Susana
Moreno Flores
* and
José L.
Toca-Herrera
*
Biosurfaces Unit, CIC biomaGUNE, Paseo Miramon 182, San Sebastian, Spain. E-mail: smoreno@cicbiomagune.es; jltocaherrera@cicbiomagune.es
First published on 3rd September 2009
Atomic force microscopy (AFM) is in its thirties and has become an invaluable tool for studying the micro- and nanoworlds. As a stand-alone, high-resolution imaging technique and force transducer, it defies most other surface instrumentation in ease of use, sensitivity and versatility. Still, the technique has limitations to overcome. A promising way is to integrate the atomic force microscope into hybrid devices, a combination of two or three complementary techniques in one instrument. In this way, a comprehensive description of molecular processes is at hand; morphological, (electro)chemical, mechanical and kinetic information are simultaneously obtained in one experiment. Hereby we review the recent efforts towards such development, describing the aim and the applications resulting from the combination of AFM with spectroscopic, optical, mechanical or electrochemical techniques. Interesting possibilities include using AFM to bring optical microscopies beyond the diffraction limit and also bestowing spectroscopic capabilities on the atomic force microscope.
![]() | Susana Moreno Flores graduated in Chemistry from the Complutense University, Madrid. This was followed by her doctorate on polymer dynamics using dielectric relaxation spectroscopy at the same university, which she finished in 2001. Between 2001 and 2003 she did postdoctoral research on the surface properties of polyelectrolyte monolayers at the Ecole Normale Supérieure, Paris. Between 2003 and 2006 she was a postdoctoral fellow at the Max-Planck Institute of Polymer Research, where she gained experienced with scanning probe microscopies in the group of Prof. H.-J. Butt. In 2007 she joined the group of José Luis Toca-Herrera at CIC biomaGUNE. |
![]() | José Luis Toca-Herrera graduated in Physics from the University of Valencia, in 1993. In 1994 he did one-year research at Max-Planck Institute of Polymer Research. From 1996 to 1999 he completed his PhD at the Max-Planck Institute of Colloids and Interfaces under the supervision of Helmuth Möhwald. He has held Postdoctoral positions at the Technical University-Berlin, the University of Cambridge, and at the Center for Nanobiotechnology (BOKU-Vienna). In June 2004, he joined the Rovira i Virgili University as Ramón y Cajal Researcher. In 2007 he moved to CIC biomaGUNE as group leader. Since June 2008 he is I3 research professor. |
The atomic force microscope certainly resembles the human analog as a blind microscope that can sense micro-and nano-objects (Fig. 1). Indeed, the instrument provides the probing stick, a microsized cantilever with a tip at its free end; a piezodriven device to move the probe over the sample (or vice versa) in three dimensions with nanometre precision; a means to get the tip position during its movement and a feedback mechanism to control how strongly the tip slides or taps over the sample surface. In this way, atomic force microscopy (AFM) has become an invaluable technique to explore the morphology of the nanoworld. Moreover, since the cantilever is a force transducer, it has been extensively used to study surface and molecular interactions.
![]() | ||
Fig. 1 AFM imaging basics as a blind microscope analog. By blindly inspecting an object just underneath we systematically move our body and stick right and left, and up and down. We can either slide or tap our stick over the surface. Our brain controls the movement and the force we exert on the surface through our hand and stick. AFM resembles the analog with a cantilever as an elastic stick, a piezoscanner that moves it along three dimensions, a laser, and a position detector that registers its position. The ways the tip can scan the surface are more numerous than we humans have, however, on most occasions the tip is either made to slide over the surface, impinging a defined force (contact mode imaging) or to tap the sample more or less gently (intermittent contact imaging). The operation is electronically controlled to ensure the scanning is being executed under the desired scan conditions. Finally, a computer processor presents the result as an image. (From ref. 31). |
In spite of its many advantages, the atomic force microscope as a stand-alone tool retains some limitations. Spectroscopic identification is beyond its reach, as are determination of contact areas or absolute distances between probe and sample. Combining the atomic force microscope with complementary techniques to build hybrid instrumentation promises more than one way to overcome these limitations. Synergy may result from such combinations. Optical microscopy bestows on AFM the eyes for precise tip positioning in microstructured materials or cells; interferometry confers the capability of measuring absolute distances, and Raman spectroscopy, the label-free chemical identification of species. Reciprocally, AFM coupled with model-based techniques such as ellipsometry, surface plasmon resonance or even quartz crystal microgravimetry is of invaluable assistance in model validation, the calculation of adsorbed mass, the estimation of trapped solvent or assessing the effect of sample heterogeneity. We will briefly summarize the state-of-the-art developments of such combined techniques.
Modification of the AFM setup to be coupled to an inverted or an upright microscope requires the creation of a obstacle-free optical path to the AFM cantilever and the substrate to be analyzed (Fig. 2). Thus, the configuration of cylindrical piezoscanners mounted upright on tips or on samples is avoided. In particular, tip-scannable atomic force microscopes are preferable when simultaneous optical and AFM imaging is required. In the case of inverted microscopes, a considerable effort in AFM design is needed so as to accommodate optical condensers of high numerical aperture (0.55) without compromising the quality of optical imaging or instrument stability.
![]() | ||
Fig. 2 Left: Scheme of the combination of a transmitted light inverted optical microscope with the AFM. To ensure simultaneous optical and AFM imaging, the AFM tip is allowed to move in three dimensions on a static stage. Right: Optical and AFM image overlay of a breast cancer cell cluster (MCF7). Bar = 25 μm. Vertical scale of the AFM image from grey to white: 10 μm (author's contribution). |
![]() | ||
Fig. 3 Left: Combined AFM and confocal imaging of mouse embryonic fibroblast. (a) and (b) are AFM topography and deflection images; (c) confocal imaging of cells exhibiting fluorescently labelled organelles (blue, nuclei; green: actin filaments; red: clathrin); (d) overlay of confocal and AFM imaging (reproduced from ref. 27 with kind permission from Springer Science + Business Media). Right: High-resolution imaging of focal adhesion points of de-roofed cells, with actin labelled in red (A) and YFP-paxillin (B, green). The partial disruption of the architecture is shown in the AFM (D) and f-actin-labelled confocal image, but not in the paxillin image, indicating that paxillin localization is more proximal to the focal adhesion while f-actin is localized beyond in the membrane-distal half. (From ref. 12, reproduced with permission of the Journal of Cell Science). |
An interesting application of the combined AFM-TIRF technique has been applied to live cells. Cells adhere to affinity substrates by developing so-called focal adhesion points, arrangements of proteins that act as anchors. TIRF microscopy (TIRFM) can map the location of these adhesion points while the AFM cantilever exerts forces on certain cell positions. In this way, force transmission from the apical membrane to the basal membrane of cells has been detected from variations in the number and arrangement of focal adhesion points in TIRFM images.29 At the molecular level, TIRFM and AFM have been successfully correlated to ascertain the morphology of myosin self-assembled filaments in a recent study3 (Fig. 4).
![]() | ||
Fig. 4 Left: TIRFM (a) and AFM (b) images of three miosin filaments. (AFM image obtained in intermittent contact mode). The inset of (a) shows the optical microscope point spread function (Gaussian fitting, standard deviation = 260 nm). Right: TIRF intensity as a function of the height measured by AFM. The intensity is a linear function of the height, which indicates that the fluorophores, situated at the heads of the myosin molecules are arranged around the filament, instead of being embedded. (Reproduced from ref. 3 with permission from Elsevier). |
![]() | ||
Fig. 5 (a) AFM-RICM setup. The tip is typically a microsized (10–20 μm) glass bead attached to the cantilever. The incoming light I0 is reflected at the substrate–medium interface (I12) and at the bead–medium interface (I23). (b) The resulting interference pattern of the glass bead. From the distance and the intensity distribution of this image is it possible to calculate the glass bead profile and the bead–substrate distance (reproduced from ref. 39 with permission from Elsevier). (c) Evolution of the interference pattern while an AFM cantilever is approaching, in contact with, and withdrawing from a substrate (reproduced from ref. 16 with permission from Elsevier). The jumps and discontinuities of the AFM cantilever are visualized in RICM as fringe displacements. (d) Deformation of a polyelectrolyte multilayer capsule. The interference pattern shows the extent of deformation while the AFM exerts an ever-increasing force on the capsule (reproduced from ref. 9 with kind permission of the European Physical Journal). |
TERS is a near-field microscopy that makes use of an apertureless probe to enhance the Raman signal exerted by a sample. The probe is a sharp metallic (e.g., Ag or Au) tip, typically a metal-coated AFM tip or an electrochemically etched metal wire, that is irradiated along its apical axis by a low-power laser at a visible wavelength (500–650 nm). When this tip is placed sufficiently close (approx. 1 nm) to the sample, field-enhancement occurs leading to molecular excitation of the nearby sample area and thus local Raman spectra can be obtained. The mechanism of this enhancement is still an issue under discussion, however it is agreed to be intimately connected to the tip material and tip geometry and it is generally believed to have a double-nature: electromagnetic and chemical.36,38 The electromagnetic enhancement has to do with the excitation of metal plasmons at the tip apex by the laser, with the tip acting as an antenna. The generated field has a reduced spatial range and can only excite molecules that are 1–20 nm apart in the vertical direction and 20–50 nm apart in the lateral direction. The chemical enhancement (also called charge transfer) is based on molecular interactions between the metal tip and the sample surface that alters the spectroscopic properties of the latter. In this regard, TERS presents an additional advantage with respect to the classical bulk or surface-enhanced Raman spectroscopy (SERS) since Raman spectra can even be obtained from poor Raman scatterers, chemical specimens that otherwise exhibit very low or undetectable Raman signals.
The most common TERS setups are shown in Fig. 6. TERS can be operated either in reflection (especially convenient for non-transparent samples, Fig. 6a)33,34 or in confocal mode (Fig. 6b),36,41 which refers to its combination with optical reflection and confocal light scanning microscopes, respectively. In both cases, a p-polarized laser is focussed on the tip so that the polarization plane is kept parallel to the tip axis. Laser focussing is done by means of an optical objective (usually 50× or 60×) which is also used to collect the Raman radiation from the sample. In the reflection version a notch or edge-filter (NF or EF) is usually placed before the spectrometer to collect only the backscattered radiation (Fig. 6a). In the confocal version, an additional beam splitter (BS) and a flipping or folding mirror (FM) are placed before the spectrometer, which redirect both the laser source and the sample emission to a photomultiplier. In this way it is possible to switch rapidly between confocal imaging and Raman spectroscopy.
![]() | ||
Fig. 6 Tip-enhanced Raman spectroscopy (TERS) in reflection mode (a) and confocal mode (b). |
TERS has been validated through its application to model systems such as dye molecules and carbon nanotubes, which both exhibit characteristic and well-known Raman signals.33,38 As far as biomaterials are concerned, studies have been accomplished on nucleic acids and biofilms such as alginates or bacterial surfaces.36 Recently TERS has been said to be the only technique suitable for the study at the nanoscale of subsurface features of thin polymer films41 (Fig. 7, right). An important and common issue in all these studies is the achievement of a high tip-enhancement contrast, defined as the ratio of the Raman signal intensity when the tip is approached, to that when the tip is withdrawn from the sample33 (Fig. 7, left).
![]() | ||
Fig. 7 Left: Effect of tip-enhancement on the Raman spectrum of Brilliant Cresyl Blue, using (a) a gold-coated AFM tip or (b) an etched bold wire (reproduced from ref. 33 with kind permission of the European Physical Journal). Blue traces correspond to the withdrawn probe, while red traces to the engaged probe. Right: Chemical identification of film constituents. (c) AFM topography image of a blend copolymer film composed of polystyrene (PS) and polyisoprene (PI) showing circular protrusions. (d) Sequence of Raman spectra collected at the positions depicted in (c). (e) Raman spectra of pure PS and PI. (f) Magnified view of spectrum collected in background positions. Protrusions are shown to be PS-enriched, on a PI-enriched background (reproduced from ref. 41 with kind permission of the European Physical Journal). |
A still pending issue of TERS is finding the optimal tip preparation technique that ensures good reproducibility and high enhancement.33,41 On the other hand, the application of TERS to complex biological samples such as biofilms is still cumbersome work that involves proper spectral interpretation and makes it difficult to locate characteristic bands that are sample-specific.36 Tip-induced interference, such as carbon contaminations or far-field effects usually appear as large background signals and they should also be properly addressed.
The geometric complementarity of both AFM and SPR methods greatly eased the issue: the AFM needs to be mounted on top of the sample surface through a tip 3D-moving configuration, while the Kretschmann configuration of the SPR setup reaches the sample from underneath (Fig. 8, left). Coupling both techniques, it was possible to monitor in situ the kinetics of the hydrolysis of biodegradable polymer films5,37 and more recently the electropolymerization and deposition of the conductive poly(3,4-ethylenedioxythiophene) (PEDOT) on gold electrodes.2 In this work, cyclic voltammetry and potentiostatic electropolymerization were performed together with SPR to trigger the electrochemical reaction and monitor the oxidation current involved in polymer formation and deposition (Fig. 8, right).
![]() | ||
Fig. 8 Left: Instrumental configuration for AFM + SPR. AFM and the optical setup reach the sample from above and below, respectively. SPR requires a Kretschmann configuration. Right: Simultaneous AFM, cyclic voltammetry and SPR data during the electropolymerization of PEDOT by a potential cycle from 0 to 1.2 V (reprinted with permission from ref. 2. Copyright 2006, American Institute of Physics). |
![]() | ||
Fig. 9 Left: Instrumental configuration for AFM + ellipsometry. Null ellipsometry requires a polarizer (P) and a compensator (C) in the laser arm and an analyser (A) in the detector arm. Right: AFM and SNEM images of a polycrystalline film of a thermotropic liquid crystal (gray scales 0–1.12 µm (a), 0–265 mV (b), 0–1.14 µm (c) and 0–42 mV in d) (reprinted with permission from ref. 20. Copyright 2001, American Institute of Physics). |
Ellipsometry detects the changes of polarization of incident light when it is reflected from a surface. The polarization change is expressed in the so-called ellipsometric angles, which are related to the optical properties of a thin film , i.e. thickness and refractive index, through the fundamental equation of ellipsometry.
Using metal-coated tips for the AFM and configuring the optical setup in order to produce total internal reflection on the sample backside surface, topology and SNEM images of thin polymer films were simultaneously obtained (Fig. 9, right). SNEM actually maps reflected light intensities from the sample backside surface rather than ellipsometric angles, and thus the origin of the optical contrast in the SNEM images is, to date, not fully understood.20
Combining the two techniques in one instrument is thus highly convenient for the interpretation of QCM data. The setup demands the simultaneous use of the QCM sensing electrode as substrate for AFM investigations (Fig. 10a). Care must be taken in the application of oscillation amplitudes, which should be below the required lateral resolution for AFM experiments.13 The coupled techniques have been applied to the study of ferritin adsorption, a quasi-spherical protein that has been shown to undergo a small deformation when adsorbed on gold. This conclusion is sustained through comparison of surface coverages obtained from AFM data, with QCM and finite element simulation data.19
![]() | ||
Fig. 10 Scheme of the experimental setup. (a) Two-in-one: combined AFM and QCM; (b) Three-in-one: combined AFM, QCM and electrochemistry (also called electrochemical QCM or EQCM). The setup is compatible with either tip-moving AFM (tip on the scanner) or sample-moving AFM (sample on the scanner). RE refers to reference electrode, CE to counter electrode and WE to working electrode (substrate). Inset in (b): top-view of the substrate, showing the configuration and geometry of the electrodes. |
One step beyond in instrumentation involves the three-in-one versions, where the sensing substrates are also employed as working electrodes to perform in situ electrochemistry4,13,18 (Fig. 10b). Again, the simultaneous application of electrochemical (approx. 0.03 Hz) and piezodriving potentials (the latter about 4.7 MHz) is possible since both electrical signals widely differ in frequency. This modality has been proved to be particularly useful in studying the electrodeposition of metals on platinum and gold electrodes by cyclic voltammetry,4,14,18 where the ratio of frequency to damping shifts could be attributed to differences in surface roughness as a consequence of metal deposition (Fig. 11). In particular for copper electrodeposition, the mass of trapped liquid has been estimated.4
![]() | ||
Fig. 11 Left column: Silver deposition on gold. Right column: Copper deposition on gold. Top image: AFM topography image showing the effects of metal deposition. Evolution of the electrochemical potential (first plot), 3rd overtone frequency and damping shifts measured with QCM (second and third plots, respectively), and current and time derivative of the frequency shift. For copper, bigger granules are deposited and the higher roughness detected with AFM may help to interpret the higher damping values obtained for this system as compared to those obtained for silver. (Reproduced with permission of ECS – The Electrochemical Society from ref. 13). |
The CPD, VCPD, is the difference of the work functions of two conductive materials, which are brought into contact. The magnitude varies with temperature, but also with the presence of a third material between the contacts, such as an oxide layer, embedded dopants or adsorbed molecules.
To measure the CPD an oscillatory and a bucking bias voltage are applied between the probe, a gold-,32,35 chromium/platinum-24 or a tungsten-coated tip,40 and the substrate, forming a capacitor. The ac voltage frequency, ωel, typically differs from that of the piezoinduced vibration of the cantilever for a better detection of both signals. As a consequence, the cantilever starts to oscillate. The resulting force has three different contributions, a bias and two oscillatory terms at frequencies of ωel and 2ωel, respectively, as the formula shows
![]() | ||
Fig. 12 Left: Scheme of the basic principle of KPFM. A sinusoidal [Vac![]() |
Developed as a combined technique in 199132KPFM has rapidly improved in versatility and lateral resolution. The technique was initially applied to the characterisation of metal surfaces32 at low, micrometric resolution. The application to organic films came shortly after, with the first characterizations of films of mixed amphiphilic molecules,8 self-assembled monolayers26 and a biological membrane.21 Recent years have witnessed the use of KPFM in characterising chemically lithographed self-assembled monolayers35 and the substantial increase of its lateral resolution to the submicron range. Examples of that are contact potential maps of films of pulmonary surfactants24 or gap structures in microelectrodes10 (Fig. 12). New applications of KPFM in biotechnology have abounded and developed in parallel. In these cases, KPFM is often presented as a non-destructive, label-free method for the localization and detection of proteins and nucleic acids and thus as an alternative technique to the classical fluorescence and mass spectrometric methods in proteomics.23,40Cell membrane characterization of living cells7 and single-molecule detection25 are among its most recent applications.
This journal is © The Royal Society of Chemistry 2009 |