EM sample preparation workflows for TEM, SEM and cryo-EM

What EM sample preparation has to achieve

EM sample preparation is the set of steps that makes a specimen compatible with an electron microscope while preserving the information the analyst needs to interpret. For TEM, this usually means preparing an electron-transparent section or film with enough contrast. For SEM, it means stabilizing the surface, controlling or removing water, and reducing charging. For cryo-EM, the target is different: hydrated material should be held in vitreous ice rather than chemically fixed in resin.

The right workflow therefore depends on the imaging question, the specimen, and the level of preparation-induced change that can be accepted. A sound preparation plan starts before the sample reaches the microscope. Fixation, drying, sectioning, staining, coating and freezing can all introduce artifacts that may resemble real structure if the workflow is not understood and documented.

Start with the imaging question, not the instrument name

Electron microscopy is often discussed as though the instrument alone determines the preparation route. In practice, the research or quality-control question should come first. A surface morphology question usually points toward SEM preparation. An internal ultrastructure question points toward TEM sectioning or tomography. A near-native hydrated structure question may require cryo-EM or cryo-electron tomography. The preparation method should protect the feature of interest, even if that means accepting trade-offs elsewhere in the specimen.

This distinction matters for biological, polymer, nanoparticle, battery, catalyst and composite samples, where preparation can change hydration state, phase distribution, surface chemistry, porosity or fine interfaces. A polished cross-section may be ideal for one materials question and misleading for another. A chemically fixed cell can show organelle organization well, but it should not be treated as a fully native hydrated state. A cryo grid may preserve hydration better, yet it can introduce ice-thickness, preferred-orientation and air-water-interface problems.

Imaging goal	Common EM route	Preparation priority	Main artifact risk
Internal cell or tissue ultrastructure	TEM of resin sections	Fixation, dehydration, embedding, ultrathin sectioning and contrast staining	Shrinkage, extraction, poor infiltration or section compression
Surface topography	SEM	Drying, mounting, conductivity and surface cleanliness	Collapse of wet structures, charging or coating obscuring detail
Macromolecules in hydrated state	Single-particle cryo-EM	Thin film formation and vitrification	Ice contamination, preferred orientation, aggregation or denaturation at interfaces
Cellular structures in 3D cryogenic context	Cryo-ET, often with FIB milling for thicker cells	Vitrification, targeting and lamella preparation	Devitrification, milling damage or insufficient thinning
Hard materials or particles	SEM, TEM, STEM or analytical EM	Clean mounting, thinning, polishing, dispersion or ion milling	Smearing, contamination, beam-sensitive changes or nonrepresentative particles

For more articles on laboratory workflows and preparation choices, see the sample preparation section.

Conventional TEM preparation builds an electron-transparent specimen

Transmission electron microscopy requires electrons to pass through the specimen. For many biological and soft specimens, the sample is chemically stabilized, dehydrated, embedded in resin, sectioned with an ultramicrotome, placed on a grid and stained with electron-dense reagents. Open-access microscopy protocols and university core-facility notes commonly describe ultrathin resin sections in the approximate 50-90 nm range for routine TEM. The optimum thickness still depends on accelerating voltage, material density, contrast method and analytical goal.

The workflow is usually described in stages. Primary fixation stabilizes proteins and overall structure. Post-fixation, often associated with osmium tetroxide in biological EM, helps preserve and contrast lipid-rich membranes. Dehydration replaces water through graded solvent steps so resin can infiltrate. Embedding polymerizes the resin to support cutting. Ultramicrotomy creates thin sections. Staining with heavy-metal contrast agents improves visibility because many biological structures are made mostly of light elements that scatter electrons weakly.

Each stage solves one problem while adding another risk. Fixatives do not penetrate all specimens at the same speed, so sample size and diffusion distance matter. Dehydration can extract soluble components and distort membranes. Resin infiltration can be uneven if the sample is too large or dense. Sectioning can compress, chatter, wrinkle or detach sections. Staining can add precipitates or create uneven contrast. A sharp TEM image is therefore not automatically a faithful specimen; it is a specimen shaped by a controlled preparation history.

Planning variables for TEM

• Specimen size: Smaller pieces generally improve fixative and resin penetration, but the sampling must still represent the material.
• Buffer and osmolarity: Biological specimens can swell or shrink if the chemical environment is poorly matched.
• Embedding resin: Hardness, beam stability and compatibility with staining or immunolabeling should match the method.
• Section thickness: Thinner sections improve electron transmission, while thicker sections may be used for tomography or special applications.
• Contrast strategy: Staining should support the question rather than simply make every region darker.

SEM preparation protects the surface and controls charging

Scanning electron microscopy mainly records signals generated near the surface, so SEM preparation focuses on surface preservation, cleanliness, mounting and electrical behavior. Dry, conductive and vacuum-compatible specimens may need little more than secure mounting. Hydrated biological materials are more demanding because water removal can collapse fine structures through surface-tension forces. Critical point drying and freeze-drying are common approaches for reducing drying damage. Environmental or low-vacuum SEM modes can sometimes reduce the need for full dehydration and coating.

Nonconductive specimens often charge under the electron beam. Charging can produce bright patches, image drift, streaking or sudden contrast changes. Sputter coating with a thin conductive layer of gold, gold-palladium, platinum, carbon or another material is a common solution, but it is not neutral. A coating that is too thick can hide nanoscale surface detail; a coating that is too thin may not control charging. Carbon may be preferred where X-ray microanalysis is important, while noble-metal coatings are often used for high-secondary-electron surface imaging.

SEM users also need to define what surface they are actually viewing: the natural surface, a fractured surface, a polished cross-section or a preparation-created surface. For particles and powders, dispersion affects whether the image represents primary particles or agglomerates. For plant, tissue or microbial samples, cleaning steps can remove debris but may also remove weakly attached structures. For coated or layered materials, cutting and polishing may smear soft phases across harder ones, making interfaces appear more continuous than they are.

Cryo-EM changes the target from preservation by chemistry to preservation by vitrification

Cryo-EM preparation aims to immobilize water as vitreous, non-crystalline ice. In single-particle cryo-EM, a small volume of sample is typically applied to a grid, excess liquid is removed to form a thin film, and the grid is rapidly plunged into a cryogen such as liquid ethane. Review literature in Nature Methods and open-access cryo-EM introductions describes sample preparation as one of the major bottlenecks because high-quality grids depend on sample concentration, buffer, grid support, glow discharge or surface treatment, blotting conditions, humidity, temperature and freezing speed.

The benefit is that macromolecules can be imaged in a hydrated, near-native environment without crystallization. The trade-off is that the thin film is fragile and variable. Particles may aggregate, adopt preferred orientations, move to the air-water interface, denature, or distribute unevenly across grid holes. Ice may be too thick for good contrast or too thin to contain intact particles. Contamination can appear as crystalline ice, frost or debris. Grid screening is therefore not a minor formality; it is often the decisive stage of the project.

For thicker cells, tissues or small organisms, plunge freezing alone may not cool the entire volume fast enough to avoid crystalline ice. High-pressure freezing can improve vitrification in thicker specimens, and cryo-focused ion beam milling can thin frozen cells into lamellae suitable for cryo-electron tomography. These methods expand what can be studied in situ, but they also add workflow complexity, equipment demands and possible artifacts from handling, targeting and milling.

Common failure modes and what they usually suggest

Preparation problems often appear first as image-quality problems. The table below connects visible symptoms with likely preparation causes. It is not a replacement for a facility protocol, but it gives users and microscopy staff a more precise way to discuss failures. See also: buying guides.

Observed issue	Likely preparation contributor	Practical checkpoint
TEM membranes look discontinuous or swollen	Delayed fixation, poor osmotic control, dehydration stress or extraction	Review sample size, time to fixation, buffer conditions and dehydration schedule
TEM sections show chatter or compression	Block hardness mismatch, knife condition, cutting speed or section thickness	Check resin cure, trim face, knife edge and ultramicrotome settings
SEM surface appears collapsed	Air drying of hydrated structures or incomplete drying strategy	Consider critical point drying, freeze-drying or cryo-SEM depending on the question
SEM image drifts or flashes	Charging from poor conductivity or weak grounding	Improve coating, mounting contact, stub grounding or use low-vacuum options
Cryo-EM particles cluster at hole edges	Grid surface, concentration, blotting or buffer effects	Screen support films, surface treatment, concentration and blotting parameters
Cryo-EM shows crystalline ice or heavy contamination	Poor vitrification, transfer warming or environmental contamination	Review cryogen condition, transfer workflow, humidity control and storage handling

The broader lesson is that artifacts are rarely random. They usually trace back to a small set of controllable variables: time, temperature, chemistry, water removal, mechanical stress, conductivity, contamination and beam sensitivity. Recording those variables is one of the simplest ways to make EM sample preparation more reproducible.

Safety and documentation should be part of the method

Several EM preparation reagents and operations carry significant hazards. CDC laboratory safety guidance identifies osmium tetroxide and glutaraldehyde as electron microscopy fixatives and emphasizes protective practices for EM laboratory work. Osmium tetroxide is volatile and highly toxic, glutaraldehyde is a sensitizer and irritant, many dehydration solvents are flammable, and resin components can irritate skin or airways. Cryogenic workflows add risks from liquid nitrogen, cold burns, oxygen displacement and pressure buildup in improperly vented containers.

Because of those risks, this article should be treated as a workflow overview, not a step-by-step operating procedure. Facility standard operating procedures, safety data sheets, local waste rules and training requirements should govern actual handling. The same applies to biological specimens, where fixation status does not automatically remove all biosafety considerations. Diagnostic, infectious or unknown samples should be processed under the correct containment and transport requirements before they enter a microscopy workflow.

Documentation is also a quality-control tool. At minimum, a useful EM preparation record should include specimen origin, sampling location, time from collection to stabilization, fixation or freezing method, solvent and resin sequence where applicable, section thickness target, grid or stub type, staining or coating conditions, storage conditions, and microscope settings used for screening. When images are later compared across batches, those notes often explain differences that might otherwise be blamed on the microscope.

How to choose a practical workflow

A practical workflow balances preservation, throughput, cost, safety and interpretability. If the question is whether a surface is rough, cracked or contaminated, SEM with careful mounting and charge control may be enough. If the question is how membranes, organelles or layered structures are arranged internally, TEM sectioning or related volume EM methods may be better. If the question is the hydrated conformation of macromolecules, cryo-EM may be the right direction, but only if the sample can produce usable vitrified grids.

For laboratories building or refining methods, the most useful improvement is often not a more complex protocol. It is a small validation set: prepare the same specimen with two plausible routes, compare expected artifacts, document screening outcomes, and choose the method that answers the question with the fewest assumptions. EM preparation is successful when the final image can be interpreted with confidence, not merely when it looks visually impressive.

Frequently asked questions

Is EM sample preparation the same for TEM and SEM?

No. TEM generally requires an electron-transparent specimen, such as an ultrathin section or thin particle film. SEM usually focuses on surface stability, conductivity and vacuum compatibility. Some early steps, such as fixation or dehydration, may overlap for biological samples, but the final specimen format is different.

Why do biological EM samples often need fixation?

Fixation stabilizes structures before dehydration, embedding, drying or vacuum exposure. Without stabilization, cells and tissues can change rapidly after sampling. However, fixation is itself a chemical intervention, so it must be chosen and interpreted with awareness of possible artifacts.

When is cryo-EM preferable to chemical fixation?

Cryo-EM is preferable when maintaining a hydrated, near-native state is central to the question, especially for macromolecular structure or in situ cellular studies. It is not automatically easier. Grid preparation, vitrification quality, ice thickness and particle behavior can determine whether the method succeeds.

Does sputter coating change SEM results?

Yes, it can. A conductive coating reduces charging and can improve image stability, but it adds a layer to the surface. If nanoscale texture, particle size or elemental analysis is the main concern, coating material and thickness should be selected carefully.

What is the most common reason EM preparation fails?

There is no single cause, but common failures include delayed stabilization, incomplete dehydration or infiltration, poor drying of hydrated samples, charging in SEM, contamination, and insufficient cryo grid optimization. Most failures become easier to diagnose when preparation variables are recorded consistently.