Atomic force microscopy provides eyes, noses, and fingers to explore the world from a molecule’s point of view.

In 1871, the Scottish physicist James Clerk Maxwell dreamed up an imaginary “demon” to help him conduct thought experiments by picking up and arranging atoms one at a time. Maxwell’s Demon (as the creature came to be known) was a tiny gatekeeper who sat at a door between two chambers and violated the second law of thermodynamics by sending all the fast gas molecules into one chamber and all the slow ones into another. Over the past 15 years, scientists have come close to making their own personal Maxwell’s demons to help them see and manipulate atoms and molecules one at a time in the real world. Imagine a demon’s tiny finger scanning a page of molecular Braille type and converting the tactile information into an image for his macroscale master. As the demon sniffs the chemical scents and tingles in the electromagnetic fields, the image takes on colors that indicate phase boundaries, defects, and foreign matter adsorbed on the surface. Equip the demon with a stylus, and he sculpts the surface to your specifications. This particular demon’s name is atomic force microscopy (AFM), and it is rapidly becoming an essential tool for examining everything from computer chips to cartilage.

AFM Basics
AFM is a relative newcomer to the analytical field, but it has been eagerly adopted by materials scientists and biochemists alike for its ability to create three-dimensional images of surfaces and manipulate and modify those surfaces. The original AFM instrument was invented in 1986 by IBM physicists Gerd Binnig and Christoph Gerber working with Stanford University electrical engineer Calvin Quate. Their prototype used a small diamond chip attached to a cantilever made from gold foil. As the diamond tip scanned the surface of a sample, the cantilever moved up and down, producing changes in a tunneling current passing between the cantilever and a second tip overhead. What emerged was a topographical map of the surface with a resolution of about 300 Å—about half the diameter of a typical virus. Tom Albrecht, a Stanford graduate student, came up with a silicon cantilever tip soon afterward that allowed him to record the first atomic-resolution image of an electrically nonconducting sample, boron nitride. He also developed the force-sensing cantilever probes that are now routinely used in AFM.

FIGURE 1: The basic AFM setup relies on a scanning tip and a detector. The tip, attached to the end of a cantilever, scans across the sample surface. A laser beam reflects from the top surface of the cantilever and is detected by a position-sensitive photodiode detector.

Today, AFM tips are commonly made from silicon or silicon nitride, although gold tips are used for some special applications and the first reports of carbon nanotube tips appeared in the literature in 1998 (1). Optical detectors have replaced the tunneling current device, typically using a laser beam reflected off the top surface of the cantilever and recorded by a position-sensitive photodiode detector (Figure 1).

The instrument can be operated in one of three modes: contact, noncontact, and tapping (2). In contact mode, a piezoelectric positioning element keeps the tip in contact with the sample as the tip is scanned over the surface. As the surface pushes and pulls the cantilever, a DC feedback amplifier compares these deflections to a reference standard. The amplifier applies a voltage to the positioning element, which keeps the AFM tip in contact with the sample surface. The resulting image is a map of the variation in voltage with the position of the tip.

Contact-mode AFM is typically performed in ambient air or with the sample and tip submerged in a liquid. In ambient air, water vapor adsorbs to the sample surface, forming a meniscus that pulls the probe toward the surface and distorts the force measurements. Nonconducting samples can accumulate static charges that further distort the measurements, and friction can damage the sample and the tip. Immersing the tip and sample in liquid provides a uniform interface and lessens these interferences, but spills and liquid leaks can contaminate the scanner and the liquid overlayer can damage the sample.

Noncontact AFM is less damaging to fragile or loosely bonded samples. A tip hovers above the sample surface and measures the van der Waals forces between the tip and the sample. The tip oscillates, and an AC detector measures changes in the amplitude, phase, and frequency of the oscillations. Because van der Waals forces are strongest within the first nanometer from the surface, adsorbed fluids can mask these forces and make variations hard to detect.

Tapping-mode AFM is the newest technique, and it solves several of the problems presented by the other modes. A piezoelectric crystal causes the tip to oscillate at 50–500 kHz, with an amplitude of 20–200 nm. The tip is scanned across the surface, but it taps the surface rather than dragging across it. When the tip comes into contact with the surface, its amplitude of oscillation changes, and a feedback loop brings the oscillation back to the reference amplitude. The forces that are measured this way reflect the vertical component rather than the shear components, since the tip does not drag across the surface. Tapping-mode AFM has a large linear operating range, and force measurements are very reproducible. The sample can be under vacuum, in ambient air, or in a liquid. Tapping-mode AFM requires a slower scan speed and more complex instrumentation, but the advantages in resolution and sample preservation outweigh the disadvantages and this technique is catching on quickly.

Tapping-mode AFM also lets you measure the phase lag that results from interactions between the surface and the tip. The resulting phase image (also called a contrast image) contains information on the chemical composition, adhesive properties, friction interactions, and viscoelasticity of the sample. Phase images can be obtained side-by-side with topographic images to reveal surface features that would not show up in either image alone (Figure 2).

Moving Beyond the Basics
Typical AFM scans provide a spatial map of height versus position. The height information is measured as a function of the force exerted by the surface as the tip scans across it. Pressing the AFM tip into the sample surface with a known force and withdrawing it provides a map of force versus distance—a measure of the mechanical and adhesive properties of the surface. Forensic investigators have used this nondestructive method to measure the ages of dried bloodstains on hard surfaces and fabrics (3). Using a calibration curve, investigators can pin down the age of a bloodstain with much greater precision than simple visual examination affords.

Modifying the AFM tip with chemical functional groups creates a “nose”, or chemical force microscope (CFM), that can distinguish between chemical species on a surface. Typically, the AFM tip is coated with gold or platinum, and the organic modifiers are bonded to the metal via a polar functional group such as a thiol. AFM tips coated with enzymes act as biosensors that produce surface maps integrating electrochemical and topographical information. This technique has been used to monitor solute permeability in diseased and healthy cartilage samples in an effort to understand the effects of degenerative diseases (4). Calcium ion concentrations in the range of 10^–4 to 10^–3 M have been measured in dental enamel samples, providing insight into local variations in dissolution activity (5).

Recently, carbon nanotube tips less than 5 nm in diameter have been fabricated and functionalized, producing chemical maps with even greater resolution (6). This kind of imaging has been used to study functional group characteristics on polymer surfaces and molecular binding and recognition in biological systems.

Gerber’s research group at IBM Zurich is taking the CFM technique one step further by dispensing with the tip all together (7). The cantilever itself is used as a sensor to measure femtojoule-scale (10^–15 J) heat dissipation in catalytic reactions and attogram-scale (10^–18 g) mass changes that occur during chemical or biological reactions. The state of the art for this technique is a cantilever 16 nm thick by 80 micrometers long that can measure mass changes to within 10^–22 g (about the mass of one cobalt atom).

It’s a short conceptual jump from sensing chemicals on a surface to mapping out magnetic and electrical fields. Equip an AFM with a ferromagnetic probe, and it becomes a magnetic force microscope. Manufacturers of magnetic recording materials find this method especially useful as they strive for greater data density and fewer defects. Electric force microscopy works on a similar principle and is used to characterize semiconductors, superconductors, and composite materials. Several research groups are working on integrating magnetic resonance imaging with AFM—one goal is to produce MRI scans of single viruses.

From Spectator to Participant
Originally, researchers weren’t thrilled when their AFM tips left tracks across their samples, so noncontact and tapping mode techniques were developed. Somewhere along the line, however, someone realized that you could mark a surface on purpose, and that this might be useful. Nano-engineers are forever looking for better ways of manipulating individual molecules, and what could be better than to push things around using a tiny AFM finger? AFM tips have been used to bend carbon nanotubes into Greek letters and other interesting shapes, using the van der Waals forces between the nanotubes and the substrate to hold them in place (8). Arrays of AFM tips have been used to produce indentation patterns on polymer surfaces, an effective, if slow (32 kbps) method of recording binary data at up to 200 Gbit/in² (9). (A single-probe setup has achieved twice this density but is too slow to be practical.)

In a nanoscale adaptation of present-day magnetic recording tape technology, researchers are manipulating the polarization direction for tiny crystal domains in piezoelectric films using AFM tips. They are still working on ways to produce uniformly sized crystallites at this tiny scale, and they are exploring ways to either eliminate or use “pinning sites”—domains that resist changes in polarization. The domain-orienting approach also works well for "writing" circuitry patterns into polycrystalline films, a form of nanolithography that could produce a new generation of very small computer chips.

As versatile as AFM is, Maxwell’s original demon won’t have to worry about being replaced by a laboratory instrument just yet. AFM and its related techniques can’t sort out gas or liquid molecules yet: They require a solid foundation to anchor things in place. And the laws of thermodynamics still apply.

References

1. Wong, S. S.; Harper, J. D.; Lansbury, P. T.; Lieber, C. M. J. Am. Chem. Soc. 1998, 120, 603–604.
2. Li, H.-Q. Atomic Force Microscopy Student Module. University of Guelph, Ontario; www.chembio.uoguelph.ca/educmat/chm729/afm/introdn.htm.
3. Crisman, E. E.; Gregory, O. J.; Platek, M. J.; Schofield, T. M. Analysis of Dried Blood Samples Using Atomic Force Microscopy. Presented at Pittcon 2002, New Orleans, LA; Abstract 849.
4. Gardner, C. E.; Unwin, P. R.; Macpherson, J. V. Development of Combined Scanning Electrochemical–Atomic Force Microscopy for the Investigation of Transport Processes in Biomaterials. Presented at Pittcon 2002, New Orleans, LA; Abstract 954.
5. Macklam, I. D.; Unwin, P. R.; Macpherson, J. V.; Guo, S. Local Imaging of Reactive Fluxes from Mineral Surfaces Using Scanning Electrochemical Microscopy. Presented at Pittcon 2002, New Orleans, LA; Abstract 1204.
6. VanLandingham, M. R.; White , C. C.; Nguyen, T. Nanoscale Characterization of Surfaces, Interfaces, and Interphases in Polymeric Materials and Systems. National Institute of Standards and Technology, Building and Fire Research Laboratory; http://ciks.cbt.nist.gov/markv/cfm.html.
7. Nano People: Dr. Christoph Gerber; Institute of Nanotechnology, Stirling, Scotland; www.nano.org.uk/personalities5.htm.
8. Nanotube Manipulation. IBM Research, Nanoscale Science and Technology Group; www.research.ibm.com/nanoscience/ manipulation.html.
9. Schewe, P. F.; Stein, B. High Density AFM Data Storage. Physics News Update, Nov 8, 2000, 511 (1); www.aip.org/physnews/update/511-1.html

Nancy K. McGuire is an associate editor with Today’s Chemist at Work. Send your comments or questions regarding this article to tcaw@acs.org or the Editorial Office, 1155 16th St N.W., Washington, DC 20036.