Single Chain Diffusion of Poly(ethylene oxide) in Its Monolayers

Dec 29, 2009 - Lateral diffusion of single chain related to the crystallization process of poly(ethylene oxide) (PEO) in its monolayers on silica surf...
0 downloads 0 Views 3MB Size
Download PDF
pubs.acs.org/Langmuir © 2009 American Chemical Society

Single Chain Diffusion of Poly(ethylene oxide) in Its Monolayers before and after Crystallization Rui Chen,† Lin Li, and Jiang Zhao* Beijing Laboratory of Molecular Science, Joint Laboratory of Polymer Science and Materials, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China. †Also affiliated with the Graduate School of Chinese Academy of Sciences. Received October 14, 2009. Revised Manuscript Received December 15, 2009 Lateral diffusion of single chain related to the crystallization process of poly(ethylene oxide) (PEO) in its monolayers on silica surfaces is studied by single molecule fluorescence microscopy and single molecule tracking techniques. Diffusion of PEO chains is observed in the supercooled state before crystallization as well as in the noncrystalline regions after crystallization. In the postcrystallization monolayers, the diffusion coefficient of PEO chains is an order of magnitude lower than that in the supercooled state before crystallization. The origin is attributed to the change of polymer surface concentration due to the consumption of polymer molecules in the crystal formation. This is supported by the results showing a monotonous decrease of diffusion coefficient with the thickness decrease of the monolayer in its supercooled state. The PEO chains take a more flattened conformation under lower surface concentration and stick stronger to the surface. As a consequence, the diffusion rate is reduced. The results clearly demonstrate a strong mutual effect between the crystallization process and the mass transportation for the polymer crystallization under surface confinement.

Introduction Polymers under confinement exhibit unique properties in a number of aspects, such as chain conformations, dynamics, phase separation, phase transitions such as glass transition and crystallization, etc., compared with their bulk analogue.1-7 The importance of researches on confined polymers lies not only on its significance in fundamental polymer physics but also on the understanding of the principles in many important applications, such as coating technology, tribology, microelectronics, nanofabrication, etc. Polymer crystallization under spatial confinement has been attracting research interests in the past decade, and it is one of most exciting fields regarding the physics of polymer metastable *Corresponding author. E-mail: [email protected]. (1) (a) Granick, S. Science 1991, 253, 1374. (b) Phys. Today 1999, 52, 26. (2) (a) Frank, C. W.; Rao, V.; Despotopoulou, M. M.; Pease, R. F. W.; Hinsberg, W. D.; Miller, R. D.; Rabolt, J. F. Science 1996, 273, 912. (b) Koutsky, J. A.; Walton, A. G.; Baer, E. J. Appl. Phys. 1967, 38, 1832. (3) Keddie, J. L.; Jones, R. A. L.; Cory, R. A. Europhys. Lett. 1994, 27, 59. (4) Forrest, J. A.; DalnokiVeress, K.; Dutcher, J. R. Phys. Rev. E 1997, 56, 5705. (5) Bezrukov, S. M.; Vodyanoy, I.; Brutyan, R. A.; Kasianowicz, J. J. Macromolecules 1996, 29, 8517. (6) Hu, Z. J.; Huang, H. Y.; Zhang, F. J.; Du, B. Y.; He, T. B. Langmuir 2004, 20, 3271. (7) Wang, Y.; Chan, C. M.; Li, L.; Ng, K. M. Langmuir 2006, 22, 7384. (8) (a) Cheng, S. Z. D.; Lotz, B. Philos. Trans. R. Soc. London A 2003, 361, 517. (b) Polymer 2005, 46, 8662. (9) Yang, P.; Han, Y. C. Langmuir 2009, 25, 9960. (10) Strobl, G. R. The Physics of Polymers: Concepts for Understanding Their Structures and Behavior; Springer: Berlin, 1997. (11) Cheng, S. Z. D. Phase Transitions in Polymers: The Role of Metastable States; Elsevier: Amsterdam, 2008. (12) (a) Sawamura, S.; Miyaji, H.; Izumi, K.; Suthen, S. J.; Miyamoto, Y. J. Phys. Soc. Jpn. 1998, 67, 3338. (b) Taguchi, K.; Miyaji, H.; Izumi, K.; Hoshino, A.; Miyamoto, Y.; Kokawa, R. Polymer 2001, 42, 7443. (13) (a) Reiter, G.; Sommer, J.-U. Phys. Rev. Lett. 1998, 80, 3771. (b) J. Chem. Phys. 2000, 112, 4376. (c) Reiter, G.; Castelein, G.; Sommer, J.-U. Phys. Rev. Lett. 2001, 86, 5918. (14) (a) Sch€onherr, H.; Frank, C. W. Macromolecules 2003, 36, 1188. (b) Macromolecules 2003, 36, 1199. (15) Despotopoulou, M. M.; Frank, C. W.; Miller, R. D.; Rabolt, J. F. Macromolecules 1996, 29, 5797. (16) Zhai, X. M.; Wang, W.; Ma, Z. P.; Wen, X. J.; Yuan, F.; Tang, X. F.; He, B. L. Macromolecules 2005, 38, 1717.

Langmuir 2010, 26(8), 5951–5956

states.8-11 Although there have been intensive and systematic studies on crystallization of polymers,8-19 polymer crystallization under confinement is still attracting intensive research attention.20-22 In polymer thin films with thickness comparable to the chain dimension (the radius of gyration of polymer chains, Rg), polymer crystallization inside has been found to be very different compared with the bulk behavior, partly because of the special chain conformation as a result of the confinement effect and partly because of the additional interaction between polymer chains and the surrounding environment, such as solid substrates. Polymer crystallization in thin films exhibits a number of new features, such as the prolonged supercooled state as a result of the high free energy barrier of primary nucleation,23 and the evolution of crystal morphology after crystallization, such as thickening process of lamellar crystals even below melting temperature due to the transition from kinetically trapped metastable state with multiple chain folding to the state with more equilibrium.11-13,16,24,25 Mass transportation has long been considered as an important issue in polymer crystallization under confinement. It not only is involved in the supercooled state in which the nucleation is initialized but also should participate in the crystal growth process (17) Zhai, X. M.; Wang, W.; Zhang, G. L.; He, B. L. Macromolecules 2006, 39, 324. (18) Zhu, D. S.; Liu, Y. X.; Chen, E. Q.; Li, M.; Chen, C.; Sun, Y. H.; Van Horn, R. M.; Cheng, S. Z. D. Macromolecules 2007, 40, 1570. (19) Zhu, D. S.; Liu, Y. X.; Shi, A. C.; Chen, E. Q. Polymer 2006, 47, 5239. (20) Kailas, L.; Vasilev, C.; Audinot, J. N.; Migeon, H. N.; Hobbs, J. K. Macromolecules 2007, 40, 7223. (21) (a) Hsiao, M. S.; Chen, W. Y.; Zheng, J. X.; Van Horn, R. M.; Quirk, R. P.; Ivanov, D. A.; Thomas, E. L.; Lotz, B.; Cheng, S. Z. D. Macromolecules 2008, 41, 4794. (b) Hsiao, M. S.; Zheng, J. X.; Leng, S. W.; Van Horn, R. M.; Quirk, R. P.; Thomas, E. L.; Chen, H. L.; Hsiao, B. S.; Rong, L. X.; Lotz, B.; Cheng, S. Z. D. Macromolecules 2008, 41, 8114. (22) Massa, M. V.; Carvalho, J. L.; Dalnoki-Veress, K. Phys. Rev. Lett. 2006, 97, 247802. (23) Keller, A.; Cheng, S. Z. D. Polymer 1998, 39, 4461. (24) Hwang, R. Q.; Schr€oder, J.; G€unther, C.; Behm, R. J. Phys. Rev. Lett. 1991, 67, 3279. (25) Brener, E.; M€uller-Krumbhaar, H.; Temkin, D. Phys. Rev. E 1996, 54, 2714.

Published on Web 12/29/2009

DOI: 10.1021/la903897v

5951

Article

Chen et al.

as well as in the crystal thickening process. For example, it was discovered that the crystal growth rate in polymer thin films is highly dependent on the film thickness,2 which should be related to different mass transportation rate in the thin films. Also, it is believed that chain diffusion can play an important role during the thickening process of lamellar crystals, serving as the pathway of transportation from one crystal to another.16,18 Polymer mass transportation under confinement is also an important topic related to many phenomena, such as wetting and dewetting, surface fluidics, and nanorheology.1 The chain diffusion under confinement is a process highly dependent on the state of the polymer chains, such as molecular weight, chain conformation, surface concentration, chain-surface interactions, etc.26-31 Therefore, it is worthy to study the relation between the crystallization process and mass transportation of confined polymers. The study can help to look into the relation between these two properties. By conventional methods, it is hard to study the microscopic feature of mass transportation (chain motion) related to polymer crystallization and melting process. Recently, single molecule fluorescence microscopy was successfully applied to study single chain diffusion, in solution, at interfaces, and during polymer crystallization.32-34 Its high detection sensitivity and fine spatial resolution help to study not only the motion of single polymer chains but also chains’ motion with respect to the sample’s morphology. In this study, we used single molecule fluorescence microscopy to study single chain diffusion of confined poly(ethylene oxide) (PEO) related to its crystallization and melting process. The results showed the change of chain diffusion coefficient before and after crystallization, which provide new information on the mass transportation rate related to the crystallization behavior in polymer thin films.

Experimental Section Materials. PEO (Mn = 5000 g mol-1, Mw/Mn = 1.08) was purchased from Polymer Source (Quebec, Canada). In order to let the center-of-mass motion of the polymer chain be observable under fluorescence microscope, a bright and stable fluorophore, Rhodamine 6G, was chemically attached to each chain end by a similar protocol to that published previously.32,33 Briefly, the original PEO chain was terminated with a methyl group at its one end and with a hydroxyl group at the other end. An esterexchange reaction was conducted (in dimethyl sulfoxide) between the ester group of Rhodamine 6G and the hydroxyl group at the PEO chain end so that one end of the PEO chain was labeled with the fluorophore. After the reaction, the fluorescence-labeled sample was carefully purified by both size exclusion chromatographic column and dialysis. After the dialysis had been performed for one month, the final product was carefully checked by fluorescence correlation spectroscopy (FCS), which measured the diffusion of fluorescent molecules in solution. The results showed the existence of pure diffusion of labeled polymer with a diffusion coefficient (D) of 6.5  10-11 m2/s and no diffusion of free dye was detected, which should have exhibit a D value of 2.8  10-10 m2/s. (The detailed description and data are provided in the Supporting Information.) (26) Granick, S. Eur. Phys. J. E 2002, 9, 421. (27) Bae, S. C.; Granick, S. Annu. Rev. Phys. Chem. 2007, 58, 353. (28) Fleer, G. J.; Cohen Stuart, M. A.; Scheutjens, J. M. H. M.; Cosgrove, T.; Vincent, B. Polymers at Interfaces; Chapman & Hall: London, 1993. (29) Sukhishvili, S. A.; Chen, Y.; Muller, J. D.; Gratton, E.; Schweizer, K. S.; Granick, S. Nature 2000, 406, 146. (30) (a) Zhao, J.; Granick, S. J. Am. Chem. Soc. 2004, 126, 6242. (b) Macromolecules 2007, 40, 1243. (31) (a) Mukherji, D.; M€user, M. H. Phys. Rev. E 2006, 74, 010601. (b) Macromolecules 2007, 40, 1754. (32) Wang, S. Q.; Zhao, J. J. Chem. Phys. 2007, 126, 091104. (33) Yang, J. F.; Zhao, J.; Han, C. C. Macromolecules 2008, 41, 7284. (34) Bi, W.; Teguh, J. S.; Yeow, E. K. L. Phys. Rev. Lett. 2009, 102, 048302.

5952 DOI: 10.1021/la903897v

Quartz coverslips (ESCO Products) were used as substrates for PEO thin films. Before use, the slips were treated successively in UV-ozone, ultrasonication in acetone, isopropanol, and water, and finally in oxygen plasma. Sample Preparation. PEO thin films were prepared by spincoating from the acetonitrile solution of unlabeled PEO with a concentration of 0.1 wt %. A trace amount of labeled PEO (at a concentration of ∼10-9 M) was mixed in the unlabeled PEO solution. The thin film samples were kept in vacuum at 22 C for more than 24 h to remove possible residual solvent. After melting the sample at 80 C, the film thickness was measured by an ellipsometer (V-2000, J.A. Woollam), and a typical film thickness was measured to be ∼2.4 nm. Microscopic Observation. The microscopic observation was conducted by a single molecule fluorescence microscope and an atomic force microscope (AFM). The fluorescence microscope was a home-built setup on an inverted microscope (Olympus IX71, Japan) equipped with an EMCCD (Andor 887, Northern Ireland). A laser beam of 532 nm was introduced to the sample via total internal reflection geometry through an oil-immersion objective lens (100 PlanApo, numerical aperture = 1.45). The timing of the image capture was controlled by external clock signals, and the exposure time was optimized (0.1 and 0.2 s) in order to have high enough image contrast without losing time resolution (data provided in Supporting Information). The surface morphologic observation was performed on a Nanoscope IIIA (Vecco) AFM under controlled temperature and atmosphere. To measure chain diffusion in the supercooled thin film before crystallization, the sample was first incubated in a hot/cold stage (TSA02i, INSTEC) at 80 C for 5 min in order to allow the crystals to melt completely. The individual sample was then cooled to its targeting temperatures (50, 38, and 22 C). After 10 min of equilibrium at each temperature, microscopic measurements were performed. To observe the chain motion after crystallization and during the melting process, the sample was incubated at 22 C for 20 h so as to let the crystallization process complete before the microscopic observation. When the observation at 22 C was finished, the sample temperature was raised to 38 C and then to 50 C for further observation. At least 10 min of incubation was always allowed before microscopic observation. In order to suppress the effect of humidity, all of the sample preparation and measurements were conducted under a controlled humidity environment (relative humidity 4 h for 22 C, for example), showing that no dewetting occurred and that the PEO films remained in their supercooled state without crystallization. This is because of the well-known effect of surface confinement: the high free energy barrier of primary nucleation. These samples are referred as “supercooled” in the following. The film thickness here was 2.4 nm, smaller than the chain dimension (Rg ∼ 2.6 nm),35 and therefore the thin films were considered to be monolayers. The samples were observed to have crystallized after 20 h: fingerlike monolamellar crystals with a thickness of ∼6 nm were formed on the quartz surface (Figure 1a).11,16,17 Afterward, the sample temperature was increased to 38 C for in-situ AFM characterization (Figure 1b) and then to 50 C (Figure 1c). 10 min of incubation was allowed for each temperature (35) Fetters, L. J.; Lohse, D. J.; Richter, D.; Witten, T. A.; Zirkel, A. Macromolecules 1994, 27, 4639.

Langmuir 2010, 26(8), 5951–5956

Chen et al.

Figure 1. (a-c) In-situ AFM height images showing the morphology evolution of PEO monolamellar crystals under increasing temperatures: (a) 22, (b) 38, and (c) 50 C. The scale in z-direction is 30 nm. (d-f) Fluorescence microscope images of PEO monolamellar crystals under increasing temperatures: (d) 22, (e) 38, and (f) 50 C. The corresponding videos showing the diffusive motion of single PEO chains are provided in the Supporting Information. The arrows denote the typical regions where the single molecule tracking analysis was conducted.

before AFM observation. The images show that the crystal melting and thickening process occurred. In some other cases, the decrease in crystal amount and the increase of their height were observed. These samples are referred as “postcrystallization” in the following text. From Figure 1a-c, it is observed that large fingerlike lamellar crystals decomposed and turned into small fragments (Figure 1b,c). These phenomena are similar to what reported by other groups.13,16 In the current experiments, the height of the lamellar crystals are measured to be about 8 and 11 nm for 38 and 50 C, respectively. These thickness values were found to be stable within the time of observation. Diffusion of PEO Single Chain in Supercooled PEO Thin Films and the Postcrystallization Sample at Different Temperatures. Under a single molecule fluorescence microscope, diffusive motion of single PEO chain in the supercooled thin films was clearly visible. Please refer to the Supporting Information for the video (Movie 1 shows the case of 50 C as an example. The movies of 38 and 22 C are similar and not shown here.) Similar to AFM observations, the fluorescence images show no stationary patterns or features at this stage, demonstrating no crystal formed in these supercooled films within the observation time. After prolonged incubation (>20 h at 22 C), the formation of fingerlike patterns of lamellar crystals was clearly observed in the fluorescence images (Figure 1d as a typical image). The results showed that most molecules were immobilized because they were trapped in the crystal lattice. However, a few slowly moving molecules outside the crystal regions were clearly observed. When the postcrystallization samples were incubated under elevated temperatures, fingerlike patterns began to decompose; i.e., large lamellar crystals turned into small locally separated fragments. Also, a large number of molecules were observed to make random motion. The moving molecules were visualized in two types of regions: in regions outside the lamellar crystals and in regions where the previous lamellar crystals resided, i.e., between the fragments after decomposition of the lamellar crystals. Typical fluorescence images for 38 and 50 C are provided in parts e and f of Figure 1, respectively. The movie for 50 C is provided in the Supporting Information as a typical example (movie 2). The images and videos were taken with the same sample but at different position due to the photobleaching effect to the fluorescence Langmuir 2010, 26(8), 5951–5956

Article

dyes. These results reveal two major facts: (1) There are still free moving PEO chains outside the lamellar crystals even after prolonged crystallization time (>20 h at 22 C), indicating that the crystallization process is still far from being completed. Normally, the process of crystallization is considered to be completed if no noticeable change of crystal morphology is observed, for example, by AFM measurements. The current observation by single molecule fluorescence microscopy provided new information on this issue. (2) The melting of the lamellar crystals is observed at temperatures much lower than the melting temperature of the bulk PEO crystals. The observations by AFM and fluorescence microscope show consistent results. The visualization of moving molecules between the fragments inside the lamellar crystals clearly demonstrate the melting process, and this also helps to exclude the possibility of artifacts induced by probing tips in the AFM observation. Quantitative measurements of diffusion coefficient of individual PEO chain were performed by single molecule tracking method. The tracking process was conducted manually. In each sample, the trajectories of 50 moving molecules for each temperature were recorded for at least 30 steps, based on which the mean-squared displacement (MSD) was calculated,34,36 adopting Pj-n 2 the definition of ÆR(tn) æ = [1/(j - n)] i=1Ri,iþn2, where Ri,iþn is the displacement of the molecule between frames i and i þ n; j is the total number of frames. ÆR2æ is a function of delay time (tn, tn = nδt, where n is the number of frames and δt is the time interval between frames). The diffusion coefficient (D) of the PEO chain was determined by its relation with MSD in two dimension: ÆR2(t)æ = 4Dt. For the postcrystallization sample, only molecules at certain distance away from the lamellar crystal region are investigated (regions indicated by the arrows in Figure 1e,f). The molecules at the edge of the lamellar crystals or inside the lamellar crystals region are not investigated. This helps to exclude the possible effects that the crystals may impose to the molecules’ motion, for example, the possible hindrance by the crystals. Figure 2a shows the temporal profile of MSD for 50 polymer chains in the supercooled sample at 50 C, and Figure 2b shows the D values at different temperatures for both supercooled and postcrystallization samples. The D values were taken from the results of histogram analysis, in which the maximum values and the width of the Gaussian fitting were taken as the average values of D and the average value ranger (please refer to Figure 3). A variation of diffusion coefficient for all temperatures was observed. As an example, the value of D ranges from 1.1  10-14 to 7.5  10-13 m2 s-1 for the supercooled monolayer of 2.4 nm thickness at 50 C. This is attributed to the inhomogeneity of the silica substrate surface, in both topography and surface chemistry.37 This fact is more clearly demonstrated in the histograms of diffusion coefficient of individual PEO chains in the supercooled sample before crystallization (Figure 3a-c) and the postcrystallization sample (Figure 3d-f) under different temperatures. The data exhibit no noticeable change of diffusion coefficient values for the supercooled monolayer at varying temperatures, but a slight increase of D value in the postcrystallization sample at higher temperature was clearly observed (Figure 2b). Also, it is clearly seen that there is a big decrease of the number of fast diffusing molecules at lower temperatures (Figure 3). Most of all, the biggest feature is the large difference in D value before and after crystallization. In the regions outside the crystals in the (36) Vrljic, M.; Nishimura, S. Y.; Brasselet, S.; Moerner, W. E.; McConnell, H. M. Biophys. J. 2002, 83, 2681. (37) Massa, M. V.; Dalnoki-Veress, K.; Forrest, J. A. Eur. Phys. J. E 2003, 11, 191.

DOI: 10.1021/la903897v

5953

Article

Chen et al.

Figure 2. (a) Temporal profile of mean-square displacement ÆR2æ of 50 individual PEO chains in the supercooled sample (film thickness: 2.4 nm) at 50 C before crystallization. Inset: typical trajectories of PEO molecules in supercooled (A and B, time interval between adjacent steps = 0.1 s) and postcrystallization sample (C and D, time interval between adjacent steps = 1 s) at 50 C. (b) Value of diffusion coefficient of PEO chains in its supercooled condition at different temperature (9) and postcrystallization samples at different temperature (b).

Figure 3. Histogram of diffusion coefficient of individual PEO chains, in which the left panel shows data for supercooled sample before crystallization and right panel shows the postcrystallization samples: (a, d) 50 C; (b, e) 38 C; (c, f) 22 C. The solid lines are the results of Gaussian fitting. The data were from tracking analysis of 50 individual molecules. The inset of (a) shows the result of 100 molecules, which is similar to that of 50 molecules.

postcrystallization samples, the average value of D is about an order of magnitude lower than those of the supercooled samples for each temperature under investigation (Figures 2b and 3). Attention is paid to investigate the origin of the vast difference of diffusion rate of PEO chains in the supercooled and the postcrystallization samples. First of all, the effect of hindrance by crystals as obstacles is excluded because the molecules under analysis were chosen as those moving at micrometers’ distance from the crystals. This is also proved by further experiments showing the dependence of diffusion coefficient on the thickness with the supercooled film (please see the text later). Excluding this possibility, attention was paid to the difference in the state of the PEO chains between these two cases. It is very likely that the PEO chains adopt different conformations due to the difference in chain concentration: the chain takes a more flattened conformation in less concentrated noncrystallized region in the postcrystallization sample compared with the supercooled sample. This concentration reduction is because a large amount of PEO chains 5954 DOI: 10.1021/la903897v

have been taken by the crystals, which has a much higher chain density than the amorphous region. As a unique property of polymers, the chain conformation depends on the surface chain concentration. In general, polymer chains on an attractive solid surface bear different portion of segments: segments in direct contact with the surface (the “train” segments) and those not in contact with the surface (the “loop” and “tail” segments).28,29 The distribution of the population of these segments depends on the surface chain concentration. For the case of extreme dilution, i.e., the isolated chains, the chains take the totally flat “pancake” conformation.28,29 At higher surface concentration, the mutual interaction between the chains results in more “loop” and “tail” segments, i.e., the more fluffy chain conformation.30,31 The more concentrated the surface is, the more fluffy conformation the chains will take in the monolayer, and vice versa. The lateral diffusion rate of polymer chains on the surface is also concentration-dependent, as a result of the conformation Langmuir 2010, 26(8), 5951–5956

Chen et al.

Figure 4. Diffusion coefficient (D) of single PEO chains in its supercooled monolayers at 50 C as a function of the film thickness. The very left datum is for the isolated chain, which shows no observable diffusion within the time of measurement (200 s).

change. Previously, both experimental studies and computer simulations showed that the surface diffusion coefficient of adsorbed polymer chain experiences a monotonous increase with the increase of surface concentration before it reaches the saturated monolayer concentration.30,31 The origin of this behavior is because of the mutual interaction between the polymer chains, which in turn results in more “loop” and “tail” segments and less “train” segments when the surface chain concentration is increased from dilute to concentrated. The “train” segments affect the diffusion rate because they interact directly with the solid surface underneath. Therefore, a more flattened chain in the less concentrated region has more “train” segments and stick stronger to the surface, and consequently, the chain diffuses slower. For the current system, the PEO chains are attracted to the silica surface via hydrogen bonds between their oxygen atoms and the surface hydroxyl groups. The surface concentration of PEO chains in the amorphous regions are affected by the crystallization process. When lots of PEO chains were consumed by crystallization, the chain concentration in the outside-crystal regions of the postcrystallization film becomes lower and the chains should take more flattened conformation than those in the supercooled films, and this induces a slower diffusion rate consequently. To verify this scenario, investigations on chain diffusion in PEO monolayers with various surface concentrations was conducted. Generally, at high enough surface coverage, i.e., when the surface concentration is higher than the overlapping concentration (c*), the conformation change of the adsorbed polymers is reflected in the value of film thickness (or monolayer thickness). Therefore, sample layers with different thicknesses were prepared by spincoating from solution with different concentrations, and measurements were performed at their supercooled state at 50 C cooled from 80 C. First of all, the experiment on extremely dilute situation (samples prepared by 10-9 M solution of labeled PEO) revealed that the isolated PEO chains were immobilized at the surface under all temperatures investigated within the time scale in our experiment, i.e., 200 s (please refer to movie 3 in the Supporting Information). As the layer gets thicker (corresponding to an increase of surface concentration), faster diffusing molecules were observed. The value of D increases with the monolayer’s thickness monotonously, as shown in Figure 4. As a comparison, the 1.3 nm thick supercooled layer exhibits a comparable D value (0.4  10-14 m2/s) with that of the 2.4 nm thick sample after crystallization at 50 C (Figure 3d). The Langmuir 2010, 26(8), 5951–5956

Article

Figure 5. Histogram of diffusion coefficient of individual PEO chains in 1.3 nm thick supercooled film (a), postcrystallization film of 2.4 nm thick sample (b), and 2.4 nm thick supercooled PEO film (c). The temperatures for three charts are 50 C. The solid lines denote the results of Gaussian fitting. A schematic illustration of the physical model is provided to the right of each chart.

detailed histogram analysis of these two samples is provided in Figure 5, in which the data for the 2.4 nm thick supercooled layer are also displayed for comparison. The D values of 1.3 nm thick supercooled layer and the 2.4 nm thick postcrystallization sample fall in a similar range, and they are both 1 order of magnitude lower than that of 2.4 nm thick supercooled layer. For a rough estimate by taking the density of PEO melt as 1.123 g cm-3, the chain density is ∼0.3 chain per nm2 for the 2.4 nm thick film and ∼0.16 chain per nm2 for the 1.3 nm thick films, respectively. The results indicate that the diffusion rate drops around 10 times when the concentration decreases to around 50%. We believe that this estimate of surface concentration is very rough because the density data for the PEO surface layer should be much lower than the bulk value because of the confinement effect. Because there is no crystal formed in these thin supercooled films, the thickness dependence of diffusion rate also helps to exclude the possibility that the slowing down of diffusion in the postcrystallization sample is due to the hindrance by crystals serving as obstacles to diffusion. All these data show that crystallization process takes the PEO chain into the crystal and leaves a much more dilute surface behind, on which the polymer chains are more flattened and stick stronger to the surface. This is the reason for the huge slowing down of the polymer chain after crystallization. Another issue is regarding the possible effect of the fluorescent dye to the mobility of the PEO chain. Control experiments were conducted to measure the diffusion of free Rhodamine 6G in 2.4 nm thick supercooled PEO layer at 50 C. The results showed that the D value of the free dye was ∼6  10-12 m2/s, which was 2 orders of magnitude higher than that of the labeled PEO chain (measurements done by FCS). This fact demonstrated that the fluorescent label has minor effect to the chain mobility. One more consideration is the possible concentration gradient formed near the edge of the lamellar crystals during the melting process at elevated temperatures. This effect is considered to be minor because, within the time of fluorescence microscopic observation, the thickness of the lamellar crystals was constant and stable, meaning they have reached a relatively equilibrate thickness at the specific temperature, as seen in the AFM images (Figure 1b,c). In this case, the concentration gradient near the crystal is minor. DOI: 10.1021/la903897v

5955

Article

The above picture is also supported by the temperature dependence of D in the supercooled and postcrystallization sample: the D value depends on temperature in the postcrystallization film, but it is insensitive to temperature in the supercooled sample (Figure 2b). This is because more crystals were molten at higher temperatures, and the noncrystalline regions become more concentrated and therefore the chains in these areas diffuse faster. Let us assume that the diffusion coefficient is proportional to the temperature (D  T, based on the thermoactivation) and the ratio of D at 50 and 22 C is ∼1.09 [(323 K)/(295 K) = 1.09], which should not be noticeable in the measurements. However, the measured increase of D with temperature for the postcrystallization sample is over 2-fold [(0.38  10-14 m2/s)/(0.17  10-14 m2/s) = 2.2]. Therefore, this difference should not be originated from thermoactivation. The change of chain concentration and therefore the conformation should play a much bigger role. The diffusion data presented here can also provide new insight in the crystal growth rate of polymer crystallization in thin films. Previous experiments show that the growth rate of polymer crystals decreased with the decrease of film thickness.2,8,14,15,37 The findings of the current study implies that the growth of the crystals can considerably slow down the chain diffusion rate in the amorphous region, which can be one of the reasons affecting the crystal growth rate when it is a diffusion-limited process. Quantitative investigation is need in connection with the Avrami and Hoffman theories.

5956 DOI: 10.1021/la903897v

Chen et al.

Conclusions Single molecule fluorescence microscopy and molecule tracking techniques have been applied to study the diffusion of single PEO chains in its monolayers on solid substrates, before and after the process of crystallization. The results show that there is a strong connection between the crystallization process and the mass transportation (chain diffusion). The consumption of PEO chains by the formation of crystals results in the reduction of the amount of amorphous polymers in the thin layers and therefore induces much more flatten chain conformation. This consequently results in a stronger interaction between the polymer chains and the surface underneath and a vast slowing down of the diffusion rate of the polymer chain. Acknowledgment. Helpful discussions with Dr. Er-Qiang Chen are highly appreciated. This project is supported by National Natural Science Foundation of China (NSFC 20634050, 20925416, 50730007, 20925416). Supporting Information Available: Fluorescence correlation spectroscopy data of fluorescence labeled PEO in aqueous solution and Rhodamine 6G; movies of supercooled, postcrystallization, and extremely dilute PEO thin films. This material is available free of charge via the Internet at http://pubs.acs.org.

Langmuir 2010, 26(8), 5951–5956