Applying Photo Induced Force Microscopy (PiFM) for Automated Defect Review (ADR) in the Semiconductor Processes

PiFM introduction

Photo-induced Force Microscopy (PiFM) is an advanced nanoscale characterization technique that integrates the high spatial resolution of Atomic Force Microscopy (AFM) with the chemical specificity of infrared (IR) spectroscopy. A tunable IR laser is focused at the probe tip, generating a tip-enhanced field that interacts with the sample and produces a photo-induced force signal (Figure 1). Sweeping the IR laser across relevant wavenumbers yields PiF-IR spectra that correlate directly with FTIR spectral signatures, enabling confident and precise chemical identification at the nanoscale. Operating the AFM cantilever in its second mechanical resonance and in non-contact mode enables high-resolution topographical imaging without physically contacting the surface, reducing the risk of tip wear or sample contamination.

Because the AFM probe and IR laser are modulated at complementing frequencies, PiFM can isolate both chemical and topographical signals, providing simultaneous physical and chemical mapping with <5 nm spatial resolution, as shown in an example of a PS-PMMA block copolymer in Figure 2. This unique capability makes PiFM ideal for analyzing monolayers, ALD films, polymer residues, process contaminants, and sub-100 nm semiconductor defects.

ADR: Current analytical approach

Defect identification on wafers is critical because even small defects can severely impact downstream chip performance and yield.  Defect review has traditionally relied on automated SEM/e-beam and EDX/EDS tools, which provide high‑resolution 2D imaging and elemental composition but lack the ability to give molecular information, which is critical in characterizing organics. While it delivers good structural detail, electron‑beams frequently damage both samples and defects, making it difficult to analyze volatile, low‑melting organic, and other fragile materials.​

ADR: PiFM advances the field

PiFM overcomes key limitations of traditional ADR technologies that struggle to measure increasingly small and complex defects, such as those that arise from organic residues, composite contamination, and monolayer films that SEM, EDX, and even TEM often cannot consistently detect or differentiate, to enable next-generation nanoscale defect characterization. Importantly, these existing ADR workflows lack non-destructive chemical identification tools capable of resolving such features, especially when the defect is invisible or ambiguous under electron-based review. Both organic and inorganic nano-sized defects are susceptible to physical damage, chemical alteration, and mass loss upon SEM/EDX analysis, which makes detailed analysis of ultra-small defects difficult in the current workflow [1].

PiFM can directly address this challenge, as it can chemically and topographically identify defects as small as 5 nm and produce correlated maps that distinguish true defect material from background texture or cleaning artifacts, even on bare silicon and complex multilayer coatings. On a wafer we recently analyzed, we were able to identify both e-beam damage from previous analysis at marked defect coordinates from SEM/EDS analysis and unaffected defects, which we were able to characterize without causing further damage (Figure 3).

Additionally, our Vista 300 instrument has the capability to automatically navigate directly to KLARF (KLA-Tencor Advanced Results File) defect coordinates on 300 mm wafers, which greatly increases both ease of defect location, a major time bottleneck in the defect review process, and significantly increase throughput. Here at Molecular Vista, we have been able to drastically increase our defect identification output, as displayed in Figure 4. Finally, because PiFM is non-contact, the wafers and defects remain intact for re-inspection and downstream analysis, eliminating unnecessary re-preparation of samples.

Conclusion

As semiconductor manufacturing continues to advance and devices shrink, the demands on defect analysis and review capabilities grow in parallel.

By providing definitive chemical identification of organic, inorganic, and complex mixed-material defects in a fully non-destructive manner, PiFM preserves wafer and device integrity for ongoing analysis while delivering sub-5 nm chemical imaging of defects that remain invisible to conventional review tools. This combination of nanoscale resolution, direct mapping from inspection tool’s defect coordinates, and support for 300 mm workflows yields actionable insights that accelerate yield learning, pinpoint root causes, and enable confident, data-driven process improvements. 

These capabilities of PiFM makes it a uniquely powerful tool for automated defect review in advanced semiconductor manufacturing.

References

1. “Particle Analysis by SEM/EDXS and Specimen Damage”, Mario Schmied and Peter Poelt, Mikrochimica Acta, 139, 171 – 177 (2002)