Periodic and Statistically Distributed Surface Texturing of Surface‐Enhanced Raman Spectroscopy (SERS) Sensors with Spatially Selective Signal Mapping

ABSTRACT Surface‐enhanced Raman spectroscopy (SERS) is a highly sensitive and noninvasive molecular fingerprinting technique already making an impact in fields such as medical diagnosis and environmental monitoring. However, challenges persist in the fabrication of SERS substrates, which delays the adoption of SERS as a standard analytical tool. Here, we experimentally demonstrate and compare two methods for achieving reproducible SERS substrates on a 150 mm wafer scale. In the first approach, we use nanoimprint lithography (NIL) to create highly periodic structures with subwavelength dimensions (< 300 nm). These periodic structures offer high signal uniformity, with a spatial variation of just 7.7% over a 20 × 20 µm area. In the second approach, we use a maskless reactive ion etching (RIE) process to create statistically distributed spikes in silicon wafers. Compared to the periodic structures, the statistical structures offer five times greater signal enhancement and twice the signal‐to‐background ratio at the cost of increased spatial variation. Both substrate types perform similarly or better than commercially available SERS substrates. In addition, we demonstrate spatially selective Raman mapping via microarray templating by photolithography, with potential for multimodal detection of analytes or use in quality control.