Precision in metrology is the closeness of agreement between repeated measurements of the same quantity under the same conditions — measuring how consistently a semiconductor metrology tool reproduces the same result, independent of whether that result is accurate (close to the true value).

What Is Precision?

• Definition: The degree of agreement among independent measurements made under stipulated conditions — quantified as the standard deviation or range of repeated measurements.
• Distinction: Precision measures repeatability and consistency; accuracy measures closeness to truth. High precision means low scatter; high accuracy means centered on the true value.
• Expression: Reported as standard deviation (σ), coefficient of variation (CV%), or range of repeated measurements.

Why Precision Matters

• SPC Effectiveness: Statistical process control requires precise measurements — if measurement scatter is large, control charts cannot distinguish real process shifts from measurement noise.
• Process Capability: Measurement imprecision inflates apparent process variation, making Cpk values appear lower than the true process capability.
• Tight Tolerances: At advanced semiconductor nodes, tolerances are sub-nanometer — measurement precision must be a small fraction of the tolerance to make reliable decisions.
• Gauge R&R: Precision is the repeatability component of Gauge R&R — the largest contributor to measurement system variation in automated semiconductor metrology.

Types of Precision

• Repeatability: Variation when the same operator measures the same feature on the same tool in rapid succession — short-term precision.
• Reproducibility: Variation when different operators, tools, or conditions measure the same feature — long-term, cross-condition precision.
• Intermediate Precision: Variation within a single lab over time — includes day-to-day, setup-to-setup, and environmental variations.
• Reproducibility (Inter-Lab): Variation between different laboratories measuring the same sample — critical for supplier-customer measurement agreement.

Precision Requirements in Semiconductor Metrology

Improving Precision

• Averaging: Multiple measurements averaged reduce random variation by √n — 9 measurements reduce noise by 3x.
• Environmental Control: Temperature stability, vibration isolation, and EMI shielding minimize environmental noise.
• Tool Maintenance: Clean optics, fresh calibration, and proper tool condition maintain optimal precision.
• Sample Preparation: Consistent sample positioning, cleaning, and orientation reduce setup-related variation.

Precision is the foundation of reliable process control in semiconductor manufacturing — without precise measurements, even the most sophisticated SPC systems and process control algorithms cannot distinguish real process changes from measurement noise.