User documentation

APT Calibrator

A practical guide to installation, the AP Suite workflow, the four core modules described in Chen et al., Microscopy and Microanalysis (2025), and the 3D analysis tools in current APT Calibrator releases.

1. Installation

APT Calibrator is a standalone Windows application. It does not require AP Suite to launch, but most calibration workflows exchange data with AP Suite 6.x.

• Operating system: Windows 10 or Windows 11
• Runtime: .NET Framework 4.6
• Installer: download page (sign in required)

Run the installer, then open APT Calibrator from the Start menu. The main windows include DAS APT, Dynamic Reconstruction, and Cluster Analysis, plus ICF and kf tools, the pole-figure simulator, and a runtime calculator. Installed copies can check dasapt.com for updates.

Initialization fields typically include flight path, lattice constants, crystal structure, and an initial tip radius. The radius can come from TEM/SEM, or from an empirical starting guess that is refined later.

2. File formats and AP Suite

Two interaction modes are supported.

Standard calibration

1. In AP Suite, use Reconstruction Wizard / Reconstruction Explorer to generate detector hit maps and spatial distribution maps.
2. Index poles and measure plane spacing as described below.
3. Enter the measured values in APT Calibrator to obtain ICF and k_f.
4. Reapply those parameters in AP Suite and rebuild the reconstruction.

Correlative dynamic reconstruction

1. Perform a rough reconstruction in AP Suite.
2. Export .apt and convert to .epos (ion sequence, voltage, mass-to-charge, detector coordinates).
3. Load the dataset in APT Calibrator and place golden seeds against TEM/STEM/4D-STEM fiducials.
4. Compute depth-resolved ICF and k_f.
5. Either reimport parameters into AP Suite, or reconstruct in APT Calibrator and export .pos.

3. ICF calibration

Image compression factor ξ is obtained from crystallographic poles on the detector hit map (Gault et al., 2008, 2009):

ξ = θcrys / θobs

θ_crys is the theoretical angle between Miller indices. θ_obs is the observed angle from pole coordinates on the detector divided by the flight path. The software uses an equidistant azimuthal / Hawkes–Kasper angular magnification scheme.

1. Open a detector hit map and identify at least two poles. The pole-figure repository and simulator in Tools help assign Miller indices, including signs.
2. Enter Miller indices and detector coordinates (x, y) for each pole.
3. APT Calibrator returns the ICF.

Symmetry-equivalent index sets depend on the projection axis. Use the pole-figure simulator to rotate the crystal until the overlay matches the hit map before locking the assignment.

4. Standard k_f iteration

Field factor follows from a calibrated apex radius:

F = V / (kf R)

1. Choose an initial tip radius, typically 20–100 nm, such that spatial distribution map peaks are visible in AP Suite.
2. Measure observed interplanar spacing at a selected pole (for example Al (111) = 0.234 nm theoretically).
3. Enter the observed spacing. The software updates the radius so that reconstructed spacing approaches the theoretical value (Ceguerra et al., 2019).
4. Rebuild the spatial distribution map with the new radius and repeat until spacing converges.
5. Combine the converged radius with evaporation field F (tabulated, Müller / Brandon charge-state selection) and the initial voltage V to obtain k_f.

The reconstruction of atoms at the pole for the z-direction spatial distribution map is documented in supplementary videos S1 and S2 of the paper.

5. Pole-figure simulator

The simulator uses an azimuthal equidistant projection, which is more appropriate for APT than the gnomonic projection used in EBSD-style Hough indexing.

• Inputs: crystal structure, lattice constants, plane-spacing threshold D_plane, flight path, ICF, and Bunge–Euler angles.
• Lower D_plane shows more zone lines (for example 0.10 nm vs 0.15 nm).
• Quick settings: LEAP 5000 XS (straight flight path) and LEAP 5000 XR (reflectron). Enable the Reflectron checkbox for XR; detector size and mounting angle differ.
• Angle correction aligns the sample reference frame with the detector hit map, which is often rotated by stage and detector mounting.
• Euler angles may be typed manually or taken from TKD or EBSD. Grain orientations can be stored with Save Grain to compute misorientation angle and rotation axis, including screening for low-Σ CSL boundaries.

Newer instruments (LEAP 6000 XR, Invizo 6000) can be approximated by updating detector size, projection geometry, and field of view. Lens-edge distortions on Invizo are not corrected; that would require Cameca’s proprietary mapping.

6. Golden-seed dynamic reconstruction

ICF and k_f change with shank angle, radius, and evaporation sequence. When poles or lattice spacing are unavailable, fiducials from TEM/STEM still constrain depth scaling.

1. Start from a static reconstruction (example values in the paper: ICF = 1.5, k_f = 3 for Sm–Co–Cu–Zr; ICF = 1.65, k_f = 1.25 for Fe–1W).
2. Place several golden seeds on distinctive features in both the APT volume and the STEM/TEM image (phase boundaries, precipitates, grain-boundary junctions).
3. Compare inter-seed distances versus depth. Systematic scaling error is converted into a depth-dependent correction of k_f/ξ.
4. Using the empirical relation k_f* ∝ (ξ*)³, separate depth profiles of k_f(Z) and ξ(Z).
5. Rebuild. Depth (flight direction) should then match the STEM geometry. Only the central APT field of view is reconstructed; the TEM periphery often lies outside the ion collection angle.

APT Calibrator adopts the Gault protocol for dv/dz rather than a simple reverse-projection area. The Hatzoglou dynamic-voltage protocol used in AP Suite is also available for comparison. In an Al–Mg–Fe example, dynamic k_f reduced deviation from theoretical (111) spacing by about 20%.

7. 3D view and RRNG editing

From version 4.0, APT Calibrator can display a reconstructed volume, edit mass ranges on the spectrum, and colour ions in 3D. Open Cluster Analysis or DAS APT. The toolbar includes RRNG, Cylinder, and related analysis tools.

Load data

1. Open a .pos or .epos file. If the EPOS already contains reconstructed XYZ, the volume is shown directly and reconstruction parameters are skipped. Otherwise a structure dialog lists the fields in the file (XYZ, m/z, TOF, voltage, detector hits, multiplicity).
2. Load a .rrng if you have one (Load RRNG), or build ranges on the mass spectrum as below. The 3D legend checkboxes hide or show each range without changing the file.
3. Optional: colour by multiplicity (singles, doubles, triples, 4+); open linked views for different ions; copy the 3D image; run a rotate animation and export GIF or MP4.

Edit ranges on the mass spectrum

Show the mass spectrum after a file is loaded. Use the RRNG button if you also want the range table. Button names below match the English interface.

1. Add a peak: right-click on a peak and choose Add peak at … Da. The software estimates a window around that peak. Type the ion name (for example Fe or Fe++).
2. Add a range by dragging: hold the right mouse button and drag across the peak (orange box). On release you get the same ion-name dialog, using the Da interval you drew. Use this when the automatic window is too wide or too narrow.
3. Adjust bounds: move the cursor to the left or right coloured edge until it becomes a left–right arrow, then drag. The selected range is drawn stronger. You can also edit numbers in the RRNG table.
4. Delete: right-click inside an existing range and choose Delete, or select the row in the RRNG table and use Delete.
5. Apply: changing ranges does not recolour the 3D view until you click Update analysis (a * appears while edits are unapplied). Spectrum zoom is kept. If you already computed an iso-surface or volume map, recompute those after the new ranges.
6. Save: Save RRNG writes the current file; Export RRNG asks for a new path. Both are also on the right-click menu.

Left-drag on the spectrum zooms; the mouse wheel zooms at the cursor; double-click restores the full range.

8. Cylinder 1D composition profile

The cylinder ROI measures composition along an axis in the reconstructed volume. You need XYZ coordinates (a .pos file, or an .epos that already contains XYZ). Click Cylinder on the toolbar.

1. A half-transparent solid cylinder appears in the 3D view, and the Cylinder 1D profile window opens. Fit to data places it on the specimen (default axis is specimen Z).
2. Move: drag a red / green / blue arrow to slide along X / Y / Z, or drag the cylinder body to move it freely.
3. Rotate: drag a coloured ring. The axis list switches to Custom when it is no longer aligned with X, Y or Z. You can also pick X, Y or Z (specimen) in the window.
4. Radius: drag the cylinder wall, or type Radius (nm). Length is set in the window (Length, nm) or by Fit to data. The 3D gizmo does not stretch the length.
5. Optional: Clip atoms to cylinder hides atoms outside the ROI in the 3D view. Show cylinder only toggles the overlay.
6. Set Bin width (nm), then click Compute 1D profile. The plot uses the current RRNG colours. Quantity can be concentration (at.%), counts, or density (at/nm³). Distance is along the cylinder axis.
7. If the cylinder is empty, increase radius or length, or Fit to data again. Export CSV writes the profile. Clear removes the ROI.

Composition follows the ranges applied with Update analysis. Edit RRNG and click Update analysis first if the ion colours should change.

9. Limitations

• Trajectory aberrations are not corrected.
• Evaporation field F is still a nominal input; heterogeneous or coated tips can vary by tens of V/nm.
• Pole-based ICF is only available when crystallographic contrast is resolved.
• Golden-seed alignment mainly constrains the specimen z-axis; small rotations about z have limited effect but should still be checked.
• The cylinder 1D profile needs reconstructed XYZ in the POS or EPOS file. Ranges that have not been applied with Update analysis are not used in the profile.

10. Citation

Chen X, Li Y, Woods E, Zhou X, Gault B. APT Calibrator: A Multifunctional C# Software Application for Reconstruction Calibration in Atom Probe Tomography. Microscopy and Microanalysis. 2025;31:ozaf092. doi:10.1093/mam/ozaf092

Supplementary videos and figures: same DOI.

Corresponding authors: Xinren Chen (x.chen@mpi-susmat.de) and Baptiste Gault (b.gault@mpi-susmat.de).