Transmission Electron Microscopy (TEM) Imaging and Analysis Guide

Transmission electron microscopy (TEM) is a powerful technique for directly imaging nanoparticles and measuring physical particle dimensions, size distributions, morphology, and structural features such as cores and shells.

Unlike solution-based sizing methods, TEM produces a high-resolution two-dimensional image of particles deposited on an electron-transparent support. This makes TEM particularly useful for measuring primary particle size and morphology, but sample preparation, image selection, and analysis methods must be carefully controlled to obtain representative results.

On this page

• What TEM Measures
• TEM Contrast and Sample Types
• TEM Sample Preparation
• Choosing an Imaging Strategy
• Particle Size and Morphology Analysis
• Common TEM Artifacts and Challenges
• TEM vs. Other Particle Sizing Methods
• Reporting TEM Results

What Does Transmission Electron Microscopy Measure?

Transmission electron microscopy directs a focused beam of high-energy electrons through a thin specimen. Electrons transmitted through and scattered by the sample are used to form a highly magnified image.

Because electrons have much shorter wavelengths than visible light, TEM can resolve nanoscale structures that cannot be directly resolved using conventional optical microscopy.

Schematic of a transmission electron microscope. An electron beam passes through a thin specimen and is focused to form a high-resolution image.

For nanoparticle characterization, TEM is commonly used to evaluate:

• Primary particle diameter
• Particle length, width, and aspect ratio
• Particle size distribution
• Particle shape and morphology
• Core-shell dimensions
• Surface and structural features that are sufficiently resolved
• Evidence of particle fusion, sintering, or other morphological changes

TEM Measures a Two-Dimensional Projection

A TEM image is a two-dimensional projection of a three-dimensional object. The dimensions obtained from an image therefore depend on particle geometry and orientation on the grid.

For approximately spherical nanoparticles, projected diameter is generally straightforward to interpret. For rods, plates, irregular particles, aggregates, or other anisotropic structures, the measurement metric should be defined explicitly and applied consistently.

TEM also examines a dried specimen rather than the original liquid dispersion. It provides direct information about the physical particle structure, but it does not directly measure hydrodynamic size or colloidal behavior in solution.

TEM Contrast and Sample Types

Successful TEM imaging depends on contrast between the nanoparticle and the surrounding support film. Image contrast can arise from differences in sample thickness, composition, atomic number, crystallinity, and electron scattering behavior.

Many metals and metal-containing nanoparticles produce strong contrast in conventional TEM. Gold, silver, platinum, iron oxide, and other relatively electron-dense materials are generally straightforward to visualize against a thin carbon support.

Silica-coated gold nanoparticles imaged by TEM. The electron-dense gold cores appear substantially darker than the surrounding silica shells.

Low-Contrast Nanoparticles

Organic, polymeric, lipid-based, and biological particles can be more difficult to visualize because their electron density may be similar to that of the support film.

For suitable samples, negative-stain TEM can increase contrast by surrounding the particle with an electron-dense stain. This approach can support imaging of materials such as polymer particles, liposomes, vesicles, viruses, and other low-density structures.

Negative staining should be interpreted carefully. Staining and drying can alter soft structures, introduce flattening or deformation, and emphasize the particle boundary rather than the native hydrated structure. The apparent dimensions therefore depend on both the specimen and the preparation method.

TEM Sample Preparation for Nanoparticles

Sample preparation is one of the most important factors affecting nanoparticle TEM data. A high-resolution microscope cannot compensate for a grid that contains excessive residue, severe drying artifacts, too many overlapping particles, or too few particles for representative analysis.

Preparing Nanoparticles on a TEM Grid

Colloidal nanoparticle samples are commonly deposited onto an electron-transparent support film, such as a thin carbon or polymer film supported by a metal grid. A small aliquot of the dispersion is placed on the grid and excess liquid is removed or allowed to evaporate before imaging under vacuum.

The objective is to deposit enough particles for representative imaging while maintaining sufficient separation between particles and minimizing background residue.

Minimize Nonvolatile Residues

Salts, excess surfactants, polymers, unreacted synthesis reagents, and other nonvolatile components can become concentrated as the droplet dries. These materials can create background films, obscure particle boundaries, decrease contrast, or make automated image analysis difficult.

When residual components interfere with imaging, an appropriate purification or buffer-exchange step may improve grid quality. The method should be selected carefully because centrifugation, filtration, or other processing can also alter the nanoparticle population.

Use an Appropriate Particle Concentration

Highly dilute samples may deposit too few particles for efficient imaging or quantitative analysis. Excessively concentrated samples can create overlapping particles, multilayers, or dense regions that make individual particle boundaries difficult to resolve.

If concentration must be adjusted, use a method appropriate for the particle formulation. Centrifugation can concentrate many nanoparticle samples but may also promote aggregation or preferentially recover larger particles. Partial solvent evaporation may change ionic strength or soluble-component concentrations.

Any concentration or purification step should therefore be documented as part of the TEM sample preparation procedure.

Drying Can Change Particle Arrangement

As a liquid droplet dries on a TEM grid, solvent evaporation, surface tension, particle diffusion, wetting behavior, and interactions with the support film can redistribute nanoparticles.

Particles that were well dispersed in solution may move together during drying and appear clustered on the final TEM grid. Conversely, a preparation method may selectively deposit certain particles or particle sizes.

For this reason, clusters observed in a conventional dried TEM specimen should not automatically be interpreted as evidence that the nanoparticles were aggregated in the original solution.

Choosing a TEM Imaging Strategy

The appropriate magnification and number of images depend on the question being asked. A measurement intended to generate statistically meaningful particle size distributions requires a different imaging strategy from one intended to document a specific structural feature.

Imaging Goal	Recommended Approach
Particle size and morphology statistics	Collect multiple representative fields at a magnification that clearly resolves particle boundaries while including enough particles for efficient measurement.
Sample survey	Collect images across several magnifications and grid locations to evaluate overall morphology, heterogeneity, particle density, and unusual populations.
High-magnification structural imaging	Image individual particles or small groups at higher magnification to examine features such as shells, surface morphology, fusion, or other fine structure.

Avoid Image-Selection Bias

Images used for quantitative particle sizing should represent the sample rather than only the most visually appealing or densely populated regions of the grid.

Use a systematic or otherwise predefined approach to select grid regions and fields of view. If obvious preparation artifacts or unusable regions are excluded, document the exclusion criteria and apply them consistently.

Survey images can be selected to illustrate interesting or unusual structures, but these images should not be presented as representative of the particle-size distribution unless they were collected using an appropriate sampling strategy.

Particle Size and Morphology Analysis

TEM images can be analyzed manually or with image-analysis software to generate quantitative particle-size and shape distributions. ImageJ and other image-processing platforms provide tools for calibrating images, identifying particle boundaries, and measuring individual structures.

Define the Measurement Before Measuring Particles

The appropriate size metric depends on particle geometry. Common measurements include:

• Projected diameter: useful for approximately spherical particles.
• Area-equivalent circular diameter: diameter of a circle with the same projected area as the measured particle.
• Feret diameter: distance between two parallel tangents to the particle boundary at a defined orientation.
• Major and minor axes: useful for elongated or elliptical particles.
• Aspect ratio: commonly calculated as major-axis length divided by minor-axis width.
• Shell thickness: useful when core and shell boundaries can be independently resolved.

Do not mix different size definitions within the same distribution. The measurement method should remain consistent across all particles and samples being compared.

Spherical Nanoparticles

For isolated, approximately spherical particles, the physical diameter can usually be measured directly from the projected image. The resulting individual particle measurements can be used to calculate mean diameter, standard deviation, coefficient of variation, and a number-based particle-size distribution.

The coefficient of variation can be calculated as:

CV (%) = (standard deviation / mean diameter) × 100

Irregular and Anisotropic Nanoparticles

A single diameter may not adequately describe rods, plates, cubes, stars, irregular particles, or other anisotropic structures. Select dimensions that reflect the particle geometry and intended application, such as length and width for nanorods or edge length for nanocubes.

Area-equivalent circular diameter can provide a useful single-number metric for irregular projected shapes, but it should not be interpreted as the actual three-dimensional particle diameter.

Overlapping, Fused, and Sintered Particles

Particles that overlap in a two-dimensional TEM projection can be difficult to distinguish from particles that are physically fused. Visible necking or continuous material between particles may provide evidence of sintering or coalescence, but a single projection may not always resolve the distinction unambiguously.

Before quantitative analysis, define how touching, overlapping, fused, and partially obscured particles will be handled. Apply the same classification criteria throughout the dataset.

How Many Particles Should Be Measured?

The number of particles required depends on the width and complexity of the distribution and the statistical confidence needed from the result. Narrow, routine quality-control samples may require fewer measurements than broad or multimodal research samples.

Rather than relying on a universal minimum, collect enough particles from multiple representative fields to achieve an appropriately stable estimate of the mean and distribution width. Larger datasets generally improve confidence in distribution statistics and the ability to identify minority populations.

Common TEM Artifacts and Challenges

Observation	Possible Cause and Interpretation
Particles appear clustered on the grid	Clusters may reflect aggregation in solution, but they can also form during droplet drying. Compare with DLS or another solution-state technique before concluding that the original dispersion was aggregated.
Particle boundaries are difficult to identify	Residual salts, surfactants, polymers, or other nonvolatile material may have deposited around the particles. Consider appropriate sample purification or alternative preparation.
Very few particles appear in each image	The sample may be too dilute or particles may deposit nonuniformly. Consider an appropriate concentration method or optimized grid preparation.
Particles overlap extensively	The sample may be too concentrated or drying may have concentrated particles locally. Reduce deposition concentration or modify preparation conditions.
Soft particles appear flattened or distorted	Drying, vacuum exposure, staining, or the electron beam may alter low-density materials. Interpret dimensions in the context of the preparation method.
Particles change during imaging	Electron-beam damage, heating, contamination, or charging may be occurring. Reduce beam exposure or use an imaging strategy appropriate for beam-sensitive material.
Particle size differs across grid regions	Drying or deposition may have produced size-selective segregation. Review the preparation method and ensure quantitative images are sampled systematically.
Automated segmentation gives incorrect boundaries	Low contrast, overlapping particles, uneven background, or residue may interfere with thresholding. Validate automated measurements against manually inspected particles.

Particle Agglomeration on the Grid

The original nanoComposix TEM guidance emphasized an important limitation that remains highly relevant: sample drying can produce particle clusters that were not present in the starting dispersion. TEM should therefore be used primarily to determine physical particle dimensions and morphology rather than as the sole method for determining aggregation state in solution.

If the goal is to determine whether nanoparticles are aggregated in the original dispersion, combine TEM with a solution-state measurement such as dynamic light scattering (DLS) or another appropriate sizing technique.

Beam Damage

Some nanoparticles and surface coatings are sensitive to electron irradiation. Organic coatings, polymers, lipid structures, biomolecules, and other low-density materials may shrink, deform, carbonize, or otherwise change during imaging.

When beam damage is possible, minimize unnecessary exposure and compare sequential images when appropriate to determine whether the structure is changing under the electron beam.

TEM vs. Other Nanoparticle Sizing Methods

Different particle-sizing techniques measure different physical properties. TEM is particularly valuable because it directly images the primary particle, while most solution-based methods infer size from particle behavior.

Technique	Primary Measurement	Key Difference from TEM
TEM	Projected physical particle dimensions and morphology	Directly images a dried specimen and provides particle-by-particle structural information.
DLS	Equivalent hydrodynamic diameter in solution	Measures diffusion behavior and is highly sensitive to larger particles and aggregates.
DCS/CPS	Sedimentation-based size distribution	Provides high-resolution ensemble size distributions but requires appropriate particle-density assumptions.
NTA	Particle-by-particle hydrodynamic size and concentration	Tracks diffusion of individual scattering particles in liquid but does not directly image physical nanoparticle boundaries.
SEM	Particle surface morphology and projected dimensions	Images surfaces rather than transmitted electrons and may be advantageous for larger particles or surface morphology.

TEM and DLS are particularly complementary. TEM measures the physical dimensions of dried particles, while DLS measures effective hydrodynamic size in the liquid dispersion. Differences between the two measurements are expected and can provide useful information about surface coatings, solvation, aggregation, or sample heterogeneity.

What to Report with TEM Particle Size Data

For reproducible and interpretable TEM characterization, report enough information to understand how the specimen was prepared, imaged, and analyzed.

• Nanoparticle material and relevant surface chemistry
• Dispersing medium before grid preparation
• Any dilution, washing, centrifugation, staining, or other sample preparation
• Grid and support-film type when relevant
• Microscope accelerating voltage
• Image magnification or calibrated pixel size
• Number of images and grid regions used for quantitative analysis
• Number of particles measured
• Particle-size or shape metric used
• Rules for excluding overlapping, obscured, or otherwise unmeasurable particles
• Mean size, distribution width, and other reported statistics
• Whether images were processed or segmented before measurement

For quantitative work, image calibration should be traceable to an appropriate reference or instrument calibration procedure, and the same measurement criteria should be used for all samples being compared.

Selected References

1. Bonevich, J.E. and Haller, W.K. Measuring the Size of Nanoparticles Using Transmission Electron Microscopy (TEM). NIST-NCL Joint Assay Protocol PCC-7. 2010.
2. Rice, S.B., Chan, C., Brown, S.C., et al. Particle size distributions by transmission electron microscopy: an interlaboratory comparison case study. Metrologia. 2013;50(6):663-678.
3. Zhang, S. and Wang, C. Precise Analysis of Nanoparticle Size Distribution in TEM Image. Methods and Protocols. 2023;6(4):63.
4. ISO 21363:2020: Nanotechnologies — Measurements of particle size and shape distributions by transmission electron microscopy. International Organization for Standardization.

---

Other nanoparticle characterization guides

• Dynamic Light Scattering (DLS) Measurement and Analysis Guide
• Zeta Potential Measurement and Analysis Guide
• UV-Vis Spectroscopy Measurement and Analysis Guide
• ICP-MS Measurement and Analysis Guide
• Endotoxin Measurement and Analysis Guide