Random laser ablated tags for anticounterfeiting purposes and towards physically unclonable functions

Counterfeiting poses a significant global challenge, causing economic losses, public health risks, and national security concerns. Attaching anticounterfeiting tags to products is a practical solution, but existing methods often suffer from limitations such as high cost, complex fabrication processes, susceptibility to cloning, and scalability issues. Current high-capacity tags, like spectral and graphically encoded tags, are deterministic and predictable, making them vulnerable to counterfeiting. Therefore, non-deterministic encoding mechanisms with strong authentication are crucial. While various physical and chemical methods have been explored using 1D, 2D, and 3D materials, they often present challenges related to speed of fabrication, cost, complexity, and mass production. Luminescent materials and chemical analysis methods, while offering high encoding capacity, lack the speed and cost-effectiveness required for widespread implementation. This research investigates the use of pulsed lasers as a solution to overcome these limitations, leveraging their ability to create random ablation patterns quickly and efficiently on laser-sensitive materials. The focus is on developing a method for generating physically unclonable functions (PUFs) for robust anticounterfeiting.

Existing anticounterfeiting tag technologies have been extensively studied, encompassing a wide range of materials and fabrication methods. These include 1D, 2D, and 3D materials/structures utilizing physical and chemical approaches. Chemical methods, attractive for their low cost and stochastic fabrication, often involve complex processes and lack speed. The use of luminescent materials like quantum dots and lanthanide complexes provides easy detection, but limitations include slow fabrication, high cost, and environmental instability. Mass-spectrometric and elemental analysis techniques offer high data capacity, yet struggle with rapid fabrication and scalability. Laser-based techniques have found applications in material surface treatment (sintering, nanowelding, patterning, annealing), but their application in generating random patterns for security tagging has received less attention. Speckle pattern recognition using lasers has been explored for authentication, but lacks the high data capacity and cost-effectiveness sought for widespread anticounterfeiting.

This study employs a nanosecond pulsed infrared (IR) laser (1064 nm) to ablate laser-sensitive materials (Si or Si/SiO2) and create random crater patterns. The key to generating randomness lies in carefully balancing laser parameters (power, repetition rate, pulse duration) and scanning line parameters (scan speed, hatch distance (HD), rotation angle (RA)). The laser and scanning line parameters were optimized to create randomly distributed craters, unlike regular patterns produced with slower scan speeds. A 4-inch wafer was used to demonstrate the scalability of the method. Image processing techniques were used to extract the outline circles of the craters, creating a binary representation (0.5 bit uniformity) of the pattern without any debiasing algorithm. MATLAB software was used to analyze the randomness of the generated patterns. Root Mean Square Error (RMSE) was calculated to assess the uniformity of crater distribution. The uniformity of the orientation of objects formed by overlapping craters, and the direction of trend lines connecting multiple craters, were evaluated using RMSE. A challenge-response pair generation mechanism was developed where a 64-bit challenge sequence was overlaid on the tag image, and the response was generated from the distribution of trend lines within defined areas. The randomness of the responses was analyzed using statistical NIST tests (specifically seven tests: Frequency, Block frequency, Run, Longest run, Serial, Approximate entropy, and Cumulative sums). The hamming distance between different responses was used to quantify the uniqueness and reproducibility of the tags. Both single-tag and multi-tag scenarios were evaluated. The application of the generated bit sequences as cryptographic keys for data encryption and decryption was also tested. Finally, stability tests were performed under harsh conditions to evaluate the robustness of the tags.

The laser ablation technique successfully generated highly random crater patterns on Si/SiO2 substrates. Optimization of laser and scanning line parameters (specifically a hatch distance of 0.3 mm and a rotation angle of 13°) resulted in highly uniform crater distributions with a low RMSE value. The challenge-response pair generation method produced responses with a fixed bit uniformity of 0.5. NIST statistical tests confirmed the true randomness of the generated bit sequences. Single tags achieved an encoding capacity of approximately 10³⁹ with false positive and negative rates on the order of 10⁻⁵⁸. Multiple tags exhibited an even higher encoding capacity (approximately 10⁵¹) and extremely low false rates (around 10⁻⁵⁰). The generated bit sequences were successfully used for data encryption and decryption, demonstrating their potential in cryptography. The tags also demonstrated good stability under harsh environmental conditions. The possibility of physical replication was addressed by proposing encapsulation using PDMS and leveraging photoluminescence (PL) imaging of the tags as a further layer of security against cloning.

This work successfully addresses the need for a rapid, cost-effective, and secure anticounterfeiting technology. The laser ablation method demonstrated significantly improved characteristics compared to existing techniques, particularly in terms of speed and scalability. The high encoding capacity, low false rates, and successful NIST test results validate the randomness and security of the generated tags. The integration of the generated bit sequences into a challenge-response system, and their application in data encryption and decryption, expands the potential applications beyond simple identification to secure authentication and cryptography. The use of photoluminescence as an additional security measure further strengthens the unclonability of the tags. The results strongly suggest the potential of this technology for a wide range of applications, especially in industries requiring high-volume, low-cost, and secure product authentication.

This research presents a novel laser ablation technique for generating highly secure, random anticounterfeiting tags with exceptional encoding capacity and low false rates. The method is rapid, cost-effective, and scalable, addressing major limitations of existing technologies. The successful application of the generated bit sequences in data encryption and decryption further demonstrates the versatility and security of the proposed method. Future research could explore the application of this technique to a broader range of materials and the development of more sophisticated challenge-response protocols to enhance the security level even further.

While the study demonstrates high randomness and security, limitations include the dependence on laser-sensitive materials and the potential for physical replication, although this was addressed by proposing encapsulation and the use of photoluminescence as an additional security measure. Further research is needed to evaluate the long-term stability of the tags under various environmental conditions and to explore more robust techniques to further prevent cloning.