Anticounterfeiting tags based on randomly oriented MoSx clusters enabled by capillary and Marangoni flow

The proliferation of counterfeit products poses a significant threat to global economies and public health. Robust and secure anticounterfeiting technologies are crucial to combat this issue. Current approaches often rely on chemical methods, offering high encoding capacity and randomness. This research explores a new approach utilizing the unique properties of randomly oriented MoSx/MoOx clusters as anticounterfeiting tags. The unpredictable nature of these clusters, generated through controlled manipulation of capillary and Marangoni flows, makes them exceptionally difficult to replicate. This approach moves beyond simple visual inspection, relying instead on precise, quantitative analysis of the cluster topography for authentication. The use of advanced microscopy techniques, along with rigorous statistical testing, ensures the security and reliability of the generated tags. The combination of material science and digital techniques holds significant potential for enhancing the security of products and supply chains.

Existing anticounterfeiting techniques often employ chemical methods to create unpredictable patterns with high encoding capacity. Examples include biomimetic microfingerprints, biological physically unclonable functions (PUFs), chaotic organic crystal phosphorescent patterns, and inkjet-printed quantum dot fluorescent labels. While these methods show promise, they may still face limitations in terms of cost, complexity, or ease of replication. The use of 2D materials like MoS2 for PUFs has also been explored, but often requires expensive characterization tools. This study aims to overcome these limitations by introducing a novel approach that combines the advantages of functional materials with efficient and cost-effective reading methods.

The fabrication process involves several key steps: 1. **Precursor Solution Preparation:** An ammonium heptamolybdate solution is prepared and filtered to remove any undissolved particles. 2. **Spin Coating:** The precursor solution is spin-coated onto a cleaned Si/SiO2 substrate. 3. **Thermal Baking:** The coated substrate is placed on a hotplate to create porous MoO3 clusters through thermal decomposition. The morphology of the clusters is precisely controlled by adjusting the concentration of 2-methoxyethanol (2-ME) in the precursor solution. This influences the interplay between capillary and Marangoni flows, determining the final arrangement of MoO3 clusters. The 2-ME concentration is optimized to achieve a random distribution of clusters, avoiding ring or dot-like patterns. 4. **Sulfurization (Optional):** For enhanced functionality, the MoO3 clusters are sulfurized using H2S gas in a CVD chamber to form MoS2 clusters. 5. **PDMS Coating (Optional):** A protective PDMS layer is applied to prevent duplication through molding techniques. 6. **Confocal Laser Microscopy:** A high-speed confocal laser microscope is employed to acquire a 3D height profile of the MoSx/MoOx clusters. This process captures the intricate, random topography of the clusters efficiently. 7. **Image Processing and Key Generation:** A custom Python code processes the microscopy data. The height profile is converted to a grayscale image, denoised using the Non-Local Means (NLM) algorithm, and the local peaks are identified. These peaks are then binarized and resized to a consistent size using a binning process. The von Neumann debiasing technique addresses potential biases in the bit distribution, resulting in a balanced bit uniformity closer to 0.5. Finally, a 128-bit digital key is generated for each tag. 8. **Tag Characterization:** The generated keys are statistically analyzed. Metrics such as entropy, bit uniformity, reproducibility (intra-device Hamming Distance, HD), uniqueness (inter-device HD), degree of freedom, and false negative/positive rates are calculated to assess the security and reliability of the tags. NIST statistical tests confirm the randomness of the extracted keys.

The study successfully demonstrates the fabrication of anticounterfeiting tags based on randomly oriented MoSx clusters. The key findings include: 1. **Controlled Randomness:** The precise manipulation of capillary and Marangoni flows allows for the creation of randomly distributed MoOx/MoSx clusters with unpredictable morphologies. 2. **Efficient Tag Reading:** High-speed confocal laser microscopy enables rapid and accurate acquisition of 3D topography data, making the reading process fast and efficient. 3. **Robust Key Generation:** A robust digitization process, incorporating denoising, peak detection, binarization, and von Neumann debiasing, produces high-quality 128-bit keys. 4. **High Security Metrics:** The generated keys exhibit high entropy (close to 1), balanced bit uniformity (close to 0.5), and excellent reproducibility and uniqueness. The calculated Hamming distances between intra-device and inter-device keys are significantly different, indicating minimal chances of false positives or negatives. The calculated false positive and negative rates were on the order of 10^-12. 5. **NIST Test Validation:** The generated keys successfully pass the NIST statistical tests, confirming their randomness and suitability for cryptographic applications. 6. **Protection Against Cloning:** The optional PDMS coating prevents duplication via molding techniques. 7. **High Encoding Capacity:** The study indicates a high encoding capacity (2^138), which makes the tags very difficult to replicate.

The results demonstrate the feasibility of using randomly oriented MoSx clusters as robust and secure anticounterfeiting tags. The combination of controlled material synthesis, advanced imaging techniques, and sophisticated data processing leads to tags with high security metrics and resilience against counterfeiting attempts. The method addresses the limitations of existing techniques, offering a cost-effective and efficient alternative. The ability to generate highly random and statistically validated keys holds significant promise for various security applications, especially in product authentication and brand protection.

This study successfully demonstrates the fabrication of anticounterfeiting tags based on randomly oriented MoSx clusters generated using capillary and Marangoni flows. The combination of controlled morphology evolution, high-speed confocal laser microscopy, and robust key generation methods provides a highly secure and efficient anticounterfeiting system. The excellent performance metrics and NIST test validation confirm the reliability of this approach. Future work could explore the integration of other functional materials, further enhancing the security and functionalities of these tags, and the use of machine learning algorithms for improved key extraction and authentication.

While the study demonstrates a high level of security, there are a few limitations to consider. The current approach relies on specialized equipment such as confocal laser microscopy, which might not be widely accessible. The complexity of the image processing and key generation pipeline could pose challenges for implementation in large-scale applications. Future studies could focus on developing more streamlined algorithms and exploring more readily available characterization techniques to increase the accessibility and scalability of this technology.