Equipment

Metrology equipment

The measurement tools that quantify what the process actually produced — film thickness, feature size, overlay, and more — so the process can be controlled.

In short

Metrology equipment measures the wafer: how thick the films are, how big the features are, how well layers align, and the surface and electrical properties. These measurements are what let engineers keep every process on target.

How does the process-control loop work?

Why it matters

You cannot control what you cannot measure. Metrology turns a process from guesswork into a controlled loop: it quantifies the result of each step so deviations are caught and corrected before they cost yield.

Beginner intuition

Every process step has some variation. Metrology is how the fab 'sees' that variation — by measuring the real wafer instead of assuming the step worked.

Different measurements answer different questions: How thick is this film? How wide is this line? Did this layer land on top of the last one?

Where it fits in manufacturing

Throughout the flow, after key steps — often on sample wafers or sample sites — feeding the data that keeps every other tool in spec.

How it works

What goes in

A processed wafer, or sample sites on it
A measurement recipe defining what to measure and where

What happens inside

Measure: an appropriate technique quantifies the property of interest — for example an optical method for film thickness, or a specialized tool for feature size and overlay.
Compare: the measured value is compared against the target and the control limits.
Feed back: deviations are used to adjust the responsible process tool before more wafers drift.

What comes out

Quantitative data — thickness, critical dimension, overlay, and more — that drives process-control decisions.

Major subsystems

Measurement sensor / optics

The core technique that quantifies the property.

Precision stage

Positions the wafer to measure the right sites accurately.

Recipe & data system

Defines what and where to measure, and records results for control.

Process parameters that matter

Accuracy & precision: How correct, and how repeatable, the measurement is — both matter for control.
Throughput vs sampling: How many wafers and sites are measured; more sampling catches more, but costs time.
Sensitivity: The smallest change the tool can reliably detect.

A note on numbers

Achievable accuracy, precision, and throughput are technique- and vendor-dependent — treat any figures elsewhere as examples, not universal specs.

What determines performance

The measurements that matter most include: film thickness (are deposited or grown layers on target?), critical dimension or CD (are features the right size?), overlay (did this layer align to the previous one?), surface properties (roughness and topography), electrical properties (for example sheet resistance as a proxy for doping and activation), and defectivity (covered under inspection).

A key trade-off is measurement quality versus speed and sampling. Measuring every wafer everywhere would be ideal for control but is impractical, so fabs sample intelligently and rely on precise, repeatable tools.

This is the heart of process control: run the process, measure the result, compare it against target, adjust the tool, and run again. Equipment provides both the sensing and the knobs; process control closes the loop; and the tighter and faster that loop runs, the higher the yield.

Common issues

What can go wrong:

A wrong or drifting measurement is dangerous — it can hide a real process problem or trigger a false correction
Poor sampling → problems missed between measured wafers
Low precision → real drifts lost in measurement noise

Yield implications

Metrology is a yield multiplier: fast, accurate measurement shortens the learning cycle and catches drifts early, while blind spots let systematic errors reach many wafers before they are discovered.