Soft X-ray induced modifications of PVA-based microbubbles in aqueous environment: a microspectroscopy study

Transcription

1 PAPER Physical Chemistry Chemical Physics Soft X-ray induced modifications of PVA-based microbubbles in aqueous environment: a microspectroscopy study George Tzvetkov,* ab Paulo Fernandes, cd Stephan Wenzel, a Andreas Fery, c Gaio Paradossi e and Rainer H. Fink af Received 27th August 2008, Accepted 7th November 2008 First published as an Advance Article on the web 8th January 2009 DOI: /b814946a We use scanning-transmission X-ray microspectroscopy (STXM) for in situ characterization of the physicochemical changes in air-filled poly(vinyl alcohol) (PVA) based microbubbles upon soft X-ray irradiation. The microbubbles were illuminated directly in aqueous suspension with 520 ev X-rays and a continuous shrinkage of the particles with an illumination time/radiation dose was observed. Utilizing the intrinsic absorption properties of the species and the high spatial resolution of the STXM, the modifications of the particles structure were simultaneously recognized. A thorough characterization of the microbubble volume, membrane thickness and absorption coefficient was performed by quantitative fitting of the radial transmittance profiles of the targeted microbubbles. Apart from the observed volume contraction, there was no significant change in the shell thickness. The chemical changes in the membranes were clarified via C K-edge near-edge X-ray absorption fine structure (NEXAFS) spectroscopy. It was revealed that the observed structural alterations go along with a continuous degradation of the PVA network associated with formation of carbonyl- and carboxyl-containing species as well as an increased content of unsaturated bonds. 1. Introduction Polymeric microcapsule and microbubble systems are of broad interest due to a variety of applications, including pharmaceutics, food design and cosmetics. Gas-filled microbubbles are highly effective when used as ultrasound contrast agents due to the large difference in acoustic impedance between the gas in the bubble and the surrounding blood and tissue. 1 5 Also, they could find applications in the encapsulation of therapeutically active gases, similarasmicrocapsulesystems Microbubbles based on airfilled denatured albumin microcapsules, phospholipids, and liposomes containing gaseous SF 6 or perfluorocarbons are currently commercially available. 5 The efficiency in ultrasound imaging and drug/gene delivery requires the microbubbles to have an average diameter of a few microns with a controlled size distribution, sustained persistence in circulation, low gas core solubility and diffusion, suitable shell thickness, stabilization in storage and elasticity for controlling ultrasound damping behavior. In other words, a Friedrich-Alexander Universita t Erlangen-Nu rnberg, Department Chemie und Pharmazie, Egerlandstrasse 3, D Erlangen, Germany. george.tzvetkov@psi.ch; Fax: ; Tel: b Swiss Light Source, Paul Scherrer Institut, 5232 Villigen PSI, Switzerland c Physikalische Chemie II, Universita t Bayreuth, Universita tsstraße 10, D Bayreuth, Germany d Max Planck Institute for Colloids and Interfaces, Potsdam, Germany e Dipartimento di Scienze e Tecnologie Chimiche, Universita` di Roma Tor Vergata, and CNR-INFM-SOFT, Via della Ricerca Scientifica, Roma, Italy f Interdisciplinary Center for Molecular Materials (ICMM), Egerlandstrasse 3, D Erlangen, Germany proper characterization of the microbubble shells is crucial for the creation of optimized microbubble materials. Particularly important is the physicochemical information about the core/shell composition and shell profile of the microbubbles during in situ release of encapsulated gases as a response to external stimuli like, e.g., temperature, pressure, ultrasound and light irradiation. Synchrotron-based soft X-ray scanning transmission X-ray microscopy (STXM) has developed into an established technique for microspectroscopy of soft matter. Combining near-edge X-ray absorption fine structure (NEXAFS) spectroscopy and zone plate imaging, the properties of many micro- and nanostructured materials can be investigated. STXM and full-field transmission X-ray microscopy (TXM) have lately been used to study a number of microcapsule systems, e.g., hollow microcapsules, 12,13 core-shell polymer microspheres, and stimulusresponsive microgels. 17 Very recently, we have demonstrated the potential of the STXM technique to observe and characterize a novel type of biocompatible poly(vinyl alcohol) (PVA)-based microbubbles directly in aqueous solution. It has been shown that STXM imaging below and above the oxygen absorption edge (520 and 550 ev, respectively) can provide unique information on the composition of microbubbles in water. 18 Furthermore, STXM provides higher spatial resolution than the confocal laser scanning microscope (CLSM), at present a widely used technique for the investigation of microcapsule systems, which makes the X-ray microscopy very advantageous for obtaining new insights into the nanoscale assemblage of such materials. Besides, additional sample preparation like, e.g., fluorescence labeling, is not used in STXM, since it utilizes the spectroscopic contrast which allows for quantitative chemical analysis Phys. Chem. Chem. Phys., 2009, 11, This journal is c the Owner Societies 2009

2 Herein, we report on the first quantitative in situ STXM investigation of the structural and physicochemical changes in PVA-based microbubbles and their stabilizing shell upon soft X-ray illumination. The morphology alterations are clarified by quantitative fitting of the radial transmittance profiles while the chemical processes in the polymeric membranes are assessed via local NEXAFS spectroscopy. The information obtained is of significance for the further development of PVA-based microbubbles as therapeutic agents. Moreover, this study is of general importance for the future investigations of polymeric microcapsule and microbubble systems with ionizing beams where radiation induced changes can occur. One of our motivations was also to extend the quite limited knowledge of the interactions of X-rays with complex multicomponent systems consisting of gases in a confined space inside the polymeric membranes. 2. Experimental Stable (air-filled) PVA-coated microbubbles were prepared by cross-linking telechelic PVA at the water/air interface. The synthesis of telechelic PVA, i.e. PVA bearing aldehydes at the chain ends, was previously described in the literature. 19,20 In summary, vigorous stirring at room temperature of telechelic PVA in double distilled water at ph 2.50 by an Ultra-Turrax T-25 at 8000 rpm equipped with a Teflon coated tip was used to generate a fine foam of telechelic PVA acting both as colloidal stabilizer and as air bubble coating agent. The cross-linking reaction was carried out at room temperature by adding sulfuric acid as catalyst and stopped by neutralizing the mixture. Floating microbubbles were separated from solid debris and extensively dialyzed against Milli-Q water. The aliphatic groups of the PVA chains at the air polymer shell water boundary are arranged at the air phase while the hydroxyl groups point toward the aqueous phase. Few unreacted chain ends bearing aldehyde groups at physiological ph are in the form of intramolecular stable hemiacetal tethered to the shell and protruding toward the solution. A sketch of an air-filled microbubble is shown in Fig. 1a. Laser scanning confocal microscopy (CLSM) analysis of fluorescent labelled microbubbles shows the presence of particles with external diameter varying from 3 to 9 mm and an average shell thickness of 0.9 mm. 20 For STXM measurements we used the so-called wet cells where approximately 1 ml of well homogenized microbubbles water suspension was sandwiched between two 100 nm thick Si 3 N 4 membranes (Silson Ltd, UK), which were then sealed with silicone high-vacuum grease to maintain the water environment during the experiment. A representative optical light image of the microbubbles in a wet cell is shown in Fig. 1b. The microbubbles were imaged in transmission mode in helium atmosphere using the PolLux STXM microscope at the Swiss Light Source (SLS), Paul Scherrer Institut. The transmitted photon flux was measured using a photomultiplier tube (Hamamatsu 647P). SLS storage ring runs at 2.4 GeV and top-up operation mode which guarantees a constant electron beam current of ma. The PolLux STXM uses linearly polarized X-rays from a bending magnet in the photon energy range between 200 and 1400 ev and it routinely provides a spatial resolution better than 40 nm. Images were recorded at selected energies through the O 1s region ( ev). The data for the beam-induced changes of the microbubbles were obtained during a normal line-by-line imaging experiment with a dwell time of 1 ms (data integration time per one image pixel). Carbon K-edge NEXAFS spectra were acquired in line mode, i.e., the transmitted signal was recorded while a line trajectory is scanned across the center of a microbubble at each value of the photon energy through the spectrum. The average spectra presented were calculated from 20 different single microbubbles spectra taken from different wet cells. Data processing was carried out using the axis2000 software. 21 Radial transmittance profiles were quantitatively analyzed with a model that considers the photon energy dependent absorption behavior of the shell material, surrounding water and microbubble interior. 22 Profile processing was performed with a homemade software. 3. Results and discussion Two STXM transmission images of eight PVA-based microbubbles in water environment recorded at 520 and 550 ev, i.e. below and above the 1s absorption threshold of oxygen, are presented in Fig. 2a and b. STXM microscopy is based on the contrast given by the absorption coefficients of the species; the transmitted photon intensity through the material is Fig. 1 (a) A sketch of the air-filled microbubble in water environment showing the PVA chains and crosslinks, as well as OH- and aldehyde groups at the polymer shell water boundary (structures not in scale). (b) A representative optical light image (magnification factor: 500) of the PVA-based microbubbles in a STXM wet cell. Fig. 2 STXM transmission images (24 24 mm 2 ) of PVA-based microbubbles in water environment recorded at (a) 520 ev and (b) 550 ev. The white arrow indicates a water-filled particle. This journal is c the Owner Societies 2009 Phys.Chem.Chem.Phys., 2009, 11,

3 dependent on the thickness, density and atomic number of each component according to the Lambert-Beer law. Taking into account the calculated transmission curves in the oxygen K-edge region of water, PVA and air components, 18 it becomes clear that the STXM image at 520 ev (see Fig. 2a) shows the PVA shells of the microballoons while the water and air absorption is weak compared to the carbonaceous material. Above the oxygen K-edge (550 ev) water environment and PVA demonstrate very strong absorption while the air shows approximately one order of magnitude higher transmission of the X-rays and apparently the gas cores of the particles appear in white. Additionally, the core of the microbubble indicated with an arrow shows essentially no contrast compared to the water background. This unambiguously suggests that the air was released through the membrane and the particle is water-filled. Thus, the contrast variations in the STXM transmission images below and above the O K-edge provide a direct evidence of the microbubble gas interior. The effect of 520 ev X-ray irradiation on a microbubble in aqueous environment was evaluated by taking a series of 20 STXM images (scans) one after the other. The 2D organization of the microbubble in the pristine state (i.e. the fist scan) and after the 5th, the 10th and the 20th scan, respectively, is compared in Fig. 3a d. Taking into account the estimated detector efficiency of 36% at 520 ev and the incident flux (I 0 = photons per second, measured outside the wet cell) an overall photon dose of 670 photons nm 3 ( ev nm 3 ) was delivered to the PVA shell during the whole experimental sequence. The result of irradiation induced particle alteration is noticeable, that is a continuous shrinkage of the microbubble. Noteworthy is the fact that, apart from the severe size decrease, the microbubble is still air-filledinthe20thscanaccordingtothecontrastatthestxm image of the same particle at 550 ev (see Fig. 3e). The radial transmittance profiles of the microbubble extracted from the STXM images are compared in Fig. 3f. In the presented curves the intensity minima correspond to the microbubbles polymeric walls while the inner parts display the higher transmission of the X-rays through the air-filled interior. According to the quantitative fitting (see below) of these profiles, in this irradiation experiment the microbubble outer diameter changes from 5.90 to 3.96 mm. The radiation induced microbubbles shrinkage upon 520 ev X-ray illumination was further evaluated for ca. 20 microbubbles with a different initial size (results not shown). Despite the different initial diameters, all microbubbles demonstrate a continuous size decrease with irradiation time. Moreover, the overall volume reduction, shell thickness and absorption characteristics of the membranes exhibit the same trends. With the aim to understand in details the changes of the PVA membrane and the microbubble interior upon X-ray irradiation, a quantitative analysis of the STXM images was performed. For that, fitting of the X-ray transmittance profiles to a model where the X-ray beam resolution and a cubic radial membrane absorption function were taken into account, was applied. A detailed description of the fitting procedure has been reported very recently. 22 In summary, the model is based on the Lambert-Beer expression for the transmitted monochromatic X-rays through a three-component system (encapsulated air, PVA-based shell and the surrounding water in the wet cell): T ¼ I ¼ e k ðy;eþwaterd Water k ðy;eþshell D Shell k ðy;eþair D Air ð1þ I 0 where I and I 0 are the transmitted and incident beams intensity, k (y,e) is the absorption coefficient depending on the sample chemical composition (y) and the photon energy (E), and D Fig. 3 (a) The 1st, (b) the 5th, (c) the 10th and (d) the 20th STXM scan (12 12 mm 2 ; 520 ev) of a microbubble in water environment. (e) STXM image at 550 ev of the particle shown in (d). (f) Radial transmittance profiles extracted from (a), (b), (c) and (d) Phys. Chem. Chem. Phys., 2009, 11, This journal is c the Owner Societies 2009

4 Fig. 4 Scheme of a gas-filled PVA-based microbubble sandwiched between two Si 3 N 4 membranes in the STXM wet cell. R e and R i are the external and internal radii, respectively, h is the shell thickness; the corresponding radial transmittance profile is shown below. the thickness of the absorbing material. Also, the model assumes a perfect spherical microbubble shell (placed between the Si 3 N 4 membranes of the wet cell) defined by an external radius R e, internal radius R i and shell thickness h = R e R i,asshownin Fig. 4. By using this fitting procedure, one is able to characterize the microbubbles three-dimensional shape and to determine with very high resolution their radius, shell thickness and absorption coefficient. The experimental transmittance profiles extracted from different image scans of a representative microbubble shown in Fig. 3 and the corresponding fitting curves are compared in Fig. 5. As one can see, the model fits well the experimental radial profiles. From the quantitative analysis of each image in the experimental series (20 scans) we derived the values for the internal (R i ) and external (R e ) radii of the microbubble, shell thickness (h) and absorption (k) changes upon 520 ev illumination. The microbubble radii changes as a function of the scan number are presented in Fig. 6a. Obviously, there is a very good correlation between the decrease of both radii; the overall reduction of R i and R e for the scanning series is 40.1 and 32.9%, respectively. The corresponding change in the PVA shell thickness, shown in Fig. 6b, is found to be less than 60 nm. The latter change is minor in comparison with the radii alteration. The changes of the shell and encapsulated gas volumes, derived from the radii analysis, are plotted against the number of image scans in Fig. 6c. The shell volume shrinks by 62.2% while the internal air volume reduction amounts to 78.5%. Essentially, if one assumes that the volume reduction is associated with air loss, the latter value is the measure of the gas release through the PVA shell as a result of the irradiation series. In Fig. 6d the mean absorption coefficient values for the present series are shown. The PVA-based microbubble wall contains pores smaller than the experimental resolution, where water and air are trapped. 22 This fact is clearly demonstrated by the much lower absorption coefficient obtained in the minimization procedure (0.21 mm 1, after the first scan) than that of dry PVA. There is a 33% increase in k between the first and the last image scan. Qualitatively, taking into account k values of air, water and PVA at 520 ev ( mm 1, 0.11 mm 1 and 0.92 mm 1, respectively), 22 the shell volume decrease and the absorption increase revealed by the analysis can be associated with the squeeze-out of water and air from the shell. However, assuming that PVA mass is conserved during the illumination series and by neglecting the air content in the shell, it appears that the absorption coefficient of the shrunk wall after the 20th scan should be 0.37 mm 1 instead of Fig. 5 Experimental radial transmittance profiles derived from (a) the 1st, (b) the 5th, (c) the 10th and (d) the 20th STXM scan at 520 ev of the microbubble shown in Fig. 2 (open circles) as well as the corresponding fitting curves. This journal is c the Owner Societies 2009 Phys.Chem.Chem.Phys., 2009, 11,

5 Fig. 6 Structural values derived from the fits of the radial transmittance profiles of the irradiation series of the microbubble shown in Fig. 3. (a) External (R e ) and internal (R i ) radii, (b) shell thickness, (c) shell and gas volume, and (d) mean absorption coefficient of the shell as a function of the scan number. The typical error bars are also shown. the derived value of 0.28 mm 1 (see Fig. 6d). We associate this discrepancy with PVA mass loss during irradiation and/or different PVA absorption due to the chemical changes. In order to examine the possible chemical processes associated with the alterations of PVA-based membranes upon photon illumination, X-ray absorption spectroscopy of the exposed microbubbles was performed. C K-edge NEXAFS spectra obtained from microbubbles in the pristine state and after the 10th and the 20th scan are displayed in Fig. 7. The dominant signatures in the spectrum of the pristine microbubbles are the peaks at ev attributed to C1s - s* C H transitions and at ev which is due to the C1s - s* C O transitions from C OH bonds in the polyvinyl alcohol molecules. A smaller shoulder at ev is also observed. We associate this feature with the carboxylic carbon (C1s - p* CQO transitions) from the aldehyde groups present in the telechelic PVA. As one can see, the changes in the NEXAFS spectrum with prolonged X-ray irradiation of the MBs are pronounced. First, a gradual suppression of the main resonances of PVA is observed. Second, an intensity increase of the feature at ev is detected and third, new peaks at 285.3, and ev are developed. Obviously, the evolution of the absorption spectra demonstrates a severe chemical modification of the microbubble membranes as a result of the X-ray illumination. The latter is in accordance with the already reported absorption coefficient changes (see Fig. 6d). Taking into consideration the existing wealth of C K-edge NEXAFS data on different polymers, we can interpret the spectral changes in Fig. 7 as follows. The intensity increase of the relatively broad feature at ev (see Fig. 7) is due to the abundance of carbonyl groups (ketons and aldehydes) in the species originating from the interaction of 520 ev X-rays with telechelic PVA. The appearance of the peak at ev can be assigned to the C1s - p* CQC transitions associated with the presence of unsaturated bonds. Fig. 7 Carbon K-edge NEXAFS spectra extracted from microbubbles after (a) the 1st, (b) the 10th, and (c) the 20th STXM scan at 520 ev Phys. Chem. Chem. Phys., 2009, 11, This journal is c the Owner Societies 2009

6 The small shoulder at B290 ev is also due to the transitions to the p* orbitals of CQC and CQO bonds. The prominent peak at ev in the spectrum recorded after the 20th image scan (Fig. 7c) can be assigned to the carboxylic carbon from COOH groups. The observed changes in the chemical speciation of the microbubbles upon X-ray illumination are in accord with the previous studies on PVA interaction with ionizing beams. Beamson and Briggs 26 have shown that 2 nm thick PVA films on different substrates experienced a severe degradation during monochromatized Al Ka (hn = ev) X-ray photoelectron spectroscopy (XPS). A continuous depolymerization of the PVA was observed along with a fall in oxygen content, a decrease in C O functionality and an increase in CH 2,CQO and CO 2 R functionality. Radiation induced degradation of polyvinyl alcohol in aqueous solutions by g-ray irradiation has also been reported. 27 The analysis on the degradation products suggested that through chain scission and formation of ketones/enol, complete mineralization of PVA could be achieved upon g-ray irradiation. Furthermore, during our very recent small angle X-ray scattering (SAXS) experiments on the telechelic PVA hydrogel slabs with the same chemical composition as that of the microbubbles, changes in the composition of the polymeric material have been detected. 28 The compositional breakdown of the polymeric shells by X-rays most likely involves a series of free-radical reactions initiated by electron impact induced bond cleavage. The ejection of energetic electrons from atoms via the photoelectric, Auger and Compton effects can cause breaking of chemical bonds by the radiation. 29 On the other hand, secondary reactions can occur initiated by the resulting free radicals. In aqueous solution the free radicals can be generated directly or through reactions with products arising from radiolysis of water. 29 On the basis of our spectromicroscopy results we assume that the increase of the carbonyl-containing species in the shell material is due to the cleavage of HCO H and H COH bonds. The carboxyl species detected by NEXAFS must result from oxidation of the carbonyl species, possibly by hydroxyl radicals. The mechanism of X-ray interaction with the microbubbles is probably more complex, consisting of numerous steps. For instance, we cannot exclude a radiation induced cross-linking resulting from recombination of radicals on two adjacent PVA chains and a loss of organic material from the membranes (a common phenomenon in soft X-ray microspectroscopy of polymers) 25 into the water medium. However, the STXM does not allow the necessary sensitivity to detect traces of fragments from the shell in the surrounding water. The chemical and colloidal stability of these microbubbles is primarily due to the amphiphilic nature of the PVA (hydrophilic OH groups and hydrophobic methylene groups) and the presence of chain segments projecting into the water medium, as shown in Fig. 1. Since our STXM observations demonstrate severe physicochemical changes of the membranes, it is surprising that there is still air encapsulated in the particles almost two times smaller than the pristine ones. One may assume two possible scenarios to explain the stability of the air bubbles. First, air and water can permeate the microbubble shells but even with the presence of new functional groups the interior is still hydrophobic and water does not wet it. Second, a lower shell permeability is developed (driven by the much denser shell as discussed above) which retains the gas inside the particles. The algorithm developed for quantitative fitting of the radial transmittance profiles of the microbubbles is not sensitive to the encapsulated gas because the air X-ray absorption is very little compared to water or PVA and consequently has almost no influence on the transmitted intensity. 22 In order to describe in details the process of gas leakage, X-ray illumination of PVA-based microbubbles filled with different gases should be performed. For instance, using SF 6 or perfluorocarbons an access to the absorption contrast at the F K-edge ( around 697 ev) is possible. The latter will give a direct evidence for the mechanism of in situ gas transport through the polymeric membrane. Such experiments are considered as part of our ongoing project. 4. Summary and conclusions The microspectroscopy results presented in this paper lead to the following picture of soft X-ray interaction with PVA-based microbubbles. Upon 520 ev irradiation of the microbubbles in aqueous environment a continuous shrinkage of the particles is observed. During the presented irradiation series (accumulation of a photon dose of ev nm 3 ) the shell thickness of the PVA-based shell change is small, but an increase of the absorption coefficient of the shell material (corresponding to its optical density for 520 ev X-rays) of 33% is derived. According to the C K-edge NEXAFS spectroscopy the morphology changes of the particles are accompanied with a continuous degradation of the PVA network and formation of carbonyland carboxyl-containing species as well as an increased content of unsaturated bonds in the supporting shells. It was found that after these substantial chemical and structural transformations of the initial polymeric entities, the microbubbles are still air-filled. Acknowledgements This work was in part performed at the Swiss Light Source (SLS), Paul Scherrer Institut, Switzerland. The research project is funded by the BMBF (contract 05KS7WE1) and through the STREP Systems for in situ Imaging and Healthcare (SIGHT). Experimental assistance by Dr Jo rg Raabe (SLS) is gratefully acknowledged. References 1 K. Ferrara, R. Pollard and M. Borden, Annu. Rev. Biomed. Eng., 2007, 9, N. de Jong, L. Hoff, T. Skotland and N. Bom, Ultrasonics, 1992, 30, E. G. Schutt, D. H. Klein, R. M. Mattrey and J. G. Riess, Angew. Chem., Int. Ed., 2003, 42, Y. Liu, H. Miyoshi and M. Nakamura, J. Control. Release, 2006, 114, S. B. Feinstein, Am. J. Physiol. Heart Circ. Physiol., 2004, 287, J. R. Lindner, Nature Rev. Drug Discovery, 2004, 3, This journal is c the Owner Societies 2009 Phys.Chem.Chem.Phys., 2009, 11,

7 7 C. S. Peyratout and L. Dähne, Angew. Chem., Int. Ed., 2004, 43, Y. Ma, W. Dong, M. A. Hempenius, H. Mo hwald and G. Julius Vancso, Nat. Mater., 2006, 5, G. B. Sukhorukov, A. Fery, M. Brumen and H. Mo hwald, Phys. Chem. Chem. Phys., 2004, 6, A. L. Klibanov, Adv. Drug Deliv. Rev., 1999, 37, M. Winterhalter and A. F. P. Sonnen, Angew. Chem., Int. Ed., 2006, 45, C. De jugnat, K. Ko hler, M. Dubois, G. B. Sukhorukov, H. Mo hwald, T. Zemb and P. Guttmann, Adv. Mater., 2007, 19, K. Ko hler, C. Déjugnat, M. Dubois, T. Zemb, G. B. Sukhorukov, P. Guttmann and H. Mo hwald, J. Phys. Chem. B, 2007, 111, I. Koprinarov, A. P. Hitchcock, W. H. Li, Y. M. Heng and H. D. H. Sto ver, Macromolecules, 2001, 34, G. Tzvetkov, B. Graf, R. Wiegner, J. Raabe, C. Quitmann and R. Fink, Micron, 2008, 39, G. Tzvetkov and R. H. Fink, Scr. Mater., 2008, 59, S. Fujii, S. P. Armes, T. Araki and H. Ade, J. Am. Chem. Soc., 2005, 127, G. Tzvetkov, B. Graf, P. Fernandes, A. Fery, F. Cavalieri, G. Paradossi and R. H. Fink, Soft Matter, 2008, 4, G. Paradossi, F. Cavalieri, E. Chiessi, V. Ponassi and V. Martorana, Biomacromolecules, 2002, 3, F. Cavalieri, A. El Hamassi, E. Chiessi and G. Paradossi, Langmuir, 2005, 21, P. A. L. Fernandes, G. Tzvetkov, R. H. Fink, G. Paradossi and A. Fery, Langmuir, 2008, 24, J. Kikuma and B. P. Tonner, J. Electron Spectrosc. Relat. Phenom., 1996, 82, O. Dhez, H. Ade and S. G. Urquhart, J. Electron Spectrosc. Relat. Phenom., 2003, 123, H. Ade and A. P. Hitchcock, Polymer, 2008, 49, G. Beamson and D. Briggs, Surf. Interf. Anal., 1998, 26, S. Zhang and H. Yu, Water Res., 2004, 38, G. Paradossi, unpublished work. 29 T. Teng and K. Moffat, J. Synchrotron Radiat., 2000, 7, Phys. Chem. Chem. Phys., 2009, 11, This journal is c the Owner Societies 2009