SEM (scanning electron microscopy) and TEM (transmission electron microscopy) both use beams of electrons instead of light to image tiny structures, but they work in fundamentally different ways. SEM scans the surface of a sample to produce detailed 3D-like images of its topography, while TEM shoots electrons through an ultra-thin sample to reveal its internal structure, down to individual atomic arrangements. This core difference shapes everything else: how samples are prepared, what you can see, how much magnification is possible, and what each technique costs to run.
How Each Microscope Works
In an SEM, a focused beam of high-energy electrons scans across the surface of a solid sample point by point, like a flashlight sweeping across a dark room. When those electrons hit the sample, two things happen. Some bounce back (called backscattered electrons), deflected by the atoms in the material. Others knock loose electrons from the sample’s own atoms, generating low-energy secondary electrons from the surface layers. Detectors pick up these signals and convert them into a grayscale image that reveals the surface shape and composition of whatever you’re looking at.
TEM takes the opposite approach. Instead of bouncing electrons off a surface, it fires them through the sample entirely. The specimen has to be thin enough for electrons to pass through it, and the image forms based on how those electrons interact with the material on the way through. Dense regions block or scatter more electrons and appear darker, while thinner or less dense areas let more through and appear lighter. The result is a view of the sample’s internal architecture: crystal lattices, protein arrangements, organelle membranes, and other structures invisible from the outside.
Resolution and Magnification
This is where the gap between the two techniques gets dramatic. SEM can resolve features down to about 0.5 nanometers, which is impressive enough to image nanoparticles and fine surface textures. TEM, though, operates on a different level entirely. Modern aberration-corrected TEMs have achieved spatial resolution below 50 picometers, roughly ten times sharper than the best SEM can manage. That’s enough to distinguish individual atoms in a crystal.
Magnification follows a similar pattern. SEM tops out at roughly 1 to 2 million times magnification. TEM can push past 50 million times. For most biological and materials science work, you won’t need the extreme end of either range, but TEM’s capacity to resolve atomic-scale detail is the reason it remains the gold standard for studying crystal structure, lattice defects, and nanoscale interfaces.
What Each Technique Shows You
SEM excels at surface detail. Because it detects electrons coming off the top layers of a sample, the images have a distinctive three-dimensional quality that makes them intuitive to interpret. You can see the shape of a pollen grain, the texture of a fractured metal surface, or the topography of a cell membrane. It’s particularly useful for studying anything where surface morphology matters: corrosion patterns, fiber structures, microbial biofilms, the architecture of nervous system components, or the 3D arrangement of synapses.
TEM, by contrast, reveals what’s inside. Researchers use it to image chromosome segregation during cell division, the organization of spindle fibers, the arrangement of protein clusters at cell membranes, and the internal lattice structure of crystals. One research group used TEM diffraction to study protein crystallization at a resolution that allowed them to map internal strain and crystallite density within individual protein crystals. If your question is about what a material looks like, SEM is the tool. If your question is about how it’s organized internally, TEM is the answer.
Sample Preparation
Preparing a sample for SEM is generally simpler. Most solid specimens can be mounted directly, though non-conductive materials (polymers, ceramics, biological tissue) need a thin conductive coating first. Without it, electrons accumulate on the surface, creating bright white patches called “charging” that distort the image. The standard fix is sputter coating: depositing a layer about 10 nanometers thick of a conductive metal. Gold has historically been the most popular choice because of its high conductivity and fine grain size. For ultra-high-resolution work, materials like iridium, tungsten, or chromium produce even smaller grains. If chemical analysis is part of the plan, carbon coating is preferred because its signal doesn’t overlap with other elements of interest.
TEM preparation is far more demanding. The sample has to be thin enough for electrons to pass through, which for biological specimens typically means slicing it to 300 to 500 nanometers, and for materials science applications often thinner still. Specimens also need to be trimmed to about 3 millimeters in diameter to fit into the microscope column. Achieving uniform sections at these scales requires specialized tools like ultramicrotomes or focused ion beam milling, and the process can take hours to days depending on the material. This preparation step is one of the biggest practical differences between the two techniques. It adds time, cost, and complexity, and it limits what kinds of samples can be studied with TEM.
Operating Environment
Both SEM and TEM operate in a vacuum, since air molecules would scatter the electron beam and degrade the image. Standard SEM uses high vacuum conditions, but a variant called environmental SEM (ESEM) can operate in low vacuum or even wet mode. This makes it possible to image hydrated biological samples or materials that would be damaged by full vacuum, without the usual dehydration and coating steps. TEM generally requires stricter vacuum conditions throughout the column, with less flexibility for wet or unprocessed samples.
Cost and Complexity
SEM systems are less expensive to purchase, maintain, and operate than TEMs. Entry-level SEMs are accessible enough that many university departments and industrial labs have one on site. TEMs, with their more demanding vacuum systems, higher voltage electron sources, and precision optics, cost significantly more and require more specialized facilities, including vibration-dampened rooms and stable power supplies.
Training reflects this complexity gap. At one university microscopy facility, SEM training requires a minimum of three hours of staff time, while TEM training requires at least four. In practice, the learning curve for TEM is steeper because the sample preparation alone involves techniques that take considerable time to master. The operating costs per session also tend to be higher for TEM, both because of the instrument’s maintenance demands and the labor-intensive specimen preparation.
Choosing Between Them
The choice between SEM and TEM almost always comes down to what you need to see. If you need surface morphology, texture, or the shape of structures at the micro or nanoscale, SEM is faster, cheaper, and easier. If you need to see internal structure, atomic arrangement, or crystallographic detail, TEM is the only option that delivers sufficient resolution. Many research projects use both: SEM to survey the surface and identify regions of interest, then TEM to examine the internal structure of those specific areas at higher resolution.
For biological work, SEM is the natural choice for imaging the outer surfaces of cells, tissues, and organisms, while TEM is essential for studying organelles, membranes, and molecular complexes inside cells. In materials science, SEM handles fracture analysis, surface contamination, and particle characterization, while TEM is used for thin film analysis, crystal defect mapping, and atomic-resolution imaging of interfaces.