Article

Advancing XPS Workflows Through Data-Driven Automation

Surface Analysis Spotlight: XPS

	by David Valley Sr. Staff Scientist

As surface analysis laboratories continue to seek faster, more reliable, and more accessible characterization workflows, automation is playing an increasingly important role. A new suite of data-driven tools is being introduced to simplify and optimize X-ray Photoelectron Spectroscopy (XPS) analysis using machine learning and advanced analytical modeling.

These tools automate critical steps in sample setup and data acquisition while maintaining the flexibility needed for advanced XPS applications.

Smart Sample Setup with Machine Learning

The first set of tools focuses on automating platen setup. Three machine learning models have been developed to:

• Automatically identify platen types.
• Align platen images with instrument coordinates.
• Detect sample locations on the platen.

Together, these capabilities automatically qualify a platen when it is introduced into the instrument and generate potential analysis locations, reducing setup time and minimizing operator intervention.

Optimizing Data Acquisition for Better Results

The second innovation addresses an important challenge in XPS workflows: determining the optimal balance between survey scans and high-resolution region scans.

Traditional XPS analysis begins with the acquisition of a broad survey spectrum followed by narrower scans designed to improve signal-to-noise ratios and reveal chemical-state information. The new approach uses statistical analysis of survey scan data to predict signal quality and quantification uncertainty, enabling automated recommendations for subsequent acquisitions.

Rather than identifying a single "best" method, the system evaluates a range of Pareto-optimal acquisition conditions. Users can then select acquisition strategies that best align with their priorities, whether those are speed, precision, or analytical depth.

Enabling More Intelligent XPS Analysis

By combining machine learning, statistical optimization, and user-defined objectives, these data-driven tools represent a significant advancement in XPS workflow automation. They help reduce complexity for new users while preserving the control and flexibility required by experienced analysts.

As laboratory workflows become increasingly data-driven, innovations such as these are helping make surface analysis more efficient, consistent, and accessible, ultimately enabling researchers to focus more on insights and less on instrument configuration.  To learn more about efficiency and consistent workflow, please join us for Dr. David Valley’s presentation "Data-Driven Automation for X-Ray Photoelectron Spectroscopy" at AVS 72 on Thursday, Nov. 12, 2026.