentrations in water never exceeded 0.00001 wt %. The samples were not mixed during the release experiments. Since the dye concentration in the solution was extremely low, the existence of saturated regions was virtually impossible (it would mean that the local dye concentration around the fibers had to be at least 104 times higher than in other regions of the solution). Dye release was investigated at four temperatures: 10, 25 (room temperature), 40, and 55 °C. The release studies at elevated temperatures were carried out by placing the immersed nanofiber samples in sealed vials on a hot plate. The cooled case (10 °C) was achieved by immersing the vials in a water bath maintaining constant temperature by cooled circulating water. Periodic variation of temperature was also used. To minimize the overall experimental error, for every experiment several nanofiber samples were cut out from the same electrospun mat and treated/analyzed identically. Additionally, for every one of these samples, several separate fluorescence measurements were averaged to minimize the error produced by the pipetting. 2.7. Staining of Nanofiber Mats. Staining was used to reveal the inhomogeneity of PAA/PVA/PNIPAM nanofiber mats related to aggregation of PNIPAM in the form of “raisins”, as it was suggested by the experiments on shrinkage and dye release. Samples of nanofiber mats were incubated in 5 mL of aqueous solution of CuSO4 3 5H2O (250 mM/L) separately for 2 h. They were then extracted from the CuSO4 solution and washed with deionized water for 3 h, with water being replenished every hour. In the nanofiber mats, only PNIPAM contains nitrogen (in distinction from PAA and PVA), which is more likely to coordinate with Cu2þ than the other atoms, e.g. such as O. As a result, the coordination effect of copper ions Cu2þ with nitrogen can be used to label PNIPAM. In particular, Cu2þ (and some other metal ions) complex can be formed because of the coordination of Cu2þ from CuSO4 solution with O and N of the amide group -CONH- in PNIPAM.13 However, a single NIPAM monomer, or a short PNIPAM chain, or any other O-containing group formed after cross-linking reactions in the fibers during annealing can provide only a few sites for coordination of Cu2þ, i.e., possess relatively low coordination numbers, which (13) Wang, Y.; Jiang, H. Mater. Lett. 2007, 61, 2779.
Langmuir 2010, 26(12), 10243–10249
Figure 2. Cumulative dye release vs time for PAA/PVA/PNIPAM nanofiber mats immersed in water at different temperatures: (a) nanofiber mats with low PNIPAM concentration, (b) nanofiber mats with medium PNIPAM concentration, and (c) nanofiber mats with high PNIPAM concentration. makes such complexes unstable. On the other hand, PNIPAM aggregates, e.g. PNIPAM “raisins”, provide with many more coordination sites, i.e., possess high coordination numbers. In the latter case, the chelation sites are close to each other making coordination of Cu2þ stable. As a result, Cu stains should be observable on PNIPAM “raisins”, i.e., Cu labels PNIPAM “raisins” alone.
2.8. Transmission Electron Microscopy (TEM) and Energy Dispersive X-ray (EDX) Analysis. All TEM observations were made using a JEOL JEM-3010 300 KV transmission electron microscope with LaB6 electron source. For TEM observation, two different TEM grids were used- a copper (Cu) grid with lacey carbon and a molybdenum (Mo) grid. For energy dispersive DOI: 10.1021/la1004177
10245
Article
Sinha-Ray et al.
Figure 3. Cumulative dye release vs time for nanofiber mats subjected to temperature modulation in time for (a) fibers with low PNIPAM concentration, (b) fibers with medium PNIPAM concentration, and (c) fibers with high PNIPAM content. (d) The corresponding temperature variation for each case (a-c). X-ray analysis (EDX) in TEM, the Mo grid was used in conjunction with a double-tilt holder made of beryllium, whereas the diffraction pattern was determined using the Cu grid in conjunction with a single-tilt holder made of copper. For TEM sample preparation, the nanofibers stained with copper sulfate were sonicated in isopropanol (IPA) using an ultrasonicator before very small drops of IPA with nanofiber fragments were deposited on grids of both types. All the diffraction pattern data were matched with the standard data from JCPDS;International Centre for Diffraction Data.
3. Experimental Results and Discussion 3.1. Fabrication and Characterization of Nanofiber Mats. Electrospinning of 3:3:2, 1:1:1 and 3:3:4 PAA:PVA:PNIPAM blends with an overall polymer concentration of 10 to 12 wt % yielded smooth nanofibers with diameters in the range of 400 to 500 nm. After electrospinning, the mats were annealed at 150 °C for 2.5 h to internally cross-link the nanofibers. The effectiveness of cross-linking in preventing leaching of water-soluble PAA, PVA and PNIPAM was tested by immersing the cross-linked nanofiber mats in water for several days. The mass loss was around 20 wt % in 1 day for the fibers with a low (3:3:2) and medium (1:1:1) PNIPAM content, and 30 wt % for the fibers with a high PNIPAM concentration (3:3:4). After that, during 4 days of immersion (in total) there was practically no additional weight loss in all three cases. Figure 1 shows SEM images of cross-linked fibers before (a, c, e) and after (b, d, f) immersion in water for 4 days. The nanofibers in Figures 1 (a), (c) and (e) do not show any sign of sintering-like fiber junctures-a clear sign that no overall macroscopic polymer flow happened during the preceding annealing (at 150 °C, slightly above the glass transition temperature of PNIPAM of about 135 °C). The morphology of the fiber mat after prolonged immersion in 10246 DOI: 10.1021/la1004177
water slightly changes to a more compact one with several interconnections between the fibers, probably due to partial local dissolution and recondensation of insufficiently linked polymer macromolecules during the immersion process. However, the fibers themselves remain mainly intact and their diameter is not visibly affected. Therefore, the cross-linking of the nanofibers can be considered successful. The weight loss probably results in formation of new nanopores, which are too small and not necessarily visible in the SEM images.11 3.2. Nanomat Behavior below and above LCST. One of the most important features of hydrogels is their swelling behavior. The uptake of water at T LCST. For example, in ref 5, it was reported that when increasing the temperature from 25 to 70 °C, PAA/PVA/PNIPAM nanofiber mats shrank to about one-fourth of their original size. The experiments with our electrospun nanofiber mats, however, yielded a very different result. Table 1 demonstrates that sample 1 without dye and with medium content of PNIPAM practically does not shrink when the LCST of 32 °C has been crossed. On the contrary, for sample 2 [a poly(methyl methacrylate/N-isopropylacrylamide] copolymer CP1-1 polymerized as in ref 9, the observed shrinkage ratio was in the range 32-38%, when water temperature crossed its LCST of 52 °C. Therefore, the PAA/PVA/PNIPAM nanofiber sample without dye practically did not shrink when its LCST was crossed, in distinction from the normal hydrogel in sample 2. Similar findings for PAA/PVA/ PNIPAM nanofiber samples with dye are presented in Table 2. We did not observe any temperature-dependent macroscopic swelling or shrinkage of the nanofiber mats, regardless of the PNIPAM content of the fibers. These observations show that in Langmuir 2010, 26(12), 10243–10249
Sinha-Ray et al.
Figure 4. Release rate modulation with temperature for (a) nanofibers with a low PNIPAM concentration, (b) nanofibers with a medium PNIPAM concentration, and (c) nanofibers with a high PNIPAM concentration. Each data point represents the change in slope of the corresponding curves shown in Figure 3.
distinction from ref 5 (where PAA/PVA/PNIPAM nanofibers swelled/shrank significantly, i.e. behaved as hydrogels), the present nanofibers did not show macroscopic signs of hydrogel behavior. The reason is most probably in a significant difference in the molecular weight of PNIPAM in the present work (Mw = 20-25 kDa) annealed for 2.5 h at 150 °C versus Mw = 79203 kDa in ref 5 annealed for 40 min at 140-180 °C. The shortchain PNIPAM molecules in the present nanofibers facilitated the aggregation of PNIPAM chains as “raisins” in the PAA/PVA matrix at the annealing stage, as the results discussed below confirm. Even though the present nanofiber mats did not show Langmuir 2010, 26(12), 10243–10249
Article
standard hydrogel behavior, they demonstrated a positive thermosensitive dye release, i.e., a significantly faster dye release at T>LCST (32 °C) as compared to TLCST (LCST =32 °C) as compared to T< LCST, as shown in Figure 2. To investigate the release kinetics from the nanofiber mats, a model compound, namely a fluorescent dye Rhodamine 610 chloride, was incorporated into the fibers by dissolving known quantities of the dye in the polymer solutions prior to electrospinning. The dye was chosen as a model compound because its amount released from the fibers over a given time period can be quantified by measuring the fluorescence of the water pool surrounding the nanofiber mats. All release studies were carried out at 10, 25 (room temperature), 40, and 55 °C: the first two values are below the lower critical solution temperature DOI: 10.1021/la1004177
10247
Article
Sinha-Ray et al.
Figure 5. PNIPAM “raisins” in nanofibers. (a) TEM image of a single nanofiber. (b) A zoomed-in area of the nanofiber shown in part a. The black “raisins” shown by arrows are of the order of 2 nm in diameter. (c) Diffraction pattern obtained from one of the “raisins”, which proves unequivocally the presence of a crystalline material. Moreover, the ring sharpness reveals crystals different from possible polymer crystallininty, since the latter is short-range and random, which would result in an integrated diffusive ring unlike the one in the image, which is characteristic of metals (in this case). The value of 30 cm in the image indicates the camera distance required to measure the d-spacing. (d) EDX spectrum obtained from a nanofiber with “raisins”, which shows the presence of Cu. The peak of Mo arises from the TEM grid made of Mo and used for EDX.
of PNIPAM, 32 °C, whereas the latter two are above it. These temperatures were chosen to represent three temperature ranges of interest. Comparing the release rates at 10 and 25 °C represents the release profile at T < LCST. By comparing the release rates at 40 and 55 °C, we can investigate the release at T >LCST, and most importantly, comparing the data acquired from the experiments at 25 and 40 °C elucidates the possible effect of PNIPAM swelling/shrinkage transition occurring at 32 °C on the release kinetics. The results in Figure 2 clearly indicate that the release of the dye increases with temperature. It is also clear that the release changes dramatically when the temperature is increased from 25 to 40 °C, while the release percentages change only marginally when the temperature is increased from 10 to 25 °C. For example, Figure 2b shows that the overall release percentage is very similar at 10 and 25 °C, being about 12% and 14% after 6 days, respectively. On the other hand, when temperature is increased from 25 to 40 °C, the 6-day release percentage increases significantly (to 39%; Figure 2b). This effect is attributed to PNIPAM swelling/shrinkage transition causing a drastic change in the release kinetics when the temperature exceeds 32 °C (LCST). An increase of temperature from 40 to 55 °C practically does not yield an increase of the 6-day release, which amounts to about 39% and 40%, respectively (Figure 2b). To further investigate the influence of the swelling/shrinkage transition of PNIPAM nanogel “raisins” and isolate it from other possible effects, experiments were conducted where the temperature was modulated in time during the dye release from the 10248 DOI: 10.1021/la1004177
electrospun nanofiber mats. Three temperature ranges were analyzed: 10 to 25 °C, 25 to 40 °C, and 40 to 55 °C, with the temperature being switched between the low and high temperature values every 1.5 to 5 h. A significant effect of the PNIPAM thermosensitivity on the release rate is expected in the interval from 25 to 40 °C, which contains the LCST. The experiments with temperature modulation (Figure 3) indicate that when the temperature was increased from 25 to 40 °C, the slope of the percent release increased visibly and decreased accordingly when the temperature was switched back from 40 to 25 °C. The same effect is also apparent for the other two temperature ranges, even though it is significantly smaller (cf. Figure 4). In particular, Figures 3 and 4 clearly show a very significant effect of the temperature response facilitated by PNIPAM nanogel “raisins”. The slope changes dramatically by as much as 300% to 700% (Figure 4a) when the temperature is switched from 25 to 40 °C. For each of the three PNIPAM concentrations studied, the change of slope observed in the other two temperature ranges is significantly smaller than in the corresponding range from 25 to 40 °C. This clearly indicates how the temperature response of PNIPAM at swelling/shrinkage transition outweighs all other effects caused by a change in temperature in the temperature intervals that do not contain the LCST. 3.4. PNIPAM Nano-“Raisins” Revealed by Staining, Transmission Electron Microscopy (TEM), and Energy Dispersive X-ray (EDX) Analysis. The absence of macroscopic shrinkage at T > LCST together with a significant thermal Langmuir 2010, 26(12), 10243–10249
Sinha-Ray et al.
Article
response of the nanofiber mats demonstrated in the release experiments in Figures 2-4 indirectly suggest that PNIPAM might be agglomerated in the mats in the form of “raisins”. When staining with copper sulfate (CuSO4), coordination of Cu2þ with PNIPAM makes this polymer visible (and distinct from PAA and PVA) in TEM images. Two different types of holders and grids in TEM were used due to the following reasons. The nanofibers were stained with copper coming from copper sulfate (CuSO4). Using EDX to prove unambiguously the existence of copperstained “raisins” in nanofibers alone, one needs to avoid any signal of copper coming from the grid and holder, which can be a source of a false signal. Therefore, a molybdenum (Mo) grid was used with a double-tilt holder made of beryllium for determining EDX. However, the double-tilt holder had an insufficiently tight fixture and it was hard to focus the electron beam on a single “raisin” to obtain the diffraction pattern. Therefore, a Cu grid was used with a single-tilt holder to obtain the diffraction pattern. The nanofiber samples were studied using TEM revealed identical structures. A set of representative images is shown in Figure 5a, b. The overall view of a single nanofiber in Figure 5a reveals black dot-like objects. Zooming-in (Figure 5b) showed that these objects are “raisins” of about 2 nm in diameter. These “raisins” are expected to be PNIPAM islands stained by Cu from CuSO4 . To prove that, diffraction pattern of the “raisins” was studied. An example of the diffraction pattern is depicted in Figure 5c. For several such images the d-spacing corresponding to the sharp crystalline ring was measured using the following formula d-spacing ðin ÅÞ ¼ ðcamera constant 2Þ =diameter of the ring in pixels
ð1Þ
The corresponding value of the d-spacing was found to be 0.148 nm. This value is close to the d-spacing for cuprite (Cu2O), where for the (h,k,l) plane the interplanar d-spacing in the (2,2,0) direction is 0.151 nm. The measured value of the d-spacing is also
Langmuir 2010, 26(12), 10243–10249
close to the one for cupric oxide (CuO) whose d-spacing of the (1,1,3) plane is 0.15 nm. However, if the “raisin”-covering material were CuO, several rings should have been seen corresponding to the d-spacing of the (3,1,1), (1,1,3) and (3,1,1) planes, namely 0.14 nm, 0.138 and 0.13 nm, respectively. The absence of these rings suggests that this material consists predominantly of Cu2O. The presence of a single ring in Figure 5c corresponding to a single d-spacing suggests that sonication flattened the nanofibers. To prove the presence of Cu on the “raisins”, EDX spectra were observed (Figure 5d). The spectra indeed demonstrated a significant amount of Cu on the “raisins”.
4. Conclusions The “raisin”-like distribution of PNIPAM in the PAA/PVA/ PNIPAM nanofiber mats corresponds to the structural model10 and explains the main reason for their lack of shrinkage/swelling behavior compared to that of such ordinary hydrogels as poly(methyl methacrylate/N-isopropylacrylamide) copolymers developed in ref 9. The “raisin”-like distribution of PNIPAM in the present case is associated with its low molecular weight and thus an elevated mobility at the annealing temperature. These nano“raisins” alone are sufficient for positive thermoresponsive dye release according to the mechanism proposed in ref 10. Namely, contraction of these “raisins” at elevated temperatures results in elastic stresses and strains in the surrounding material, which can lead to nanocracking. The latter brings more dye in contact with water and facilitates dye release in those areas. This is indirectly corroborated by the positive thermosensitivity observed experimentally. The nanofiber mats studied in this work represent a novel class of resins with nanohydrogel “raisins”. Bulk materials of this type can also be obtained. Acknowledgment. The work was supported in part by the National Science Foundation under Grants NIRT CBET-0609062 and NER CBET-0708711. Alan W. Nicholls and Ke-bin Low are gratefully acknowledged for their help in the interpretation of the TEM results.
DOI: 10.1021/la1004177
10249
