Direct measurement of electronic band structure (energy + momentum) at the interface, as the material stack is being grown — using condensed-matter physics tools (STM, ARPES) integrated with industrial deposition equipment (PVD, PLD, MBE).
Why it matters
The semiconductor industry’s metrology stack measures atoms (TEM, XRD, atomic-scale lithography characterisation) and averaged electrical behaviour (I-V, R, C on finished devices). It does not measure electrons directly.
Resistance, mobility, current — these are averaged proxies for what electrons actually do at material interfaces. As nodes shrink and interface area grows as a fraction of device volume, this proxy gap is now the dominant source of device-level performance variance.
If you can measure electrons directly during growth, you close the loop:
- Faster discovery: months instead of years for a new contact material
- Cheaper: coupon-scale, no lithography, no full device fab
- Higher ceiling: approach the ballistic-transport theoretical limit, not just incremental atomic-scale tweaks
The instruments
- STM (Scanning Tunneling Microscopy) — scanning tunneling microscopy. Tip-on-surface imaging of individual atoms and their local electronic state. Slow but direct.
- ARPES (Angular-Resolved Photoemission Spectroscopy) — angular-resolved photoemission spectroscopy. Shine UV/X-ray light on a surface, photoelectrons fly out; measure their energy + emission angle to reconstruct the electronic band structure E(k). Fast, integrative.
What’s novel
- In-situ integration with growth. STM+ARPES measure samples in the same vacuum-continuous platform where PVD/PLD/MBE deposits them. No breaking vacuum, no transferring between tools.
- Industrial application. Historically these tools have been used on exotic correlated materials (high-Tc superconductors, manganites, multiferroics) where you write Nature/Science papers. Industrial silicon-stack discovery is unconventional.
- Volume. 75-target PVD + multi-modality (PLD, MBE) means combinatorial materials sweep at coupon scale.
Frontier questions
Comparable approaches (incumbent)
- DFT-based materials screening at IMEC + every foundry’s internal materials team. Computational, not experimental.
- Combinatorial materials discovery (cf. Intermolecular — sold to Merck KGaA for ~$60M after IPO, cautionary tale).
- In-line metrology inside foundries — at-wafer X-ray photoelectron spectroscopy (XPS), but only on finished films, not during growth.