Atomic Force Microscopy (AFM)¶
Atomic force microscopy (AFM) is a surface metrology technique in which a sharp probe on a flexible cantilever is scanned across a sample while the force between tip and surface is held constant. The vertical motion required to maintain that force is recorded at every point, producing a calibrated three-dimensional height map with sub-nanometre vertical sensitivity.
How atomic force microscopy works¶
The probe is a tip a few nanometres across at the end of a cantilever. A laser reflects off the back of the cantilever onto a segmented photodetector, so any deflection of the cantilever moves the laser spot and is measured directly. A feedback loop drives a piezoelectric scanner to keep that deflection constant as the tip moves, and the scanner's vertical correction is the topography.
Two modes cover most work:
- Contact mode keeps the tip in continuous contact and tracks the surface directly. It is fast and tolerant of setup error, and it suits hard, flat samples. Lateral drag can damage soft material.
- Tapping mode oscillates the cantilever near resonance so the tip touches the surface only briefly at the bottom of each cycle. Shear force falls sharply, which is what makes polymers, biological material, and loosely bound particles measurable.
Two further channels record material contrast rather than height. Phase imaging measures the lag between drive and response, distinguishing regions of differing stiffness or adhesion even where they are level. Lateral force measures cantilever twist and maps friction.
Because AFM senses force and not current, it needs neither vacuum nor a conductive coating - the fundamental difference from electron microscopy, and often the reason it is chosen.
When to use AFM¶
AFM is the correct choice when the question is quantitative and vertical:
- Measuring surface roughness as a calibrated Ra or RMS value
- Measuring step height of a deposited or etched film
- Confirming thin-film uniformity and continuity after deposition
- Measuring the height and pitch of patterned nanostructures
- Distinguishing phases of differing stiffness by phase contrast
- Characterising surfaces that must not be coated, dried, or placed in vacuum
If you need a number in nanometres rather than a picture, AFM is the instrument.
What AFM cannot do¶
- The scan area is small. 50 by 50 micrometres per image. Whole-wafer survey work belongs on optical profilometry.
- It is slow. A high-resolution image is minutes, not seconds. It is a measurement tool, not a survey tool.
- Vertical range is limited. Features taller than 17 micrometres exceed the scanner.
- It reports no chemistry. AFM measures shape and mechanical response. For elemental composition use SEM-EDS.
- The tip convolves the image. Measured lateral width is the true feature broadened by the tip shape. Steep walls and deep trenches are systematically distorted; vertical heights remain accurate.
AFM at MPaCT Lab, Flagstaff, Arizona¶
The MPaCT Lab at Northern Arizona University operates an AFM Workshop B-2 atomic force microscope in Flagstaff, Arizona. The scanners are linearized in all three axes with closed-loop strain gauges, so lateral distances and step heights are measured against a calibrated scale rather than inferred from piezo drive voltage.
It runs in ambient air. No vacuum, no sputter coating, no fixation - a sample can be measured and returned in the same condition it arrived. For thin films, coatings, and patterned substrates that must continue to a downstream process step, this is decisive.
The instrument supports both research and teaching, and is available to NAU researchers, external academic users, and industry partners on a fee-for-service basis or as a trained hands-on user.
| Specification | Value |
|---|---|
| XY scan range | 50 um x 50 um |
| Z range | 17 um |
| Noise floor | < 0.3 nm standard; < 0.15 nm with vibration isolation table |
| Scanners | Linearized X, Y, Z with closed-loop strain gauges |
| Imaging modes | Vibrating (tapping), non-vibrating (contact), phase, lateral force |
| Environment | Ambient air, acoustic enclosure |
Noise floor figures are as published by AFM Workshop for the B-2.
Full AFM Workshop B-2 specifications and booking →
Sample requirements¶
AFM asks less of a sample than any other technique in the lab:
- Flatness. Total height variation under 17 micrometres across the scan area. Very rough samples are the one common disqualifier.
- Size. Must sit stably on the stage. Small coupons, chips, and cleaved pieces are ideal.
- Cleanliness. Loose particles and residue attach to the tip and corrupt the scan. Blow off with dry nitrogen before submitting.
- Conductivity. Not required.
- Vacuum compatibility. Not required.
- Coating. Not required, and not wanted - a sputtered layer changes the roughness being measured.
Frequently asked questions¶
What is the difference between AFM and SEM for surface roughness?¶
AFM measures height directly with a physical probe and returns calibrated roughness values in nanometres. SEM encodes topography as image brightness, which looks three-dimensional but is not a height measurement. For a quantitative Ra or RMS roughness figure, AFM is the correct instrument.
Does an AFM sample need to be conductive or coated?¶
No. AFM senses force rather than current, so insulators, polymers, and oxides are measured directly with no conductive coating. Measurements run in ambient air rather than vacuum, so the sample is returned unaltered.
How large an area can the AFM scan?¶
Up to 50 by 50 micrometres laterally, with 17 micrometres of vertical range. Features taller than that, or surveys across a whole wafer, are better handled by optical profilometry on the Keyence VK-X3000.
Why does my AFM image look streaked or wider than expected?¶
Almost always tip convolution. The measured image is the true surface broadened by the shape of the tip, so steep walls slope and narrow trenches appear shallow. Lateral widths on high-aspect-ratio features are therefore systematically overstated. Vertical heights stay accurate, which is why AFM step-height data is trustworthy even where widths are not.
Request time on this instrument¶
MPaCT Lab - Building 98E, South Engineering Lab
561 E Pine Knoll Dr, Flagstaff, AZ 86001
Phone: 928-523-2343 · Email: mpct.nano@nau.edu