Center for Research in Multiscale Science and Engineering

PhD Thesis Defense - Pol Benítez Colominas

Modeling Lattice-Mediated Optoelectronic Response in Anharmonic Semiconductors: A First-Principles and Machine Learning Study

When

Nov 12, 2026 from 10:00 AM to 01:00 PM (Europe/Madrid / UTC100)

Where

Sala Polivalent - Edifici A - Campus Diagonal-Besòs (UPC)

Add event to calendar

iCal

As part of the Doctoral Programme in Material Science and Engineernig, Pol Benítez Colominas will be defending his PhD thesis on Tuesday, 12th Nobember 2026, at 10:00 AM, at the Escola d’Enginyeria de Barcelona Est (EEBE)

  • Name of the author: Pol Benítez Colominas

  • Thesis title: Modeling Lattice-Mediated Optoelectronic Response in Anharmonic Semiconductors: A First-Principles and Machine Learning Study

  • Thesis director: Prof. Claudio Cazorla Silva

  • Thesis co-director: Edgardo Ademar Saucedo Silva

Solar panels need materials that absorb sunlight well and last a long time. However, finding the right materials has always been slow and difficult. This research shows how combining supercomputer simulations with artificial intelligence (AI) helps scientists discover and improve materials for future solar energy much faster.

Key Highlights & Findings:

  • Solving a Material Puzzle: Explained how a promising group of new solar materials stays stable when heated.

  • Heat Changes Light Absorption: Showed that natural heat vibrations inside the material help it absorb sunlight twice as efficiently as scientists originally thought.

  • Control with Electricity: Discovered that using a small electric field can fine-tune how the material works without needing heat.

  • Teaching AI Real Physics: Built AI tools trained on real atomic data, allowing the AI to learn how atoms stick together and predict how materials behave.

  • Superfast Simulations: Used AI to instantly test complex material mixes that used to take supercomputers way too long to calculate.

By understanding how tiny atoms move and using AI to speed up testing, this research creates a simple blueprint to invent cheaper, more efficient solar panels.