AFM Tutorial
Within the past decade, a family of powerful surface imaging techniques, known collectively as scanned probe microscopy (SPM), has developed in the wake of the invention of the scanning tunneling microscope (STM). Each scanned probe technique relies on a very sharp probe positioned within a few nanometers above the surface of interest. Some combination of probe and/or substrate positioning is required to provide sub-nm-resolution, three-dimensional motion of the probe relative to the substrate. When the probe translates laterally (horizontally) relative to the sample, any change in the height of the surface causes the detected probe signal to change. In general, if the probe signal decreases, this means that the point on the surface directly beneath the probe is farther from the probe than the previous point was. Conversely, if the probe signal increases, then the point on the surface is closer to the probe than the previous point.
The electronic circuit that controls the vertical position of the probe relative to the sample uses these changes in the probe signal as sensory feedback to decide which direction (up or down) to move the probe to maintain a constant probe signal. When the probe signal decreases, the circuit realizes that the surface is now farther away, so it moves the probe down until the signal increases to the same level that was measured at the previous point. Similarly, the circuit responds to increases in probe signal by moving the probe up, away from the surface, until the signal decreases back to the desired level. The distance that the probe is moved up or down to return the probe signal to the desired value is therefore related to the height at each point.
Figure 1. General principles of a scanned probe microscope.
To create a 3-dimensional map of surface height, the probe is scanned horizontally (let's say parallel to the x axis) along a series of parallel lines, while the height at each point along each line is recorded. Each scan line is displaced laterally from the previous line by a small distance along the y axis, so that, after several hundred lines have been scanned, the total displacement in the y direction is equal to the length of each line in the x direction, and a square region of the sample has been measured. We now have a measure of height at each point in a 2-dimensional region of the sample, or z(x,y), which is exactly analogous to a topographic map of (if the surface is very rough, for example) the Grand Canyon, except that the dimensions are scaled down by a factor of a trillion (1012) or so!
Probe signals that have been used to sense surfaces include electron tunneling current, interatomic forces, photons, capacitive coupling, electrostatic force, magnetic force, and frictional force. In two prominent cases (STM and atomic force microscopy, AFM), the probe signals depend so strongly on the probe-substrate interaction that changes in substrate height of as little as 0.01 nm can be detected. In addition, the STM and AFM probes can interact with regions of the substrate that are of atomic-scale lateral dimensions, which allows the substrate height to be measured with sub-nm lateral resolution as well. It was the ability of the STM to image individual atoms on surfaces that won the inventors of the STM (Gerd Binnig and Heinie Rohrer of IBM Research in Zurich) the Nobel Prize for Physics in 1986.
Figure 2. Examples of atomic resolution images acquired with scanned probe microscopes: a.) STM of silicon [Si (111) 7x7 reconstruction; 2x2 nm]; b.) AFM of graphite (Highly Oriented Pyrolytic Graphite; 3.5x3.5 nm).
The Scanning Tunneling Microscope
STM image information is derived from measurements of the electron current that can flow when two electrodes, one a sharp metal tip, and the other a relatively-flat, conducting sample, are brought to within about one nanometer of each other. When the two electrodes are so close together (a few atomic radii), electrons can pass from one electrode to the other by tunneling through the potential energy barrier (think of it as a wall) that normally confines them inside each electrode. Electrically biasing the tip electrode relative to the sample allows more electrons to travel in one direction than in the other, so a net current flow (which can be measured) is established. This is our probe signal, the tunneling current.
The magnitude of the tunneling current is a very strong function of the distance (we'll call it s) between the probe tip and the sample. In fact, the current is so sensitive to the probe/sample spacing that, under normal operating conditions, it changes by about a factor of ten for a change of only 0.1 nm in the separation distance (a distance smaller than the radius of a single atom!). Now think about this: The probe tip is made of atoms, and, if the tip is sharp, chances are good that one of the these atoms sticks out a little farther than the others. Therefore the tunneling barrier is thinnest right below this protruding atom so it is much easier for electrons to tunnel between the tip and the sample at this point than anywhere else. Consequently, the majority of the tunneling current will flow through this single atom, allowing us to sense changes in the sample height with atomic-scale lateral resolution.
In practice, the tip is first approached toward the sample until a tunneling current is detected, at which point a constant current feedback loop is turned on. The feedback circuit responds to changes in the current and varies the voltage applied to the tip-positioning piezoelectric element until the current reaches the desired value (the set current). When the tip is moved laterally to a new position above the sample, the current will change if the tip-to-sample distance changes. The control unit then moves the tip up or down until the current matches the set current, which is equivalent to restoring the tip-to-sample distance to its previous value. Since the tip is moved to the same tip-to-sample distance above each point on the surface, the STM is actually tracing out a replica of the surface topography, as is shown in Figure 3. A record of the voltage applied to control the tip height at each point can therefore be converted into a constant current image of the topography of the surface.
An alternative mode of STM imaging, referred to as the constant height mode, uses the tunneling current directly as a measure of topographic changes. The scan rate of the STM is increased, while the gain of the constant current feedback loop is reduced to the point that the controller cannot respond to the current changes induced by individual features on the surface. The current changes themselves are then recorded as the image information. This mode has the advantage that images can be acquired more quickly, however, current changes do not provide as direct a link to feature heights.
Figure 3 Basic concepts of STM imaging. Tunneling current (Itun) is exponentially dependent on distance between tip and sample (usually ² 1 nm). As the tip is moved from x1 to x2, the current increases as the tip-to-sample distance decreases due to the change in sample height. The increase in current causes the control loop to move the tip away from the sample until the error signal is again zero. Recording the value of the tip height (z) as a function of position (x,y) allows the 3-dimensional topography to be reconstructed.
Problem 1.
The tunneling current at a tip bias of 0.1 volt is found to be given by , where I0 and s0 are constants , and s is the distance between the tip and the sample. For I0 Å10 microamperes, ands0 = 0.0434 nm, plot log10(Itun) vs s , and find the distance between the tip and sample at which the tunneling current is 1 nanoampere. What is the effective electrical resistance of the tunneling barrier when the distance between the tip and sample is 1 nm? (clue: remember Ohm's Law)
THE ATOMIC FORCE MICROSCOPE
Instead of tunneling current, an atomic force microscope (AFM) senses interatomic forces that occur between a probe tip and a substrate. The AFM probe tip, which is similar in purpose to the stylus on a phonograph (you remember phonographs, don't you?), is normally integrated into a microfabricated, thin film cantilever. Once contact between the probe tip and the sample surface has been established, the sample is translated laterally relative to the probe tip, while the vertical position of the cantilever is monitored. Variations in sample height cause the cantilever to deflect up or down, which changes the position sensor output, thus generating the error signal that the feedback circuit uses to maintain a constant cantilever deflection (constant force). Normal imaging forces are in the 1-50 nanonewton range (1 newton is little less than 1/4 pound, after cooking), and cantilever deflections of less than 0.1 nm can be detected.
Detection of cantilever deflection
Given that most of the positioning, vibration isolation, and feedback hardware used for STM can be adapted for use in any scanned probe instrument, the additional instrumental requirement for AFM is only the ability to detect sub-nm deflections of a cantilever. Unlike STM, in which the probe signal is always detected with a preamplifier in series with the tunneling current, the AFM probe signal can be derived from several possible methods for detecting cantilever deflection.
Tunneling sensor:
The first realization of atomic force microscopy relied on an STM tip to establish a tunneling current to the back side of the cantilever so that deflections due to cantilever-substrate forces could be detected.1 Although a tunneling sensor is very sensitive to cantilever deflection, this method proved to be very inconvenient to implement, and the STM-based sensing technique was quickly superseded by simpler methods.
Optical lever:
First, and still foremost among alternatives to tunneling detection is the optical lever scheme.2 A laser beam is trained on the back surface of the cantilever, and the reflected beam is sent to a photodiode that is divided into two sections, A and B. Due to the macroscopic length of the reflected light path, any deflection of the cantilever causes a magnified lateral displacement of the reflected laser spot on the photodiode. The relative amplitudes of the signals from the two segments of the photodiode change in response to the motion of the spot-the difference signal (A-B) is very sensitive to cantilever deflection; detection of deflections of