Scanning probe microscopies for the characterization of porous solids: strengths and limitations

J.I. Paredes; A. Martínez-Alonso; J.M.D. Tascón    Instituto Nacional del Carbón, CSIC, Apartado 73, 33080 Oviedo, Spain

1 INTRODUCTION

The term scanning probe microscopy (SPM) comprises a whole class of techniques that measure surface properties of materials on an extremely local scale. Although at the very beginning only topography could be examined, further advances extended the measurement capabilities to other surface properties, such as friction, adhesion, electric and magnetic forces or specific interactions between molecules. The scanning tunneling microscope (STM) was the first instrument of this family to be conceived and put into practice, some 20 years ago [1], for the study of conducting surfaces in vacuum. As a logical sequel to this invention, a few years later the atomic force microscope (AFM) was developed [2], expanding the realm of SPM application to nonconducting materials. Since then, the number of SPM-based techniques has been growing at a rapid and steady pace and currently amounts to a few tens of variants [3,4]. Nevertheless, the bulk of SPM work performed to date has been (and still is) carried out using either STM or AFM in any of its operation modes.

The unique resolution capabilities of SPM are based on the combined use of two elements: an ultrasharp tip, which probes the sample surface and enables their mutual interaction (and therefore the measured property) to be highly localized, and a piezoceramic scanner, which controls the tip-sample relative position in the three spatial directions with subangstrom precision. This approach has allowed the imaging of surfaces with atomic resolution in STM/AFM, as many examples have demonstrated [5]. To attain such resolution level, the sample under study is mainly required to be atomically flat and highly crystalline.

As a result of the development of probe microscopy, many areas of research have experienced a remarkable progress and the variety of materials that have been visualized and characterized by STM, AFM and/or related techniques is correspondingly broad. These include metals [6], semiconductors [7] and superconductors [8], layered inorganic materials [9] and self-assembled monolayers [10] or polymers [11] and macromolecules (including biomacromolecules) [12,13].

In this context, the SPM techniques (and especially STM and AFM) appear, a priori, ideally suited for the direct visualization of the porous structure of materials at scales which are not so readily accessible by other means (e.g., scanning and transmission electron microscopies). However, the performance of such a task is confronted with two major limitations. The first one arises from the fact that detection with SPM is exclusively restricted to the outermost surface of the sample. Accordingly, this implies that only the most external porosity of the material can be probed, whereas no information on the bulk (inner) porosity, which might not be identical to the former, is revealed. The second drawback is related to the finite dimensions of the probing tip, which limits the size of the voids (pores) physically accessible (and thus detectable) by the tip on the sample surface. Obviously, pores significantly smaller than the tip diameter will pass unnoticed to the instrument when the surface is scanned. As a specific example, the tips normally employed in AFM are not sharp enough to provide access to the whole mesopore range (between 2 and 50 nm).

In spite of the mentioned disadvantages, useful information has been obtained from SPM imaging of a number of porous materials. To illustrate such point, the present review examines the research that has addressed the visualization of the porous structure of solids by SPM. A wide variety of materials is covered, such as porous silicon, activated carbon materials, aluminas, synthetic membranes or biological materials.

2 OVERVIEW OF THE SCANNING PROBE MICROSCOPY TECHNIQUE

All SPMs are based on the same principle of operation: a sharp tip and the sample surface of interest are brought to a close proximity (several Å or nm), or to mechanical contact. As a result, a physical interaction is established between tip and sample, which, in turn, gives rise to a measurable signal. The signal usually has a sharp dependence on the tip-sample distance, so it can be employed to accurately control such separation and track the surface topography of the sample under study. This is accomplished by a feedback system, which, as the tip scans the sample, adjusts the tip-sample distance in order to keep the measured signal constant. The relative motion between tip and sample in the three spatial directions is realized by a piezoceramic scanner, whose position can be controlled with subangstrom precision, enabling atomic-scale imaging.

The main differences between the different SPM techniques lie in the type of interaction that is used to control the tip-sample distance. Although the SPM offspring are remarkably numerous [3,4], here we will focus only on STM and AFM, as they remain the most widely used and the best suited for high resolution imaging of surface structures.

2.1 Scanning tunneling microscopy

The STM was the first scanning probe microscope to be developed [1]. Although the original aim was to image the surface topography of metals on the atomic scale and in ultrahigh vacuum, it was later applied for the study of other (conducting) materials [5] and in additional environments (air and liquids). In STM, a bias voltage is applied between the sample and a metallic tip. On approaching both electrodes to just a few Å, a tunneling current is established. This current decreases exponentially with tip-sample separation and is used as the signal to be controlled (i.e., to be kept constant) by the feedback system. Provided that the surface to be studied is flat enough, STM can render images with atomic resolution [5,14]. However, in general such images contain both topographic and electronic information and their interpretation is frequently a difficult task. By contrast, on large distance scales topographic effects usually dominate over electronic ones and the features observed in the images are more straightforward to elucidate [15].

2.2 Atomic force microscopy

The AFM was developed to overcome one of the main limitations of the STM, namely, the necessity of electrically conducting samples [2], and is based on the interaction forces (short- or long-ranged, attractive or repulsive) that exist universally between atoms and molecules. To measure the tip-sample interaction force, the former is attached to one end of a flexible lever, referred to as cantilever [5]. The forces exerted on the tip (~ 10- 9 N [16]) induce a measurable deflection of the cantilever, which is, essentially, the signal to be controlled by the feedback system. Basically, an atomic force microscope can be operated in three different modes: contact, intermittent contact (also referred to as tapping) and noncontact mode [5]. In contact mode, the tip is brought into physical contact with the sample and the repulsive force that originates thereof is used to track the sample topography. In the intermittent contact mode, the cantilever is vibrated at its resonance frequency and the tip allowed to intermittently contact the sample surface. Such interaction reduces the cantilever oscillation amplitude (compared to the noninteracting situation), which is then used as feedback signal [17]. In the noncontact mode, the instrument is operated in the net attractive force regime and the frequency shift arising from the interaction is employed as feedback [18].

2.3 Resolution in STM and AFM

Provided that the sample to be imaged is reasonably rigid, resolution in STM/AFM at the nano- and micrometer scale is fundamentally limited by the shape of the probing tip [19,20]. Thus, a certain degree of distortion will always be present on the images [21,22], so that prominent features on the surface will appear wider than they really are (the blunter the tip, the wider the apparent size). By contrast, depressions in the topography, such as pores, will appear narrower (again, the blunter the tip, the narrower the pore will appear) or will not be detected at all. Such effect has important implications for imaging the porous structure of surfaces by these techniques. Considering the tip radii of curvature of commercial AFM cantilevers (a few tens of nm), the micropore and small mesopore (  10 nm) range will not be accessible by AFM,...