Response Surface Methodology (RSM) is an advanced DOE technique that models the relationship between process inputs (factors) and outputs (responses) as a mathematical surface, enabling optimization — finding the factor settings that maximize, minimize, or target a specific response value.

Why RSM?

• Factorial designs (2-level) identify which factors are important and provide linear models — but real processes rarely have purely linear responses.
• RSM uses 3+ levels per factor to fit quadratic (second-order) models that capture curvature, minima, maxima, and saddle points in the response landscape.
• Once the response surface is modeled, the optimal operating point can be found mathematically.

The RSM Model

A second-order RSM model for $k$ factors:

This includes:

• Linear terms ($\beta_i x_i$): Main effects.
• Quadratic terms ($\beta_{ii} x_i^2$): Curvature — how the response bends.
• Interaction terms ($\beta_{ij} x_i x_j$): How factors interact.

RSM Workflow

• Screening (Phase 1): Use fractional factorial to identify the 2–5 most important factors.
• Path of Steepest Ascent (Phase 2): Use a first-order model to move quickly toward the optimum region.
• Response Surface Design (Phase 3): Run a central composite design (CCD) or Box-Behnken design near the optimum to fit a second-order model.
• Optimization (Phase 4): Use the fitted model to find the exact optimal factor settings.

Common RSM Designs

• Central Composite Design (CCD): The most popular RSM design. Consists of factorial points + axial (star) points + center points. Provides enough data to fit the full second-order model.
• Box-Behnken Design: An alternative to CCD that avoids extreme corner points — useful when extreme conditions are impractical or risky.

Semiconductor Applications

• Etch Process Optimization: Model CD, uniformity, and selectivity as functions of power, pressure, and gas flow — find settings that simultaneously optimize all three.
• CMP Optimization: Model removal rate, uniformity, and dishing as functions of pressure, speed, and slurry flow.
• Lithography Process Window: Map CD and DOF as functions of dose, focus, and PEB temperature to find the most robust operating point.

Multi-Response Optimization

• In semiconductor manufacturing, multiple responses must often be optimized simultaneously (e.g., maximize etch rate while minimizing roughness and maintaining selectivity).
• Desirability Function: Combines multiple responses into a single optimization target, weighting each response by importance.

RSM is the standard method for process optimization in semiconductor manufacturing — it transforms experimental data into actionable mathematical models that identify the best operating conditions.