A Second Silver Layer Turns a Silent Melamine Sensor Into a 190,000× SERS Signal
Two SPR peaks instead of one: adding a second silver deposition step after laser dewetting turns an undetectable melamine signal into a 190,000× SERS enhancement.
Melamine detection keeps coming back as a food-safety problem because the compound is cheap, nitrogen-rich, and easy to add to milk and other protein-tested products to fake a higher protein reading — the practice behind China’s 2008 tainted-milk scandal. Surface-enhanced Raman scattering (SERS) is one of the faster, simpler screening routes available: no derivatization, no chromatography column, just a molecular fingerprint read directly off a nanostructured metal surface. But that fingerprint is only as strong as the nanostructure underneath it, and laser-dewetted silver films — a popular, low-cost way to build that nanostructure — don’t automatically deliver a detectable signal.
The Obstacle — One Dewetting Pass Builds Particles, Not Hotspots
Laser-induced dewetting is an efficient way to turn a flat, continuous silver film into a field of nanoparticles: a focused laser locally melts the film, and surface tension pulls it into isolated islands as it cools. Push the laser power from 15 to 25 W and the resulting particles shrink from about 162 to 125 nm. The problem is that a single dewetting pass produces one particle population at a time, and larger particles — while optically active — tend to sit too far apart to create the tight nanogaps that generate real SERS hotspots. In this study, that gap showed up directly in the data: aqueous melamine solution placed on any of the traditionally dewetted films produced no obvious Raman peak at all.
The Innovation — Deposit a Second, Thinner Silver Layer After Dewetting
In a study published in Optics & Laser Technology (Lin, Ding, Yen, Chen & Lee, 2025, DOI: 10.1016/j.optlastec.2024.111673), researchers at National Pingtung University of Science and Technology and National Kaohsiung Normal University tested a hybrid dewetting process: a continuous-wave 1070 nm fiber laser (R4 HS Series, SPI) first dewets a 10 nm silver film into large nanoparticles as usual, then a thermal evaporator deposits a second, 2 nm-thick silver layer directly onto that dewetted surface.
The second silver deposition fills the gaps between the large dewetted nanoparticles with a population of much smaller ones, creating the mixed-size nanostructure at the heart of the hybrid process.
That second layer doesn’t just add mass — it forms its own population of roughly 20 nm nanoparticles, sitting in the gaps between the original, larger 142–167 nm particles. Instead of one particle size, the hybrid film carries two, and the smaller particles are small and close enough to generate the tight electromagnetic hotspots the larger particles alone couldn’t.
Key Results — From No Signal to a 190,000× Enhancement Factor
The team built the case in three steps: show that the hybrid process succeeds where traditional dewetting fails, establish how low a melamine concentration it can still detect, then confirm the signal is reproducible enough to trust.
Only the Hybrid Film Detected Melamine’s 701 cm-1 Fingerprint Peak
For 10-3 M melamine solution at all three tested laser powers, the traditionally dewetted films (a–c) showed no obvious fingerprint peak, while every hybrid dewetted film (d–f) produced a clear signal at 701 cm-1. At the best-performing condition — 25 W laser power, 100 mm/s scanning speed — the hybrid film reached an analytical enhancement factor (AEF) of 1.9×105.
Detection Held Down to 10-7 M — Five Times Below the Legal Limit
Across a melamine concentration series from 10-3 down to 10-7 M, the 701 cm-1 fingerprint peak weakened but stayed resolvable at every step. That lower bound — 10-7 M, or 0.0126 ppm — sits roughly 200 times below the 2.5 ppm melamine safety limit set by the US and EU.
10 Random Points, 4.37% Relative Standard Deviation
Measuring 10-3 M melamine at 10 randomly chosen points on the 25 W / 100 mm/s hybrid substrate, the 701 cm-1 peak intensity varied by a relative standard deviation of just 4.37% — the kind of point-to-point consistency that separates a usable screening substrate from a one-off lab result.
Note: This work was supported by Taiwan’s National Science and Technology Council, Project No. NSTC 112-2637-E-020-004.
Why ProTrusTech Is Featuring This Research
Every Raman measurement in this study — the traditional-versus-hybrid comparison, the concentration series, and the 10-point reproducibility test — was collected on a confocal Micro Raman system (MRI532S) at 532 nm excitation. This is the same lab’s third consecutive melamine-SERS publication on our site to use this Raman platform: the HfO2-shelled nanoparticle work we featured previously (also from H.K. Lin’s group at NPUST) used the identical instrument to verify a different route to the same 701 cm-1 fingerprint peak.
Two independent fabrication strategies — a protective ceramic shell in one paper, a second metal deposition step in this one — both converging on the same Raman platform to validate melamine detection is exactly the kind of repeat, cross-project use that tells us an instrument is trusted by a research group, not used once and set aside.
Engineering the Solution — A Second Thin-Film Step, Not a New Process
The manufacturability story here is that the upgrade from a weak substrate to a highly sensitive one costs almost nothing extra in process complexity. Both dewetting and thermal silver deposition are standard thin-film steps already available on the same evaporator system used for the original film — no wet chemistry, no new equipment, no exotic precursor. Tuning laser power and scanning speed gives direct control over the large-particle population, while the second deposition’s thickness controls the small-particle population that fills the gaps. That’s a two-parameter process a fabrication line can dial in, not a bespoke recipe that only works once.
Featured Instruments:
Why This Matters
Beyond melamine, this result is a clean demonstration of a general SERS design principle: hotspot density depends on having small nanogaps between particles, and a single dewetting step tends to produce particles that are either too large and too far apart, or too small and too weakly resonant, to deliver both at once. Deliberately building a mixed-size nanostructure — large particles for a strong plasmonic response, small particles to bridge the gaps between them — is a route other SERS substrate designs can borrow, independent of the specific analyte or the specific metal.
What’s Next
- Combining hybrid dewetting with a protective shell layer, pairing this paper’s hotspot-density gains with the stability improvements from ProTrusTech’s earlier-featured HfO2-shelled substrate.
- DDA-guided optimization of particle spacing, using the simulation framework already validated here to target an even higher enhancement factor.
- Extending the two-step deposition approach to other small-molecule food and environmental contaminants beyond melamine.
Evaluating a SERS-based screening substrate for your own food safety or contaminant detection workflow? Get in touch with our team.
Key Takeaways
- A hybrid dewetting process — laser dewetting followed by a second, 2 nm silver deposition — builds a mixed-size nanoparticle population (~20 nm and 142–167 nm) that single-step dewetting cannot.
- Traditionally dewetted films showed no detectable melamine Raman signal; hybrid dewetted films showed a strong 701 cm-1 fingerprint peak at every tested laser power.
- The best condition (25 W, 100 mm/s) reached an analytical enhancement factor of 1.9×105 and a limit of detection of 10-7 M (0.0126 ppm) — roughly 200x below the 2.5 ppm US/EU melamine safety limit.
- Point-to-point reproducibility across 10 random measurement locations came in at 4.37% relative standard deviation.
- All Raman measurements were collected on a confocal Micro Raman system (MRI532S) at 532 nm — the same platform behind ProTrusTech’s earlier feature on HfO2-shelled melamine SERS substrates from the same lab.
