Machine learning for Random Laser characterization and phase retrieval

• Grant number: 23/08033-7
• Support Opportunities: Regular Research Grants
• Start date: February 01, 2024
• End date: January 31, 2026
• Field of knowledge: Physical Sciences and Mathematics - Physics - Condensed Matter Physics
• Agreement: National Research Council of Italy (CNR)
• Principal Investigator: Davinson Mariano da Silva
• Grantee: Davinson Mariano da Silva
• Principal researcher abroad: Neda Ghofraniha
• Institution abroad: Consiglio Nazionale delle Ricerche (CNR), Italy
• Host Institution: Faculdade de Tecnologia São Paulo (FATEC São Paulo). Centro Paula Souza (CEETEPS). Secretaria de Desenvolvimento Econômico (São Paulo - Estado). São Paulo , SP, Brazil
• Associated researchers: Marco Isaías Alayo Chávez

Abstract

The development of photonics would not have been possible without suitable materials for transmitting and manipulating light. The development of lasers was one of the important milestones for the development of this important field, which today is seen as strategic for increasing the competitiveness of various industrial sectors. The lasers manufacture involves obtaining gain media with a very low concentration of structural defects. This is because structural defects are deleterious and can greatly compromise the performance of laser devices. Unlike conventional lasers, random laser (RL) sources do not have a predefined resonant cavity. In these types of lasers, the multiple scattering of light, which is avoided in conventional lasers, is precisely the phenomenon that generates the necessary feedback for optical amplification. For this reason, random lasers may have less complex manufacturing processes, and allow the use of a greater range of materials as possible means of gain. RLs are emitted in many directions and have very low temporal and spatial coherence, which is advantageous, for example, for obtaining high-definition images, free of speckle. However, the same properties that make RLs attractive are also responsible for the difficulty in characterizing them. In this sense, phase retrieval combined with replica symmetry breaking (RSB) theory may constitute a new way to investigate RL emission, since the non-linear equation of phase retrieval appears in many different scenarios, from X-ray imaging to optical computing. Furthermore, many algorithms have been proposed, from flexible gradient descent routes to more specialized spectral methods and recent advances in machine learning have contributed a lot with these investigations. Thus, the proposed project aims to develop new mechanisms for the characterization and improvement of RL emitting systems through machine learning algorithms associated with phase retrieval. This will allow not only the demonstration of laser action on materials, but also help in the optimization and development of new materials and photonic devices, which can be used in the generation of random lasers used in the most diverse areas. All works proposed in this project are unpublished and aim to generate knowledge about new photonic devices, with different possibilities of application in the field of photonics. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

CAMARA, JOSIVANIR G.; DA SILVA, DAVINSON M.; FERRETTI, STEFANO; GENTILINI, SILVIA; CONTI, CLAUDIO; GHOFRANIHA, NEDA. Random Laser and Replica Symmetry Breaking in SiO₂-Rhodamine 6G xerogel powder. CNR '09: SECOND INTERNATIONAL WORKSHOP ON COMPOUND NUCLEAR REACTIONS AND RELATED TOPICS, v. 309, p. 2-pg., 2024-01-01. (23/08033-7)

CAMARA, JOSIVANIR G.; DA SILVA, DAVINSON M.; KASSAB, LUCIANA R. P.; PALACIOS, G.; DE ARAUJO, CID B.. Upconversion random lasing in Er3+/Yb3+co-doped germanate-lead-magnesium glass-ceramics. Optical Materials, v. 157, p. 9-pg., 2024-11-01. (23/08033-7)