Gap-enhanced Raman tags for physically unclonable anticounterfeiting labels

Counterfeiting poses a significant global challenge, causing economic losses and threatening public health. While various anticounterfeiting methods exist, including holograms, watermarks, and nanostructured surface labels with surface-enhanced Raman scattering (SERS) signatures, these are often susceptible to forgery due to their deterministic nature. Physical unclonable functions (PUFs), based on inherent randomness in physical objects, offer a more secure approach. PUF labels leverage stochastic processes to create unique, difficult-to-duplicate patterns. The encoding capacity, or the maximum number of unique labels, is crucial for PUF security. Authentication methods for PUF labels typically involve point-by-point comparison of digitized patterns or pattern recognition. Chemical methods, such as using luminescent nanomaterials or molecular tags, offer advantages in terms of stochasticity, encoding capacity, and cost-effectiveness. Optical PUF labels, particularly those based on SERS, are attractive due to their non-contact, fast readout, and potential for high encoding capacity using multiple spectral features. SERS offers advantages over fluorescence due to narrower spectral linewidths and higher material and photostability. This study introduces a novel SERS-based PUF label using GERTs, aiming to achieve high encoding capacity and fast, reliable authentication.

The authors review existing anticounterfeiting techniques, highlighting the limitations of deterministic methods like holograms and conventional SERS-based labels. They discuss the advantages of PUF-based systems, emphasizing the importance of encoding capacity and low false-positive rates. The use of chemical methods for PUF label fabrication, particularly those employing optical techniques such as fluorescence or Raman scattering, is examined. The superior properties of SERS tags over fluorescence dyes for PUF applications—such as higher encoding capacity, better photostability, and narrower spectral linewidths—are discussed. The authors cite previous work on various types of metallic nanoparticles used in SERS-based anticounterfeiting, along with different methods for encoding information. Existing work on optical PUF labels employing excitation-selected lanthanide luminescence and plasmonic nanogels is also referenced.

The researchers synthesized ten types of GERTs, each with a different thiolated aromatic molecule as a Raman reporter. These GERTs were characterized using TEM and SERS spectroscopy to confirm their core-shell structure, size, and unique spectral profiles. PUF labels were fabricated by drop-casting aqueous GERT solutions onto silica substrates. Raman mapping using a confocal Raman system (785 nm laser) with varying resolutions (2×2, 10×10, and 50×50 pixels) was performed to acquire Raman signals. The readout signals were digitized using coarse-grained coding methods. For labels composed of a single type of GERT, the Raman intensity at a specific wavenumber was used for digitization. For labels with multiple GERT types, a non-negative least squares (NNLS) method was employed to demultiplex the Raman spectra and determine the contribution of each GERT type at each pixel. Both binary and quaternary encoding of Raman intensity levels were used. A similarity index (I) was calculated by comparing digitized patterns pixel-by-pixel, quantifying the reproducibility of the same label and the distinguishability of different labels. A global search (GS) algorithm was employed for threshold determination. High-speed readout using a DuoScan mode with a SWIFT mode was also demonstrated. Finally, the practical application of the PUF labels was tested by fabricating them on transparent Scotch tape, which was then affixed to printing paper. The PUF labels on Scotch tape were read out using a Raman spectrometer to confirm feasibility.

The study successfully synthesized and characterized ten types of GERTs with distinct Raman spectral profiles. The GERT-based PUF labels exhibited a high degree of randomness due to the stochastic nature of drop-casting. Using a 50×50 pixel resolution and quaternary encoding, a theoretical encoding capacity exceeding 3 × 10¹⁵⁰⁵¹ was achieved. The authentication experiments demonstrated high reproducibility (similarity index I > 85% for binary encoding, > 70% for quaternary encoding) for the same label under repeated measurements, and clear differentiation between different labels (I < 57% for binary encoding, < 30% for quaternary encoding). High-speed readout using the DuoScan/SWIFT mode reduced the acquisition time for a 50×50 pixel Raman map to 6 seconds, showing significant improvement compared to the conventional STAGE mode. The study successfully demonstrated the feasibility of applying the GERT-based PUF labels to transparent Scotch tape, suitable for transferring onto various product surfaces.

The results demonstrate the effectiveness of GERTs as a robust platform for creating highly secure PUF labels. The exceptionally high encoding capacity achieved ensures a practically unbreakable anticounterfeiting system. The ability to perform high-speed readout significantly enhances the practicality of the method, making it potentially suitable for real-world applications. The use of Scotch tape as a substrate further broadens the potential applications. The high reproducibility and distinguishability observed confirm the reliability and security of the authentication process. Further optimization and improvements, including the refinement of demultiplexing algorithms and improved laser stability, could lead to even better performance. The potential vulnerability of data substitution during cloud-based authentication can be mitigated by employing advanced encryption techniques.

This research successfully developed gap-enhanced Raman tags (GERTs) based PUF labels for high-security anticounterfeiting applications. The extremely high encoding capacity, combined with high-speed readout and robust authentication, makes this technology a promising solution for combating counterfeiting. Future work could focus on optimizing the demultiplexing algorithms, enhancing the photostability of the tags, and developing more compact and user-friendly readout devices for practical implementation.

While the theoretical encoding capacity is exceptionally high, the actual achievable capacity may be limited by practical factors such as variations in the manufacturing process and the sensitivity of the Raman system. The current readout speed, although significantly improved, still needs to be further optimized for high-throughput industrial applications. The authentication relies on cloud-based comparison, making it vulnerable to potential data breaches or attacks. Finally, while the Scotch tape method shows promise, further research is necessary to determine the long-term stability and environmental robustness of these PUF labels.