← Latest papers
⚛️ high-energy theory

On Thermodynamic Universalities and Optics of B-deformed Reissner--Nordström--AdS Black Holes

This paper employs machine learning techniques to analyze the thermodynamic Van der Waals behaviors and optical shadows of B-deformed Reissner–Nordström–AdS black holes, successfully training neural networks to detect critical phenomena and constraining stringy deformation parameters using Event Horizon Telescope observations of M87* and Sgr A*.

Original authors: Adil Belhaj, Maryem Jemri

Published 2026-09-23
📖 6 min read🧠 Deep dive

Original authors: Adil Belhaj, Maryem Jemri

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Deep in the cosmos, where gravity is so intense that not even light can escape, black holes serve as nature's most extreme laboratories. For decades, physicists have studied these invisible giants not just as objects of light and darkness, but as thermodynamic systems, much like the steam engines of the nineteenth century. By treating the space around a black hole as a fluid that can be compressed or expanded, scientists have discovered that these cosmic objects undergo phase transitions similar to water boiling or freezing. This connection between the microscopic rules of quantum mechanics and the massive scale of gravity has become a vital bridge in modern physics. Recently, the Event Horizon Telescope, a global network of radio dishes, has captured the first direct images of the shadows cast by two of the universe's most famous black holes, M87* and Sgr A*. These images provide a rare, tangible way to test theories about how gravity behaves under the most extreme conditions, allowing researchers to check if their mathematical models match the reality captured in the sky.

In a new study, researchers Adil Belhaj and Maryem Jemri have taken this investigation a step further by exploring a specific, theoretical version of a black hole that includes subtle corrections from string theory. They focused on a type of black hole known as Reissner–Nordström–AdS, which is charged and sits within a universe that has a specific type of curvature. However, they did not stop at the standard model. They introduced a "deformation" inspired by non-commutative geometry, a concept from string theory that suggests space-time itself might have a fuzzy, grainy structure at the smallest scales. In their model, this fuzziness is controlled by a parameter linked to an invisible field called the B-field. The team set out to see how this stringy modification changes the black hole's behavior, specifically looking at two things: how it behaves like a thermodynamic fluid and what its shadow looks like when observed from Earth.

To tackle these complex questions, the researchers turned to advanced computational tools, including machine learning, to sift through vast amounts of simulated data. First, they examined the thermodynamic properties of these deformed black holes. In the standard view, certain black holes behave like a Van der Waals fluid, a type of gas that can condense into a liquid under specific pressure and temperature conditions. The team calculated whether their stringy, deformed black holes exhibited this same behavior. They found that the answer depends entirely on the value of the stringy parameter and the electric charge of the black hole. For most values, the fluid-like behavior disappears, but they identified specific regions where the black hole does indeed mimic a Van der Waals fluid. To confirm this, they built a computer program using a fully connected neural network, a type of artificial intelligence designed to recognize patterns. They fed the machine nearly half a million simulated scenarios, teaching it to distinguish between black hole configurations that showed this fluid-like behavior and those that did not. The result was a highly accurate classifier that could instantly identify the correct thermodynamic behavior, proving that machine learning is a powerful tool for mapping the hidden rules of these cosmic objects.

The researchers then turned their attention to the optical side of the problem: the shadow. When a black hole sits in front of a bright background, it blocks the light, creating a dark silhouette known as a shadow. The size and shape of this shadow depend on the black hole's mass, its charge, and the curvature of the space around it. The team simulated how the shadow of their stringy, deformed black hole would appear to a distant observer. They discovered that the electric charge and the stringy parameter both influence the size of the shadow, but in opposite ways. Increasing the charge makes the shadow smaller, while increasing the stringy deformation makes it larger. Interestingly, the cosmological constant, which relates to the expansion of the universe, also enlarges the shadow. Despite these changes in size, the shadow remained perfectly circular because the researchers were studying non-rotating black holes. This circularity is a crucial detail, as it means the shape itself does not change, only the scale.

Finally, the team brought their theoretical models into contact with real-world data. They compared their simulated shadows against the actual observations of M87* and Sgr A* captured by the Event Horizon Telescope. By adjusting the parameters of their stringy model, they searched for the specific combination of values that would produce a shadow matching the size and shape seen in the telescope images. They found that the stringy parameter must be negative and fall within a very narrow range to be consistent with the observations. If the parameter were outside this range, the predicted shadow would not match the data. To make this process even more robust, they used their machine learning model again. This time, the neural network was trained to look at the parameters and predict whether a given configuration would produce a shadow that fits the telescope data. The AI successfully learned to separate the "allowed" configurations from the "forbidden" ones with over 99% accuracy.

The study concludes that while the theoretical space for these stringy black holes is vast, the universe is much more selective. Only a tiny fraction of the possible mathematical configurations are compatible with what we actually see in the sky. The research highlights that the stringy parameter acts as a strict control knob, determining whether a theoretical black hole can exist in our universe as we observe it. By combining high-speed numerical simulations with machine learning, the authors have provided a clear, data-driven way to constrain these exotic theories. Their work suggests that while the mathematics of string theory allows for many possibilities, the physical reality of black holes, as revealed by the Event Horizon Telescope, imposes tight limits on how these theories can manifest. This approach offers a promising path forward for testing other complex gravitational theories, using the universe's own images as the ultimate judge.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →