Calibration of gas flow, vacuum, temperature, RF and exposure

Why calibration drifts on older tools

A process recipe is only as good as the numbers the tool reports. On a tool that has run for twenty years, a mass flow controller can read a few percent off its true flow, a capacitance manometer can drift by several percent, and a thermocouple can age until the furnace runs hotter than its display says. Drift of this kind rarely triggers an alarm. It shows up as a slow change in etch rate, film thickness or line width, and as recipes that no longer transfer between tools.

Regular calibration of process equipment brings the reported values back in line with reality, and the as-found record tells you whether past results may have been affected.

What we calibrate

The scope depends on the tool. Typical items in scope are:

• Mass flow controllers: zero and span, checked with a flow reference or by rate of rise in a known volume, depending on the gas and the set-up.
• Vacuum gauges: capacitance manometers, Pirani and ion gauges compared against a reference gauge at several pressures.
• Temperature: thermocouples and controllers on furnaces, hot plates, chucks and heated lines, checked with a reference probe or profile thermocouple.
• RF power: forward and reflected power readings compared with a calibrated RF power meter into a dummy load.
• Exposure intensity: UV intensity and field uniformity on mask aligners, measured with a calibrated UV intensity meter at the wavelengths the tool uses.
• Chiller and flow sensors: setpoint against measured temperature and flow.

For preventive maintenance customers we usually combine calibration with a scheduled visit, which keeps downtime to one window.

Tolerances and what gets adjusted

For most items the tolerance is the original specification; where your process needs a tighter limit, we use yours, and the record states which applies. Items found within tolerance are left alone, because adjusting a working instrument adds risk without benefit. Items outside tolerance are adjusted only where the hardware allows it. Some cannot be brought back, such as a mass flow controller with a contaminated sensor or a gauge head at the end of its life; those are flagged and quoted for replacement or rebuild rather than forced into range.

Downtime and what you provide

A calibration visit on a single tool typically takes from a few hours to a day, depending on how many items are in scope and whether the tool needs to reach process temperature. Furnace profiling takes longer because each zone has to stabilise.

We need:

• The tool free of users and safe to work on, with process gases available where flow checks need them
• The recipe tolerances that matter to your process, if they are tighter than the original specification
• Previous calibration records, so we can track drift over time

Results and documentation

You receive a calibration record per tool listing each item, its tolerance, the as-found and as-left values, the reference instrument used and its own calibration date. Items that were found out of tolerance are flagged with a note on which processes they may have affected, so you can judge whether earlier results need a second look. For mask aligners the record includes the intensity map; on platforms such as the SUSS MA6 this is also the starting point if you are considering a UV-LED source.