Time-of-flight secondary ion mass spectrometry imaging, commonly written ToF-SIMS imaging, maps the lateral and, with depth profiling, the vertical distribution of chemical species across a semiconductor sample by pulsing a focused primary ion beam and time-resolving the secondary ions it ejects. Because the time-of-flight mass analyzer records essentially the full mass spectrum from every raster pixel in parallel, the technique yields chemically specific two-dimensional and three-dimensional maps rather than a single bulk composition number, which is what makes it a preferred tool for defect review, contamination mapping, dopant imaging, and metallurgy studies across the semiconductor fab and the failure-analysis lab. Unlike scanning mass filters that step through one mass channel at a time, a pulsed ToF source captures the entire spectrum per pixel in one shot, so a chemical image can be reconstructed for any species after the raster is complete without ever revisiting the sample.

Pulse the beam and time-resolve every mass at once.

Time-of-flight secondary ion mass spectrometry departs from scanning mass analyzers by pulsing the primary beam for roughly 1 ns and letting every secondary ion packet drift the same field-free path before it strikes the detector. Ions are accelerated to a fixed potential near 2000 V, so lighter fragments outrun heavier ones and arrive at distinct times; a single pulse therefore yields the complete mass spectrum for that pixel rather than one mass channel at a time. Flight times through a drift path of about 120 cm typically span from a few hundred ns up to 30 µs, and a reflectron can fold the path to roughly 2 × its physical length, raising mass resolution without enlarging the tool footprint. Because acquisition is parallel across mass, dwell time per pixel is spent once rather than once per element, which is what makes full-frame chemical imaging practical within a workable acquisition frame time near 200 ms per line.

Sharper images cost some mass resolution at the pixel level.

Lateral resolution and mass resolution trade against each other because both draw on the same finite ion budget delivered by the primary column. Narrowing the probe below 100 nm sharpens the image and can reach lateral resolution near 50 nm, but fewer primary ions land on each pixel, which thins the secondary ion signal available to define a mass peak cleanly. Bunching the pulse to preserve timing precision keeps duty cycle near 1 %, so operators routinely choose between a high-resolution imaging mode optimized for lateral detail and a high mass-resolution spectroscopy mode optimized for isobar separation, sometimes toggling between the two settings on the same die to answer different questions from the same defect.

Turn a raster into a two-dimensional chemical map.

Rastering the pulsed beam across a defined field converts point spectra into a spatial data cube: every one of, for example, 256 x 256 or 512 x 512 raster pixels carries its own complete mass spectrum. Any mass channel can be extracted after acquisition to render a two-dimensional ion-distribution map for a specific species without re-running the experiment, and several channels can be overlaid as a false-color composite to show co-location or exclusion of elements across a field between 100 µm x 100 µm and 500 µm x 500 µm. Pixel size as fine as 0.4 µm keeps sub-µm features resolvable, and correlating the ion map with an AFM topography scan or an XPS chemical-state survey over the same region strengthens the interpretation of what a bright pixel actually represents on a device that a wafer map alone cannot explain.

Acquire a baseline 2D ion map at the native surface
  -> select mass channels for the target dopant or contamination species
  -> raster the full field and store per-pixel mass spectra
  -> sputter a controlled depth increment across the same field
  -> re-acquire the raster to capture the next depth-resolved slice
  -> repeat raster and sputter cycles through the region of interest
  -> stack the slices into a 3D chemical volume
  -> co-register the volume with SEM, AFM, or XPS evidence
  -> report defect, dopant, or metallurgy distribution to the process team

Depth-resolved imaging builds a chemical volume one slice at a time.

Three-dimensional imaging is assembled from successive raster-and-sputter cycles rather than one continuous erosion. An analysis raster records a 2D slice, a separate sputter raster removes a controlled increment often near 0.5 nm per cycle, and the analysis raster then returns to record the next slice from the fresh crater floor; stacking the slices yields a depth-resolved 3D chemical volume with depth resolution that can approach 2 nm under favorable matrix conditions. Because crater-edge and shadowing effects distort the outer few µm of the field, the sputtered area is normally kept several times larger than the analyzed area, and dopant profiles extracted this way are often cross-checked against Hall effect carrier measurements or four-point probe sheet-resistance mapping to confirm that ion-count depth profiles track active carrier concentration rather than total chemical dose alone.

Defect review and contamination mapping lean on ToF-SIMS specificity.

Defect review campaigns route a stage-navigated coordinate from an optical or SEM inspection tool directly into the ToF-SIMS imaging recipe, so the same particle or residue that triggered an excursion at a 20 nm design rule is the one interrogated for chemistry. An ion map covering the defect and a clean reference area distinguishes a genuine foreign-material signature from ordinary matrix variation, and NIST-traceable reference materials support the quantification needed to compare counts against a control region rather than an assumed baseline. Because ToF-SIMS is destructive at the analyzed spot, teams typically confirm morphology and location with SEM first and reserve the ion beam for the chemical question that other tools cannot answer directly, then archive the ion map alongside the wafer map and recipe history for the disposition record.

Dopant and metallurgy imaging round out the fab application set.

Dopant imaging maps boron, phosphorus, or arsenic distributions across a die or across a full 300 mm wafer, revealing implant-angle effects, shadowing near topography, and anneal-driven redistribution that a single-point measurement would average away entirely. Metallurgy imaging follows barrier and liner diffusion in interconnect stacks, flagging where a Ta or Ti barrier has thinned locally or where Cu has migrated toward a dielectric before it becomes a reliability failure years later. These chemical maps are strongest alongside electrical evidence: a Keithley source-measure unit or a Keysight parametric analyzer can localize a leakage path to a device, Semilab corona-Kelvin metrology can flag a surface-charge anomaly, ellipsometry can confirm whether a film thickness shift near 2 % accompanies the chemical signature, and DLTS can characterize whether a trap level near 0.3 eV correlates with the contaminant species identified in the ion map.

Viewed through a chemical-imaging-localization lens, ToF-SIMS imaging earns its place in the metrology flow not by replacing SEM, AFM, XPS, four-point probe, or electrical test, but by answering the one question those tools cannot resolve alone: which element or compound sits at this exact pixel, and how does that distribution change with depth. A raster that starts as a single pulsed-beam spectrum and ends as a registered 3D chemical volume turns a contamination call, a dopant excursion, or a metallurgy failure into a defensible, spatially resolved answer rather than an inference drawn from bulk composition data.