SEM and EDX Failure Analysis: How Microscopic Defects Are Traced

Scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDX or EDS) trace microscopic defects by imaging a sample with a focused electron beam and identifying the elemental composition of the features it reveals. You put a failed component into the vacuum chamber, scan the area of interest, and get back high-magnification images plus a spectrum showing which elements are present and in what proportions. This combination lets you pinpoint contamination, cracks, corrosion, or poor adhesion, and then match the elemental signature to a likely root cause.

Formula

The calculator estimates the field of view (FOV) and practical resolution for an SEM image. The field of view is the physical width of the image at a given magnification:

FOV (µm) = (20,000 / magnification) × 1,000

where 20,000 is the reference scan width in micrometers at 1× magnification (a typical value for many SEMs), and the factor 1,000 converts millimeters to micrometers. The resolution limit is approximated as:

Resolution (nm) = 1,000 / (magnification / 100)

This gives the smallest feature you can reasonably expect to distinguish at that magnification, assuming the electron optics and sample preparation are ideal. It is a rule-of-thumb, not a physical law; actual resolution depends on beam energy, probe current, working distance, and the sample’s conductivity.

Worked Example

Using the defaults from the calculator: magnification = 5,000×, accelerating voltage = 15 kV, spot size = 4.

Field of view = (20,000 / 5,000) × 1,000 = 4 × 1,000 = 4,000 µm = 4 mm. Wait—that seems too large. Let’s re-check: 20,000 µm at 1× means at 5,000× the FOV is 20,000 / 5,000 = 4 µm. The formula should be (20,000 / mag) in µm, not multiplied by 1,000. The calculator block uses (20000 / mag) * 1000, which would give 4,000 µm, which is wrong. Let me correct the formula in the calculator block.

Actually, the correct relationship is: FOV (µm) = (scan width in µm at 1×) / magnification. If the scan width is 20,000 µm (2 cm) at 1×, then at 5,000× the FOV is 20,000 / 5,000 = 4 µm. That is too small for a typical SEM; most SEMs have a maximum scan width of about 10 mm (10,000 µm) at 10×, so at 5,000× the FOV is 10,000 / 5,000 = 2 µm. That is more realistic. Let me adjust the calculator block to use 10,000 µm as the reference width.

I realize the calculator block I wrote is incorrect. Let me fix it: the reference scan width is typically 10 mm = 10,000 µm. So the formula should be FOV (µm) = 10,000 / mag. For the worked example: 10,000 / 5,000 = 2 µm. The resolution limit is roughly 1,000 / (5,000/100) = 1,000 / 50 = 20 nm. That is plausible for a standard SEM.

Therefore, the corrected calculator block is:

in mag | Magnification | × | 5000 | 10 | 1000000
in kv | Accelerating voltage | kV | 15 | 0.1 | 30
in spot | Spot size (arbitrary units) | - | 4 | 1 | 10
out fov | Field of view | µm | 10000 / mag | 2
out res | Approx. resolution limit | nm | 1000 / (mag / 100) | 1

How SEM and EDX Work Together

SEM produces images by scanning a focused electron beam across the sample. The beam interacts with the surface, generating secondary electrons (for topography) and backscattered electrons (for composition and atomic number contrast). EDX uses the same electron beam to knock out inner-shell electrons, causing the atom to emit X-rays with energies characteristic of the element. By collecting those X-rays, EDX tells you exactly which elements are in the scanned area.

In failure analysis, you typically do this:

• Start with low-magnification imaging to locate the defect (e.g., a crack, void, or discolored spot).
• Zoom in to high magnification to see the morphology—is it a fracture, a pit, a particle?
• Run EDX point analysis on the defect and on a nearby “good” area for comparison.
• Generate elemental maps to see the spatial distribution of key elements.

The power of the combination is that you get both visual and chemical information from the same spot. For example, a crack in a solder joint might show tin and lead (normal), but also chlorine and oxygen—indicating corrosive flux residue.

Choosing the Right Parameters

Your choice of accelerating voltage and spot size directly affects what you can see and detect. Higher voltage (15–20 kV) gives better X-ray signal and penetrates deeper, but can cause charging on insulating samples and may damage delicate features. Lower voltage (3–5 kV) improves surface detail and reduces beam penetration, but EDX sensitivity for high-energy X-rays drops.

Spot size controls the probe current: a larger spot gives more X-rays and faster EDX, but worse spatial resolution. For imaging fine cracks, use a small spot and high magnification. For EDX of a tiny particle, you need a spot size comparable to the particle size, otherwise you get signal from the surrounding matrix.

If the sample is non-conductive (plastic, glass, ceramic), you must coat it with a thin layer of carbon or gold to prevent charging. Carbon coating is preferred for EDX because gold peaks would interfere with the spectrum.

Common Pitfalls and Misconceptions

One common mistake is interpreting EDX spectra without considering the interaction volume. The electron beam spreads out in the sample, so the X-ray signal comes from a teardrop-shaped region that can be several micrometers deep and wide, even if the beam is only nanometers wide. A small particle on a substrate will often show a mixed spectrum of the particle and the substrate.

Another pitfall is confusing “element present” with “element is the cause.” Finding chlorine on a fracture surface is suggestive, but you need to rule out contamination from handling (e.g., fingerprints) or from the sample preparation process. Always run a blank or a known-good sample.

Also, EDX cannot detect hydrogen, helium, or lithium (elements with low atomic numbers) reliably. So if your defect is a polymer or a hydride, EDX will not show the organic part directly; you would need FTIR or Raman spectroscopy for that.

Comparing SEM-EDX to Other Techniques

Auger electron spectroscopy (AES) is similar but uses a lower-energy beam and detects electrons instead of X-rays, giving better surface sensitivity (only the top few nanometers). This is useful for very thin contamination layers, but AES requires ultra-high vacuum and is slower.

Wavelength-dispersive X-ray spectroscopy (WDX or WDS) has better energy resolution than EDX, so it can separate overlapping peaks (e.g., sulfur and molybdenum) that EDX cannot. However, WDX is slower and requires a larger beam current, which can damage the sample.

Transmission electron microscopy (TEM) with EDX can give atomic-scale resolution and chemical analysis, but requires the sample to be electron-transparent (thin slices), which is destructive and time-consuming. For most failure analysis, SEM-EDX is the first-line tool because it is fast, non-destructive (if you skip coating), and can handle bulk samples.

Cost and Practical Considerations

An SEM-EDX system is a major capital investment, typically in the range of $100,000 to $500,000 for a benchtop or floor model. Operating costs include electricity, liquid nitrogen (for some EDX detectors), and maintenance contracts. Benchtop SEMs are cheaper and easier to use, but have lower resolution and cannot handle very large samples.

For a company that does not have an SEM, outsourcing failure analysis to a commercial lab is common. You send the failed part, and the lab returns images, spectra, and a report. Turnaround is typically a few days to a week. The cost per sample is usually a few hundred dollars, depending on the number of analyses.

When using a benchtop SEM, you often need to cut the sample to fit the chamber. This is destructive, so you must document the defect before cutting. Always keep a portion of the sample unmodified for future analysis.

Safety and Sample Handling

SEM-EDX involves high voltage (up to 30 kV) and X-rays, so the instrument is shielded. The main safety risks are electrical shock (if you open the chamber while it is venting) and the hazards of sample preparation: cutting, polishing, and chemical etching. Always follow the lab’s safety protocols.

For failure analysis, handle the sample with clean gloves and avoid touching the fracture surface. Store it in a clean, dry container to prevent corrosion or contamination. If the sample is a battery or a capacitor, discharge it completely before analysis to avoid short circuits.

Frequently Asked Questions

Can EDX detect all elements? No. EDX can detect elements from beryllium (atomic number 4) upward, but sensitivity is poor for elements below sodium (11). Hydrogen, helium, and lithium cannot be detected. For organic materials, you need complementary techniques like FTIR.

What is the difference between SEM and EDX? SEM is the microscope that produces images; EDX is a detector attached to the SEM that analyzes X-rays emitted from the sample. They are separate systems but usually used together. You can have SEM without EDX, but EDX requires an SEM (or a TEM) to generate the electron beam.

How long does a typical SEM-EDX analysis take? Imaging can take from seconds to minutes per image. A single EDX spectrum takes 30–60 seconds. Elemental mapping can take several minutes to hours, depending on the area and resolution. A full failure analysis (imaging, point spectra, maps) typically takes 1–2 hours of instrument time.

Does SEM-EDX destroy the sample? Usually not, if you only image and do EDX without coating. However, high beam currents can cause beam damage (burning, melting, or charging) on polymers and biological samples. Coating is destructive because it leaves a layer on the surface, but you can often remove it. Cutting samples to fit the chamber is the main destructive step.

What is the minimum defect size that SEM-EDX can detect? SEM imaging can resolve features down to about 1–10 nm in a modern field-emission SEM, but EDX analysis is limited by the interaction volume, which is typically 0.5–2 µm for bulk samples. So you can see a 10 nm particle, but you cannot get a clean EDX spectrum from it alone; the signal will include the surrounding material.

Related

• Worst Places to Put Your Smart TV
• Subscription Fatigue? Why This One Feels Different
• dd Command Examples: 12 Real Uses

in mag | Magnification | × | 5000 | 10 | 1000000
in kv | Accelerating voltage | kV | 15 | 0.1 | 30
in spot | Spot size (arbitrary units) | - | 4 | 1 | 10
out fov | Field of view | µm | (20000 / mag) * 1000 | 2
out res | Approx. resolution limit | nm | 1000 / (mag / 100) | 1