AFM Modes: Contact vs Tapping vs Non-Contact

Atomic force microscopy (AFM) builds an image by dragging or tapping a sharp tip across a sample surface and tracking how the tip deflects, rather than by focusing light or electrons at the sample. Because it senses physical force rather than optics, AFM can resolve features from micrometers down to sub-nanometer scale on almost any surface — conductive or insulating, in air, in liquid, or in vacuum — and it returns true quantitative height (z) data, not just a projected image. The single biggest decision an operator makes before scanning is which of the three imaging modes to run: contact, tapping (intermittent contact), or non-contact. That choice sets the force applied to the sample, the risk of damage, and ultimately what the image can and can’t tell you. This guide compares the three modes directly, walks through a generic engage-and-image procedure, and covers the artifacts operators run into most often.

How AFM works: the core components

Every AFM, regardless of mode, is built from the same four elements:

Cantilever and tip. A microfabricated cantilever, typically silicon or silicon nitride, with a sharp tip (radius commonly in the 5-15 nm range for a new tip) mounted at its free end. The tip radius sets the practical limit on lateral resolution — a sharper tip resolves finer lateral detail, and tip radius increases as the tip wears, which is why images can visibly degrade partway through a session.
Optical lever detection. A laser is reflected off the back of the cantilever onto a position-sensitive (typically quadrant) photodiode. As the cantilever bends or its oscillation changes, the reflected spot moves on the photodiode, and that signal is what the instrument actually measures — cantilever motion, converted to a voltage.
Piezoelectric scanner. Either the sample or the tip is moved in x, y and z by piezoelectric actuators capable of sub-nanometer positioning. The z-axis piezo is what maintains the tip-sample interaction as the feedback loop dictates.
Feedback loop. A proportional-integral (PI) controller adjusts the z-piezo in real time to hold a chosen setpoint (a deflection, an oscillation amplitude, or a frequency shift, depending on mode) constant as the tip scans across topography. The z-adjustment the controller makes at each pixel, not the raw deflection, is what becomes the topography image.

The three imaging modes

Contact mode

In contact mode the tip stays in continuous physical contact with the sample surface — in ambient air, effectively through the thin adsorbed water/contaminant layer present on most surfaces. The feedback loop holds cantilever deflection (and therefore the force pressing the tip into the surface, via Hooke’s law, F = -kx) constant, and the z-piezo’s response as the tip tracks up and down over features becomes the topography signal. Contact-mode cantilevers are deliberately soft — a low spring constant, achieved by making the cantilever thin — so that a small, controllable force can be maintained without excessive loading. Because the tip never leaves the surface, contact mode is comparatively fast and mechanically simple, and on hard, flat, robust samples (freshly cleaved mica, many inorganic crystals, some polymers) it can resolve very fine detail. Its drawback is lateral (shear) force: as the tip drags across the surface it exerts sideways as well as vertical force, which can drag, tear, or otherwise damage soft samples — biological specimens, polymers below their glass transition, loosely bound particles — and can also drag the tip itself across abrasive samples, dulling it faster.

Tapping mode (intermittent contact / AC mode)

Tapping mode oscillates the cantilever at or near its resonant frequency, with a free-air oscillation amplitude typically in the tens of nanometers, and lets the tip touch the surface only briefly at the bottom of each oscillation cycle rather than staying in continuous contact. The feedback loop holds the oscillation amplitude constant (the amplitude setpoint) as the tip scans; surface features change how much the tip’s oscillation is damped, and the z-piezo’s compensation for that damping becomes the topography signal. Because contact is intermittent, lateral shear forces are largely eliminated, which makes tapping mode the default general-purpose choice for soft, loosely bound, or biological samples in air, and it is generally regarded as capable of resolution comparable to contact mode without the same damage risk. Tapping-mode cantilevers need a comparatively high spring constant and resonant frequency to overcome tip-sample adhesion forces on each tap and recover cleanly, which is the opposite requirement from contact mode — the two probe types are not interchangeable. Tapping mode also produces a phase channel alongside topography (see below), which is one of its most useful features for heterogeneous samples.

Non-contact mode

Non-contact mode also oscillates the cantilever, but at a small amplitude and at a frequency slightly above resonance, and the tip never touches the surface at all — it stays in the regime of long-range attractive forces (primarily van der Waals) a few nanometers above the sample. The feedback loop tracks the shift in resonant frequency or amplitude caused by that attractive force gradient and uses it to maintain constant tip-sample separation. Because the tip never contacts the sample, non-contact mode applies the lowest force of the three modes and causes essentially no mechanical wear to either tip or sample, which makes it the preferred mode for extremely fragile or easily contaminated samples and for ultra-high-vacuum (UHV) work, including true atomic-resolution imaging. Its practical drawback in ambient air is the same adsorbed fluid layer that complicates contact mode: the attractive-force signal is weaker and noisier through that layer, so lateral resolution in ambient non-contact imaging is normally lower than either contact or tapping mode. Non-contact mode is used far more in vacuum systems than on ambient benchtop instruments for this reason.

Contact vs. tapping vs. non-contact: a comparison table

Property	Contact mode	Tapping mode (AC / intermittent contact)	Non-contact mode
Tip-sample contact	Continuous	Intermittent (brief contact at bottom of each oscillation)	None — stays in the attractive-force regime
Force regime sensed	Repulsive contact force (constant deflection)	Repulsive, briefly, at each tap (constant amplitude)	Attractive long-range forces, chiefly van der Waals
Typical tip force	Lowest achievable is still a sustained contact force; commonly nN-scale and higher	Lower time-averaged force than contact mode; brief nN-scale taps	Lowest of the three — sub-nN, no sustained loading
Sample damage risk	Highest — lateral/shear drag on soft or loosely bound material	Low — shear forces largely eliminated by intermittent contact	Lowest — no mechanical contact at all
Best-suited specimens	Hard, flat, robust surfaces (cleaved mica, many inorganic crystals, robust polymers)	Soft, biological, loosely bound, or easily damaged samples in air or liquid; general-purpose default	Extremely fragile or contamination-sensitive samples; UHV and atomic-resolution work
Typical lateral resolution	Can reach sub-nanometer to atomic-scale on ideal flat, hard samples	Comparable to contact mode on most real-world samples	Normally lower than contact or tapping mode in ambient air; can reach atomic resolution in UHV
Cantilever spring constant	Low (soft, thin cantilever)	Higher (stiffer, to overcome tip-sample adhesion on each tap)	Higher, similar family to tapping-mode probes
Typical environment	Air or liquid	Air or liquid	Almost always vacuum for best results; degraded performance in air
Extra channels available	Deflection/error signal	Amplitude error and phase (compositional contrast)	Frequency-shift signal

Choosing a mode for your sample

In practice, mode selection is driven mostly by sample fragility and by what question the image needs to answer:

If the sample is hard, flat, and robust (a cleaved crystal, an etched semiconductor, an inorganic film) and you want the fastest scan with the finest achievable detail, contact mode is a reasonable default.
If the sample is biological, polymeric, soft, or only loosely adhered to the substrate — which describes most cell, tissue, protein, membrane, and particle samples — tapping mode is the standard general-purpose choice, in air or in liquid.
If the sample is exceptionally fragile, prone to contamination, or the work requires true atomic-resolution imaging under vacuum, non-contact mode is the appropriate choice, with the understanding that it requires a UHV or controlled-vacuum system to perform well.

When in doubt with an unfamiliar or precious sample, starting with the lowest force the instrument’s tapping-mode setpoint allows, and only increasing force if tracking is poor, is the safer default than starting in contact mode.

Step-by-step: engaging and imaging a sample

The sequence below describes a generic tapping-mode workflow, the most common starting point for new samples; contact-mode steps differ mainly at steps 4 and 5, where a deflection setpoint replaces an oscillation setpoint and no resonance tuning is needed.

Select the cantilever/tip. Choose a probe rated for your intended mode and matched to your sample — a soft, low-spring-constant probe for contact mode, or a stiffer, higher-resonant-frequency probe for tapping or non-contact mode.
Mount the sample. Secure the sample flat and stable on the sample stage (commonly a magnetic puck), minimizing tilt, since excessive tilt can exceed the z-piezo’s range partway through a scan.
Load the probe and align the laser. Install the cantilever in its holder, then align the laser spot onto the back of the cantilever near the tip, and center the reflected beam on the photodiode detector so the baseline deflection and sum signal are within the instrument’s expected range.
Tune and set the operating point. For tapping or non-contact mode, run a frequency sweep to find the cantilever’s resonant peak and set the free-oscillation drive amplitude; for contact mode, instead zero the deflection setpoint against the free (non-contacting) baseline.
Set the imaging setpoint. Choose the target deflection (contact mode) or the target amplitude/frequency ratio relative to free oscillation (tapping/non-contact mode) — this setpoint directly sets how hard the tip presses or how close it approaches the surface, and should be the gentlest value that still gives stable tracking.
Engage the tip. Bring the tip toward the surface with the coarse-approach motor, then trigger the automated engage sequence, which advances the tip in small steps until the feedback loop detects the setpoint has been reached and locks on.
Set scan parameters. Choose scan size, pixel resolution (line and frame density), and scan rate — smaller, slower, higher-resolution scans generally track topography more accurately but take longer and increase drift exposure.
Optimize feedback gains. Increase the proportional (P) and integral (I) gains until the trace and retrace scan lines start to show ringing or oscillation artifacts, then back the gains off just below that point — too low and the feedback lags behind real topography; too high and it overshoots and rings.
Capture and inspect the image. Compare trace and retrace lines for agreement (poor agreement indicates the feedback isn’t tracking cleanly) and check for obvious artifacts before saving; adjust setpoint, gains, or scan rate and re-image if needed.
Withdraw the tip. Fully retract the tip before removing or moving the sample, and before switching probes, to avoid an uncontrolled tip crash.

Reading the output: topography, amplitude, and phase

A tapping-mode scan typically produces at least three data channels per pixel, and reading them together, not just the height image, is what makes an AFM dataset interpretable:

Topography (height). The z-piezo’s feedback response at each pixel — the quantitative height map, and the channel most people mean by “an AFM image.”
Amplitude (error signal). The residual difference between the actual oscillation amplitude and the setpoint at each pixel — ideally near zero everywhere if feedback is tracking well; a non-zero, edge-highlighting error image is a useful diagnostic for where the feedback loop is struggling to keep up, often at steep edges.
Phase. The phase lag between the cantilever’s driven oscillation and its actual response, which is sensitive to local material properties — stiffness, adhesion, viscoelasticity — independent of height. Phase contrast is widely used to distinguish different materials or domains within a sample (for example, separating a filler phase from a polymer matrix) but it is a qualitative, comparative signal rather than a direct, calibrated measurement of any single material property.

AFM compared with other microscopies

AFM is frequently chosen alongside, or instead of, electron and optical microscopy, and each has a different trade-off profile:

Versus scanning electron microscopy (SEM). SEM generally images a much larger field of view faster and doesn’t require the sample to be scanned point-by-point in the way AFM does, but it images a 2D projection rather than quantitative height data, typically requires vacuum, and requires a conductive (or conductively coated) sample for standard imaging. AFM instead returns true 3D topography and works on insulating samples without coating, but over a much smaller field of view and at slower scan speeds.
Versus confocal microscopy. Confocal is light-based, works well on living cells, and is diffraction-limited to roughly a few hundred nanometers of lateral resolution; see our confocal microscopy guide for the setup and when to use it. AFM resolves finer surface detail and gives direct height data, but only images the surface, not internal structure, and cannot match confocal’s optical sectioning through a sample’s depth.
Versus structural techniques. For atomic-resolution structural information about a purified, crystallizable macromolecule, X-ray crystallography and cryo-electron microscopy (archived via resources like EMPIAR) answer a different question than AFM — atomic structure of a molecule in a crystal or vitrified sample, rather than the surface topography of an intact specimen under near-native conditions.

Troubleshooting common AFM imaging artifacts

Symptom	Likely cause	Fix
Streaky or doubled lines running across features	Dull, damaged, or contaminated tip	Retract and inspect; swap to a fresh tip; reduce setpoint force to slow further wear
Features look unnaturally rounded or wider than expected for the sample	Tip broadening — tip radius large relative to the feature size	Use a sharper tip for the feature scale involved; treat lateral dimensions as tip-convolved, not exact, unless deconvolved
Fine cross-hatched noise or ringing along scan lines	Feedback gains (P/I) set too high	Lower the gains until ringing disappears, then raise slightly for the fastest stable tracking
Image looks smeared or lags behind real topography, worse in one scan direction	Feedback gains too low, or scan rate too high for the gains set	Raise gains modestly, or reduce scan rate
Sudden discontinuity or the tip stops responding mid-scan	Tip crash from an excessive setpoint or a rough, uncontrolled approach	Reduce setpoint force, re-engage carefully, and check the tip/cantilever for damage before continuing
Whole image appears to drift or shift progressively during the scan	Thermal drift or piezo creep, especially soon after engaging	Allow the system to thermally equilibrate before capturing data; rescan the same area to confirm drift has settled
Sample appears torn, dragged, or scratched (soft samples)	Contact-mode force too high, or contact mode used on a sample that needed tapping/non-contact mode	Switch to tapping or non-contact mode; lower the setpoint force
Regular repeating pattern or moiré-like artifact across the image	External mechanical or acoustic vibration coupling into the scanner	Use vibration isolation (an isolation table or platform) and an acoustic enclosure
Tip fails to track steep edges or tall features cleanly	Scan rate or gains insufficient for the topography, or tip aspect ratio too blunt for the feature	Slow the scan rate, raise gains within stable limits, and use a sharper, higher-aspect-ratio tip for tall or steep features

Frequently asked questions

What is atomic force microscopy used for?

AFM is used wherever nanoscale-to-micron-scale surface topography, mechanical properties, or fine surface features need to be measured directly in three dimensions — materials science (thin films, nanoparticles, polymers, semiconductors), biology (cells, membranes, proteins, DNA), and any application needing quantitative height or roughness data that optical or electron microscopy can’t provide directly.

Which AFM mode is best for biological samples?

Tapping mode, run in air or in liquid, is the standard general-purpose choice for biological and other soft, loosely bound samples, because intermittent contact largely eliminates the lateral shear forces that can drag or damage soft material in contact mode.

Can AFM image samples in liquid?

Yes — both contact and tapping mode can be operated in liquid using an appropriate liquid cell and probe, which is one of AFM’s advantages over techniques that require vacuum, and is widely used for imaging biological samples under near-native, hydrated conditions.

How does AFM resolution compare with an optical or electron microscope?

AFM’s lateral resolution is set primarily by tip radius and can reach sub-nanometer to atomic scale under ideal conditions (a very sharp tip, a flat hard sample, and often vacuum for true atomic resolution), well beyond the diffraction limit of optical microscopy; its vertical (height) resolution is generally its strongest feature, often sub-angstrom. Electron microscopy can match or exceed AFM’s lateral resolution over a much larger field of view, but does not return direct, quantitative height data the way AFM does.

Why does AFM image quality degrade partway through a scan?

The most common cause is a wearing or contaminated tip — tip radius increases with use, which broadens and can streak features — followed by thermal drift, which shows up as a slow shift or distortion of the same features over repeated scans.

Atomic Force Microscopy: Contact, Tapping and Non-Contact Modes Compared

Ask about Atomic Force Microscopy: Contact, Tapping and Non-Contact Modes Compared