Researchers at Istanbul Technical University have identified a previously overlooked mathematical solution in teleparallel gravity theory that suggests dark energy may have shifted from negative to positive approximately 9.5 billion years ago.

Prof. Özgür Akarsu led the international team, which included postdoctoral researcher Dr. N. Merve Uzun and doctoral student Bilal Bulduk from ITU, along with Prof. Nihan Katırcı of Doğuş University and Prof. Antonio De Felice of Kyoto University's Yukawa Institute for Theoretical Physics. The study appeared in Physical Review D in 2025.

The team reexamined teleparallel f(T) gravity, an alternative framework to Einstein's general relativity that describes gravity through torsion rather than spacetime curvature. While previous researchers had focused on one mathematical branch of exponential teleparallel gravity, Akarsu's group investigated what happens when the assumption that dark energy density must always be positive is removed.

Their calculations revealed two possible cosmic histories. One resembles the standard model of cosmology. The other predicts that effective dark energy had negative density in the distant past, reaching zero at redshift 1.5 (roughly 9.5 billion years ago) before becoming positive. The model requires no additional parameters beyond the six used in the standard Lambda-CDM model.

Akarsu stated: "For several years we have been studying a scenario in which dark energy switches sign, and it has held up remarkably well against data such as DESI's. The open question was what physical mechanism could produce such behavior. So we deliberately looked at a branch of a well-known theory of gravity that had been set aside because it violates the usual expectation that the dark energy density stays positive."

Observations from the Dark Energy Spectroscopic Instrument have raised questions about whether dark energy behaves as a fixed constant. When combined with measurements of the cosmic microwave background, supernovae, and weak gravitational lensing, DESI's findings strengthen indications that dark energy might change over time. However, the evidence remains statistically insufficient to claim a definitive discovery.

DESI has completed its five-year survey of more than 47 million galaxies and quasars. The researchers have not yet tested their model against the complete DESI dataset.