AFM: Not Just for Materials Science Anymore

The atomic-force microscope (AFM) was developed 20 years ago, but only recently has it become a significant tool for biologists.

Karen Heyman(
Dec 4, 2005

Joelle L. Bolt

The atomic-force microscope (AFM) was developed 20 years ago, but only recently has it become a significant tool for biologists. Irene Revenko, applications scientist at Santa Barbara, Calif.-based AFM manufacturer, Asylum Research, says when AFMs first came into biology, most of the experiments were essentially just replicating earlier findings that had been done on other instruments. "In the last five years people began to get new data with AFMs that they weren't getting with other microscopy techniques."

With their "scanning probe" design, AFMs can be used in experiments that map a sample's mechanical compliance (hardness), charge, or magnetic field, as well as in force spectroscopy studies. The tool's primary biological application, however, remains imaging, where its high resolution and ability to image live samples now make it a device worth considering for researchers seeking to combine electron microscope-level resolution with the sample variety afforded by optical microscopes....


Paul Hansma of UC-Santa Barbara is widely acknowledged as the dean of AFM development. "The family of cantilever-based, scanning-probe microscopes is generically called atomic-force microscopes, in honor of the name given them by the inventors of the first such instrument," says Hansma.1 "In order to help people with literature searches, [referees try] to keep the name AFM ... despite billions of variations you can do with cantilever-based instruments."

The Next Generation of Cantilevers

Several academic labs, including Paul Hansma's, are working on smaller cantilevers. "We're going to 10 microns, an order of magnitude smaller than what's commercial," he says. Julio Fernandez of Columbia University has worked with Hansma's prototypes and notes a number of problems, although some cantilevers were good enough that, "We got really fast recordings with them."

Fred Sachs at the University of Buffalo, NY, and his student Arthur Beyder are working on cantilevers in the 20-micron range. With innovative double-hinges at right angles to each other, these cantilevers tilt left and right and up and down independently, allowing an investigator to measure two things at once within the same system. The design could be especially useful for removing drift and other external noise, says Sachs. "You use one axis to measure the position of the substrate, the other to measure the position of the sample on the same cantilever. Now there's no more drift; you can measure for hours." Sachs is now looking for a manufacturer to produce the cantilevers.

Roukes' lab is attempting an even more radical reduction in size. They are aiming for nanoscale cantilevers. Crucially, nanoscale cantilevers would be smaller than the wavelength of light, making laser detection much more difficult. Some AFMs already use an electrical detection system based on piezoresistance (stress-sensitive resistance) rather than lasers, and Roukes' work is an extension of that principle.

Since a patent is pending, Roukes cannot give details, but Fernandez, who is familiar with the concepts, finds Roukes' idea exciting. "If they successfully develop these devices, and the means to detect their deflection," says Fernandez, "It has the potential for impacting physics and biology quite extensively."

A cantilever is like a tiny diving board, made of silicon or silicon nitride, with an upside-down pyramid-shaped tip at the end. The cantilever's dimensions set the force resolution, while the tip's radius governs the imaging resolution. The system noise floor (inherent noise) is also a limiting factor.

"Any noise that contributes to cantilever deflection is a fundamental limit on the force that you can resolve or on the scale of the image that you can resolve," says Matthew Thompson, life sciences development manager in the Nanobio Instruments Group at Veeco Instruments of Woodbury, NY. Whether the noise source is acoustic or mechanical, anything that causes additional change between the head (which holds the tip) and the sample is going to translate into changes in the cantilever deflection.

The microscope's probe (cantilever and stylus tip) scans the sample in raster fashion: across, down, and across again, like an old dot-matrix printer. Raster comes from the Latin for rake, which is a fairly accurate approximation of what actually happens during a scan: in certain modalities the tip makes continuous contact with the sample, in others (tapping mode) it touches down intermittently. "One of the advantages of the AFM is it's a kind of 3-D microscopy – you measure X, Y, and Z, so you can measure real heights of features," says Hansma.

The force applied on the surface by the cantilever as it flexes over the surface is given by Hooke's Law, F = ks, where F is applied force, s is the deformation of the body caused by that force, and k is a spring constant. The cantilever's deflection (e.g., s) is most commonly detected through the beam-bounce method, in which a laser beam, bounced off the tip of the cantilever, hits a detector. When the cantilever deflects, the beam's movement is detected with a split photodiode detector, which emits an electrical signal that is proportional to the extent of the deflection.

Though most biologists use AFMs for imaging, the biophysical community has adapted them for use in force-spectroscopy experiments. "It's essentially the same instrument," says Calvin Quate of Stanford University, who, with Gerd Bennig and Christoph Gerber, invented the first AFM. "It's modified in the sense there's no scanning required, the tip just dips down and picks up a protein molecule, and lifts back and stretches it and it unfolds, and they can monitor the unfolding. Then they stretch it until it breaks, and they get some idea of the [molecule's] elastic properties." (Veeco Instruments' 30-page A Practical Guide to SPM,2 lists four primary imaging modes, 11 secondary imaging modes, and nine nonimaging modes.)

The tip can also be biofunctionalized for various kinds of experiments, says Patrick Collier of the California Institute of Technology. For example, Collier cites dippen nanolithogra-phy, Chicago-based NanoInk's 21st-century advance on, well, inkwells. "You coat the AFM tip with molecules (your 'ink') and dip the tip into the substrate (your 'paper')," he explains.

Or, suggests Fred Sachs of the University at Buffalo, NY, "You could put a specific ligand on the tip, if you're looking for a neuro-transmitter receptor .... You go poking around in this raster, wherever it sticks, you find a high force, and wherever it doesn't stick, you find a low force, and you could make a map of where the physical location of receptors are."


AFMs have one well-known weakness, says Michael Roukes of Caltech: "The time it takes to acquire an image." They typically scan almost 30 times slower than an electron microscope. As a result, certain biological events, such as nuclear pore-complex activity or lipid thermodynamics and other high-speed dynamic processes, cannot be captured using current AFMs.

That wouldn't be the case, however, if the cantilevers were made smaller than the current 100 microns (see sidebar). "The resolution limit that we have is fluctuations in the cantilevers," says Hermann Gaub of Ludwig-Maximilian University of Munich. "The smaller a cantilever is, the less it dissipates when it moves through liquid, and if it has less dissipation, it will also have less fluctuation."

"There's a lot more to the instrument than the cantilevers," says Thompson. For optimal performance, the feedback loop and actuator should also be adjusted to follow the much higher changes on the sample surface detected by smaller cantilevers.

But, when small cantilevers are fully developed, there may still be a Catch-22 in terms of costs. Typically you clip a cantilever into the microscope. But current AFMs are tailored to standard-size cantilevers, and after a certain shrinkage, it's no longer possible to simply swap-in new cantilevers. Both commercial and academic sources say next-generation commercial AFMs will likely address this problem.

To Buy or Not to Buy, and Why

Research-grade AFMs can cost from $100,000 to $300,000, depending on the configuration, and an integrated AFM-confocal microscope could run much higher. But AFMs don't come in a one-size-fits-all model. Even the people whose business is AFM sales say that it's impossible to determine whether you need an AFM or what kind you'll need, without knowing exactly which questions you want answered. They do, however, offer some general advice:

1. Try before you buy. Ask the manufacturer to give you a thorough instrument demonstration on your own samples. Even borrowing a colleague's AFM may not show you how a new microscope would work on your own, unique problem.

2. Check out the noise. Noise measurements are the primary criteria for good resolution in biological experiments. "The baseline noise level should be smaller than the samples you want to look at," says Asylum Research's Irene Revenko. Make sure to ask whether you are being given raw scores (which is what you want) or scores averaged over several time periods.

3. Don't buy more than you need. High-end instruments with optimal noise specifications could be overkill if you're not interested in single-molecule biophysics. But, if you decide to choose a lower-end instrument, pay attention to the upgrade limits.

4. Location location location. "The first thing you should think about for an AFM is whether there's a suitable location," says UCLA's Mari Gingery, "Is there a vibration-free environment? Vibration control is underestimated as an issue ... you've got vibrations of a couple angstroms and you're trying to image at a couple of angstroms."

5. Sample prep is key. Finally, it isn't about instrumentation alone, says Fred Sachs of the University at Buffalo, NY. "As with all microscopy, sample preparation is the key. Expect to spend the time on sample preparation."

Selected Atomic-force Microscope Vendors

Asylum Research

BioForce Nanosciences

EXFO Burleigh


JPK Instruments


Molecular Imaging

Nanofactory Instruments


Nanonics Imaging



Novascan Technologies



Omicron Nano Technology

Pacific Nanotechnology


Quesant Instrument

RHK Technology


Surface Imaging Systems


Veeco Instruments


Interested in reading more?

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?