1. Phenomenon and consequence

Self-aligned double patterning (SADP) uses sidewall spacers around a lower-density mandrel pattern to create twice as many lines. In the ideal one-dimensional construction, spacer thickness and mandrel geometry are symmetric, and the resulting line centers are evenly spaced. Pitch walking appears when successive gaps alternate between two populations instead of repeating at one pitch.

The cadence matters even when the average pitch looks correct. A downstream cut or transfer step sees local edges, not just the array average, so an alternating displacement consumes edge-placement margin in opposite directions on adjacent features. Mandrel critical dimension, mandrel placement and profile, spacer deposition, and spacer etch can all contribute; pitch walking should not be assigned to lithographic overlay alone.

Educational scope. This article and preset explain a deliberately simplified geometry model using generalized dimensions. They do not reproduce a qualified process, predict yield, or replace CD/overlay metrology and process-window experiments. No claim of first-hand production results is made.

2. Mechanism in the simulator

The top-down mandrel polygons are authoritative geometry. At the selected cut Y, the 2D engine finds each sampled mandrel edge, places a spacer line beside it using the selected spacer thickness, and moves each left/right pair outward by the explicit pitch walk value. Outside the polygons' Y extent there is no spacer, so the worker-generated 3D view terminates the lines at the same layout boundary. Changing numeric mandrel pitch or width regenerates the preset's periodic polygons; manually dragged vertices remain authoritative until one of those two controls is changed again. Pitch walk is still a user-chosen geometry perturbation, not a result derived from deposition or etch chemistry.

For an idealized mandrel pitch P, mandrel width M, and spacer thickness S, the two spaces are governed by different parts of the construction: one inherits the former mandrel region, while the next inherits the region between adjacent mandrels after both spacers are present. A change in M moves those two space populations in opposite directions. A change or asymmetry in S changes spacer-line width and the adjacent spaces. The preset makes this alternating signature visible without claiming a unique physical root cause.

Use three comparisons:

Comparison Hold fixed Inspect
pitch walk = 0 vs preset value Pitch, mandrel width, spacer dimensions Whether line-center spacing changes from one population to two.
Mandrel width − small / nominal / large Pitch and spacer dimensions Opposite movement of the two gap populations.
Spacer thickness − small / nominal / large Mandrel geometry Coupled change in spacer CD and free-space width.

3. Open this case in the simulator

Open the SADP pitch-walking case in the simulator →

Compare the saved spacer result with pitch walk = 0 for the symmetric control. Then change one geometry input at a time. Dragging a mandrel vertex changes the sampled edge directly; changing numeric pitch or width rebuilds a periodic rectangular mandrel array while preserving its current Y extent.

4. Further reading

These primary conference papers demonstrate spacer-defined patterning and quantify how sacrificial-pattern and unit-process variation produce multiple spacing populations:

  1. C. Bencher, Y. Chen, H. Dai, W. Montgomery, and L. Huli, “22 nm half-pitch patterning by CVD spacer self alignment double patterning (SADP),” Proceedings of SPIE 6924, 69244E (2008). https://doi.org/10.1117/12.772953
  2. W. H. Arnold, “Towards 3nm overlay and critical dimension uniformity: an integrated error budget for double patterning lithography,” Proceedings of SPIE 6924, 692404 (2008). https://doi.org/10.1117/12.782311
  3. M. J. Maslow et al., “Co-optimization of exposure dose and etch process for SAQP pitch walk control,” Proceedings of SPIE 10587, 1058704 (2018). https://doi.org/10.1117/12.2297345