Article pubs.acs.org/Langmuir
Structure and Volta Potential of Lipid Multilayers: Effect of X‑ray Irradiation S. K. Ghosh,*,†,⊥ B. Salgin,*,‡ D. Pontoni,§ T. Reusch,† P. Keil,‡ D. Vogel,‡ M. Rohwerder,‡ H. Reichert,§,∥ and T. Salditt*,† †
Institute for X-ray Physics, University of Göttingen, Göttingen, Germany Max-Planck-Institut für Eisenforschung GmbH, Düsseldorf, Germany § European Synchrotron Radiation Facility, Grenoble Cedex, France ∥ Max-Planck-Institut für Metallforschung, Stuttgart, Germany ‡
ABSTRACT: The effect of hard X-ray radiation on the structure and electrostatics of solid-supported lipid multilayer membranes is investigated using a scanning Kelvin probe (SKP) integrated with a high-energy synchrotron beamline to enable in situ measurements of the membranes’ local Volta potential (Vp) during X-ray structural characterization. The undulator radiation employed does not induce any detectable structural damage, but the Vp of both bare and lipid-modified substrates is found to undergo strong radiation-induced shifts, almost immediately after X-ray exposure. Sample regions that are macroscopically distant (∼cm) from the irradiated region experience an exponential Vp growth with a characteristic time constant of several minutes. The Vp variations occurring upon periodic on/off X-ray beam switching are fully or partially reversible depending on the location and time-scale of the SKP measurement. The general relevance of these findings for synchrotron-based characterization of biomolecular thin films is critically reviewed.
■
INTRODUCTION With the advent of synchrotron radiation sources, the structural characterization of organic thin films by hard X-rays has reached an unprecedented degree of accuracy. A wealth of information on macromolecular structure and assembly at interfaces has been derived from reflectivity and diffraction studies using very brilliant and highly collimated synchrotron X-ray beams. Particular interest is devoted to the molecular structure, interaction forces, and thermal fluctuations of model lipid membranes deposited on solid substrates both as single bilayer1−3 and as multilamellar thin films.4−10 Whereas the structural information is provided by Thomson X-ray scattering, the interaction of high energy X-ray photons with the investigated materials is significantly affected by the absorption cross-section, even in samples that appear as essentially phase contrast samples, such as soft matter thin films in hard X-ray beams. Absorption of hard X-ray photons leads to the generation of high energy electrons by the photoelectric and Auger effects. The released energy can lead to bond breaking, generation of free radicals, and electronic excitations, which generally results in an entire cascade of chemical processes. X-ray-induced chemical damage was, for example, studied in semifluorinated self-assembled monolayers, and found to be caused by the electron-stimulated breaking of C−F and C−C bonds.11 The damage in hydrocarbon monolayer was also discussed as an effect of chemical © 2012 American Chemical Society
transformation, such as dehydration followed by cross-linking, branching, double-bond formation, rearrangements, etc.12 Because the absorption cross section for the X-ray photons is much higher in the solid substrate than in the film, dosage and damage are dominated by the electrons generated close to the substrate surface and escaping into the film. With a comparatively large interaction cross section in the film, these photoelectrons become the primary cause for damage in the sample.13 The possible appearance of beam-induced damage in organic thin film samples needs to be taken into account during design, execution, and interpretation of synchrotron X-ray experiments. In practice, this issue is often dealt with by repeating scans on the same sample spot to monitor the structural changes, and by measuring previously unirradiated regions (“fresh spots”) to assess the sample uniformity. For the case of lipid membranes and monolayers, we have drawn the following conclusions from past experiments: (i) Reducing Xray exposure to the actual time intervals of data acquisition by means of fast shutters and avoiding the overhead during scanning of motors and detector readout are the most significant dosage and damage reduction strategies. In addition, 2D detectors should be used for parallel acquisition whenever Received: October 18, 2012 Revised: December 7, 2012 Published: December 11, 2012 815
dx.doi.org/10.1021/la304139w | Langmuir 2013, 29, 815−824
Langmuir
Article
Figure 1. Schematic illustration of the experimental setup. For X-ray reflectivity measurements, the incident angle αi is equal to the exit angle αf in the plane of incidence. The sketch illustrates the case in which the SKP tip allows one to measure the Volta potential near the point of X-ray irradiation. The tip however does not touch the sample surface, nor is it directly exposed to the X-ray beam. Multilamellar lipid membranes are deposited as a highly oriented stack, with the membrane normal oriented along the z-axis. The electron density profile (ρ(z)) of a bilayer is shown schematically.
possible, rather than resorting to time-consuming serial scanning methods based on point detectors. (ii) Raising the photon energy to and above 20 keV leads to a significant reduction of irradiation-induced structural changes possibly because the photoelectrons are generated over a larger volume in the solid due to the reduced stopping power and increased attenuation length for hard X-rays. This reduces the number of electrons escaping into the organic material, thus opening a larger time window for X-ray exposure. (iii) Fully hydrated samples immersed in buffer solution are much less prone to beam-induced changes than samples hydrated from vapor in air. While a combination of the above strategies is certainly suitable or even necessary to achieve reproducible experimental results that are unaffected by radiation damage, we take a different and more sensitive approach to study X-ray-induced effects, notably by monitoring the electrostatic changes induced in the sample by X-ray irradiation. Long before the structures investigated are altered by bond breaking, pile-up of charges in particular near interfaces is likely to become the single most important cause of beam-induced sample alterations, in particular if the system’s structure is susceptible to electrostatic effects. As in electron microscopy, the transport properties of charge carriers, electric conductivity, and correct sample grounding assume central importance for the preparation of successful experiments. More generally, and beyond the specific topic of X-rayinduced sample modifications, the precise control and accurate measurement of surface/interface Volta potentials can lead to considerable advances in our fundamental understanding of a wide variety of thin film materials in which the electrostatic boundary conditions are fixed by the intervening interfaces. In fact, electrostatic potentials at interfaces play a major role in determining key physical, chemical, and biological properties in several classes of hard and soft materials. Lipid membranes represent a typical example of organic soft interfacial systems whose properties depend strongly on the charge and polarizability of the constituting lipid head groups. In biological systems, membrane potentials are important not only in nerve conduction, membrane fusion, protein, or drug binding,14,15 but also in transport phenomena such as in voltage-gated ion channels. Membrane potentials also affect the functional structure of a wide range of membrane proteins.16
The direct measurement of membrane potentials can be quite challenging. In addition, membrane science suffers from the lack of much needed experimental setups allowing simultaneous determination of structural and electrochemical properties. The SKP, a highly sensitive probe enabling work function measurements, is an ideal tool to satisfy this increasing demand for combined structural/electrochemical sample characterization tools. In the 1990s,17 the range of applications of the SKP has been extended to the fields of electrochemistry and corrosion, where it is mostly used for monitoring the delamination of organic coatings from metallic substrates,18 studying ion transport19 at buried metal/polymer interfaces, and investigating electrochemical reactions under ultrathin electrolyte films and on dry surfaces.20,21 In this work, we have investigated by X-ray reflectivity (XR) and grazing incident small angle X-ray scattering (GISAXS) oriented lipid multilayers formed on hydrophilized silicon and glass substrates. During the X-ray measurements, we monitored in situ the thin films’ Volta potential changes by SKP. After experimental optimization along the lines mentioned above, including in particular the use of high energy X-rays (∼70 keV), no detectable structural changes in the lipid multilayer were observed during several repeated X-ray data acquisitions. At the same time, however, the beam was found to significantly and almost instantaneously alter the electrostatic properties of the biomolecular thin films. The effects are maximum in the region directly exposed to the X-ray beam, but the Volta potential changes spread to unexposed and macrosopically distant surface regions, as detailed in the following sections.
■
EXPERIMENTAL SECTION
Sample Preparation. Dioleoyl-sn-glycero-3-phosphatidylcoline (DOPC) and dioleoyl-sn-glycero-3-phosphatidilserine (DOPS) were purchased from Avanti Polar Lipids (Alabaster, AL) and used without any further purification (purity >99%). Weighted amounts of lipid powders were mixed in glass bottles to obtain stock solutions in chloroform at concentrations ∼10 mg/mL. Polished silicon wafers with ⟨100⟩ orientation (p-type, 1−20 Ω cm) purchased from CrysTec (Berlin, Germany) were cut to a dimension of 15 × 10 mm2. The glass substrates were purchased from Menzel Gläser (Braunschweig, Germany) and cut to the same dimension. The substrates underwent 15 min sonication cycles in methanol (twice) and ultrapure water (three times) followed by ultrapure nitrogen flow drying and a final 4 min air plasma treatment (Harrick Plasma, U.S.). Such a surface treatment applied to the bare Si and glass substrates eliminates any 816
dx.doi.org/10.1021/la304139w | Langmuir 2013, 29, 815−824
Langmuir
Article
residual organic contaminants and renders the thin native oxide surface layer hydrophilic, thanks to the hydroxyl terminating unit (−OH) of the surface silanol groups (Si−OH).22 75 μL of lipid solution was spread onto a substrate, followed by a 2 h period for the bulk solvent to evaporate. The samples were then stored in a vacuum (
