AFM sample preparation guide for reliable nanoscale imaging

What AFM sample preparation must accomplish

AFM sample preparation is the work of presenting a surface so the probe measures the feature of interest, not dust, salts, glue, fingerprints, loose particles, drying damage, or the background from an unsuitable substrate. For most atomic force microscopy work, a good sample is clean, stable, firmly mounted, compatible with the imaging environment, and flat enough for the scanner to follow the surface without losing feedback. The right method depends on whether the experiment involves a hard film, a soft polymer, a biological specimen, a nanoparticle dispersion, or a conductive structure. The practical rule is to prepare the surface for the measurement question, not for a generic ideal surface.

This matters because AFM is highly surface sensitive. A sharp tip interacts with the uppermost features, so a thin residue layer or a few mobile particles can dominate the image. In routine laboratory work, poor AFM images often start with preparation problems rather than instrument faults. A polished semiconductor cross-section, a freshly cleaved mica sheet, a rinsed biomolecule sample, and a hydrated tissue section all require different handling. Treating them as the same specimen type usually creates artifacts.

For more articles on laboratory handling and preparation workflows, visit the sample preparation section.

Match the substrate to the measurement

The substrate is not a neutral background. It contributes roughness, surface charge, hydrophilicity, adhesion, electrical behavior, and contamination risk. The best substrate is usually smooth relative to the feature size, chemically compatible with the sample, and practical to prepare reproducibly.

Substrate	Typical use in AFM	Preparation advantage	Main limitation
Freshly cleaved mica	DNA, proteins, nanoparticles, thin soft materials	Molecularly smooth, hydrophilic surface after cleavage	Surface charge and strong hydration can affect adsorption chemistry
Highly oriented pyrolytic graphite	Hydrophobic molecules, carbon-related studies, some conductive measurements	Cleaves to a flat graphitic surface	Less suitable for hydrophilic adsorption without surface modification
Silicon or silicon dioxide wafer	Thin films, microfabricated structures, polymer coatings, semiconductor sections	Flat, widely available, compatible with many cleaning routes	Native oxide, surface charge, and organic residue must be controlled
Glass coverslip	Cells, larger biological specimens, optical correlation	Convenient, transparent, compatible with cell culture workflows	Usually rougher and less uniform than mica or wafer-grade substrates
Gold-coated substrate	Self-assembled monolayers, thiol chemistry, conductive AFM	Useful for chemical functionalization and electrical connection	Evaporated gold roughness and grain structure may complicate nanoscale topography

A frequent mistake is choosing a substrate because it is familiar rather than because it supports the measurement. If the target feature is only 1 to 5 nm high, background roughness and contamination must be below that scale. If the target is a living or hydrated structure, a chemically flat dry substrate may still be the wrong choice when immobilization damages the specimen. If the goal is electrical contrast, mechanical smoothness is only one part of the preparation; contact resistance, grounding, and oxide layers also affect the result.

A practical AFM sample preparation workflow

Define the surface and imaging mode first

Before handling the specimen, define the region of interest, expected height range, feature size, acceptable environment, and intended imaging mode. Tapping or intermittent-contact modes are often preferred for delicate samples because they reduce continuous lateral forces. Contact mode can be suitable for robust surfaces or force mapping, but it may drag weakly attached material. Liquid imaging can better preserve hydrated biological structures, while air imaging is simpler and often suitable for hard, dry materials.

The surface height range should fit comfortably within the instrument z travel and the selected scanner. Tall steps, tilted mounts, dried droplets with thick edges, and uneven adhesive can all prevent stable feedback. If the sample has macroscopic roughness, trim, polish, microtome, cleave, or select a flatter region before attempting high-resolution scanning.

Clean without changing the surface you need to measure

Cleaning should remove unwanted residue while preserving the feature being studied. Solvent rinsing, ultrapure water rinsing, nitrogen drying, UV-ozone treatment, plasma cleaning, and thermal treatment are all used in AFM laboratories, but none is universally safe. Plasma can make many inorganic surfaces more hydrophilic, yet it can also oxidize, etch, or chemically modify polymers and organic layers. Aggressive oxidizing mixtures may clean glass or silicon in some facility protocols, but they require approved safety procedures and can destroy chemically sensitive surfaces.

Handle substrates with clean tweezers and gloves, and avoid touching the region of interest. Use filtered solvents and clean containers. A freshly cleaned sample placed on a dusty bench can become unsuitable within minutes. Storage should match the surface chemistry; covered dishes, wafer boxes, desiccators, or nitrogen cabinets may be appropriate.

Deposit or immobilize the material reproducibly

The deposition method controls sample distribution. Drop casting is simple, but evaporation can concentrate material at the droplet edge and leave coffee-ring patterns. Spin coating can improve uniformity for polymers and dispersions, although spin speed, solution concentration, solvent volatility, and humidity affect thickness. Dip coating and controlled incubation can work well for biomolecules and nanoparticles, but adsorption time and buffer composition must be documented.

Weak immobilization is a common cause of streaking, repeated feature displacement, and tip contamination. If particles or biomolecules move during scanning, the image may show the path of the tip rather than the original sample. Possible strategies include optimizing ionic strength, using an appropriately charged substrate, applying a coupling layer, drying only when compatible, or imaging in liquid with gentler scan parameters. The correct choice depends on whether the measurement is intended to preserve native structure, quantify topography, map mechanics, or characterize adhesion.

Rinse and dry with artifact control

Rinsing is not just a finishing step. Salts, buffers, surfactants, and unbound material can dry into crystals or films that are larger than the nanoscale structures being measured. For samples that tolerate it, gentle rinsing with ultrapure water or a compatible volatile solvent can reduce residue. For samples that do not tolerate dilution or solvent exposure, liquid AFM or buffer exchange may be better than drying.

Drying should be gentle and repeatable. Strong gas flow can push loosely attached particles into streaks or leave directional artifacts. Air drying can collapse soft structures through capillary forces. Critical-point drying, freeze drying, or imaging in liquid may be considered for selected hydrated specimens, but each approach has equipment requirements and may change morphology. A useful method sheet should not simply say to dry the sample. It should define how drying is done and require verification that the drying method does not create the features being measured.

Mount firmly and keep the region accessible

Mounting affects both mechanical stability and imaging access. The sample should sit flat on the puck or holder, with adhesive kept away from the scan region. Double-sided tape, wax, magnetic disks, epoxy, silver paint, and mechanical clamps are common choices, but they differ in outgassing, creep, conductivity, and contamination behavior. For electrical AFM modes, the mounting route must provide a reliable conductive path without spreading conductive paste onto the area of interest.

After mounting, inspect the surface optically if possible. AFM should not be the first tool used to find a millimeter-scale tilt, a fiber, a cracked substrate, or a fingerprint. A few minutes of optical inspection can prevent hours of failed scanning.

Preparation choices for common sample types

Hard materials, coatings, and cross-sections

Hard samples still require careful preparation. Thin films, wafers, ceramics, metals, and polished cross-sections can carry polishing slurry, abrasive particles, oxide changes, or embedded residue. When the measurement is roughness, do not over-polish away the functional surface. When the measurement concerns a buried interface or doped region, the cross-section must expose the region cleanly with minimal damage. Instrument guidance for semiconductor AFM commonly emphasizes low roughness, cleanliness, and minimal surface damage because rough or charged surfaces can distort contrast in electrical modes.

Nanoparticles and powders

Nanoparticle AFM sample preparation is a balance between isolation and representativeness. A dispersion that is too concentrated produces aggregates and overlapping features; a dispersion that is too dilute wastes scan time and may bias the visible population. Sonication, centrifugation, filtration, and dilution can help, but they can also change the agglomeration state. A good workflow records solvent, concentration, sonication time, deposition volume, incubation time, rinse method, and drying method. Imaging a blank substrate and a solvent-only control helps separate real particles from residue. See also: buying guides.

Biomolecules, cells, and hydrated specimens

Biological AFM sample preparation depends on immobilization without excessive distortion. Reviews of biological AFM preparation repeatedly emphasize that specimens must be well attached to a solid substrate by methods that preserve the structure under study. For DNA and proteins, mica and modified mica are common because they provide a smooth background and tunable adsorption chemistry. For cells, glass, coated glass, culture-compatible substrates, or functionalized surfaces may be selected for adhesion and optical compatibility.

Dry biological samples are easier to image in air, but drying can flatten, shrink, crystallize, or reorganize soft structures. Liquid imaging better preserves hydrated states, although it requires cleaner fluid handling, stable temperature, compatible probes, and more attention to drift. If fixation is used, the fixation chemistry should be treated as part of the experiment rather than as a neutral handling step. Different fixatives can change surface ultrastructure and mechanical properties.

Conductive and functional AFM modes

Conductive AFM, Kelvin probe force microscopy, piezoresponse force microscopy, and related modes add preparation requirements. A topographically clean sample may still fail if it has poor grounding, an insulating contamination layer, mobile surface charge, or an unstable contact. Keep conductive paste and silver paint away from the scan field. Confirm that the back contact is reliable. Avoid cleaning routes that change the work function, oxide state, or polarization unless those changes are intended and documented.

Artifact checks before imaging and during troubleshooting

A useful AFM preparation method includes controls. The fastest control is a blank substrate prepared with the same cleaning, mounting, rinsing, and drying steps. If the blank shows particles, residues, terraces, scratches, or streaking at the same scale as the sample, the preparation method is not clean enough for interpretation.

Observation	Likely preparation issue	Check or corrective action
Large angular or faceted features after drying	Salt or buffer crystals	Use compatible rinsing, lower salt residue, or image in liquid
Streaks in scan direction	Loose material, tip contamination, or excessive force	Improve immobilization, reduce force, change tip, image a blank
Repeated displaced particles	Weak adhesion to substrate	Adjust substrate chemistry, incubation, or drying protocol
Unexpected uniform film	Solvent, surfactant, polymer, or airborne residue	Prepare solvent-only control and review containers and pipette tips
Loss of contrast in electrical AFM	Poor contact, oxide, charge, or insulating residue	Verify grounding, contact path, surface cleaning, and bias limits
Features change between scans	Sample damage, hydration change, drift, or mobile contaminants	Reduce force, control humidity or liquid environment, rescan at lower load

Tip condition should also be considered, but it should not become the default explanation for every problem. A contaminated or worn tip can broaden features and create repeated shapes, yet a dirty sample can contaminate a fresh tip within the first scan. When troubleshooting, change only one variable at a time: substrate, rinse, drying method, immobilization chemistry, probe type, or imaging environment. Otherwise, a successful scan may not show which change solved the problem.

Documentation, controls, and reproducibility

Reliable AFM sample preparation should be a written method, not a memory. Record substrate lot or source, cleaning method, solvent grade, deposition concentration, volume, incubation time, rinse volume, drying route, storage time, humidity when relevant, mounting material, imaging environment, probe model, scan size, scan rate, setpoint, and feedback settings. For quantitative work, add calibration checks, roughness analysis settings, image flattening choices, and exclusion rules for defects or particles.

Metrology organizations such as NIST have long emphasized calibration, traceability, and reproducible measurement in AFM, while ISO guidance covers areas such as probe characterization for nanostructure measurement. These metrology practices do not replace sample preparation, but they show why preparation and measurement must be treated together. A calibrated instrument cannot recover reliable dimensions from a surface covered with mobile residue, and a well-prepared sample can still produce misleading numbers if the probe shape, scanner calibration, or image processing is inappropriate.

The most useful information gain in an AFM preparation study often comes from a simple comparison set: blank substrate, substrate plus solvent, substrate plus buffer, fully prepared sample, and a replicate prepared on a different day. This set helps separate substrate background from process residue and real specimen features. It also gives future users a practical reference for what a clean preparation should look like at the same scan size.

Frequently asked questions

What is the best substrate for AFM sample preparation?

There is no universal best substrate. Freshly cleaved mica is excellent for many nanoscale biological and particle samples because it is very smooth, while silicon wafers are common for films and device materials. Glass is useful for cells and optical correlation, and gold is chosen when thiol chemistry or conductivity is needed. The best choice is the one that supports the measurement with the least background and chemical interference.

Should AFM samples be completely dry?

Only if the sample and the measurement can tolerate drying. Hard coatings and inorganic particles are often imaged dry, but hydrated biological structures, gels, some polymers, and weakly attached assemblies may deform during drying. Liquid AFM can preserve hydrated states, although it requires more control of cleanliness, drift, and probe-sample interactions.

How clean does an AFM sample need to be?

It needs to be cleaner than the scale of the feature being interpreted. If the feature of interest is a 2 nm molecular layer, submicron dust and dried salt crystals are obvious problems, but thin organic residue can also matter. A blank substrate control at the same scan size is one of the most practical ways to judge whether cleanliness is sufficient.

Why do AFM images show streaks or dragged particles?

Streaks often indicate loose material, tip contamination, excessive imaging force, or an unstable surface. Improve immobilization, reduce interaction force, try a fresh probe, and compare with a blank substrate. If particles move repeatedly in the scan direction, the preparation method is probably not holding the sample firmly enough.

Can plasma cleaning be used before AFM?

Plasma cleaning can be useful for removing organic contamination and changing surface wettability on some inorganic substrates, but it is not universally safe. It can modify polymers, organic films, biological coatings, and functional surfaces. Use it only when the intended surface chemistry can tolerate the treatment and when the method is documented as part of the experiment.