Cosmologists Propose Cosmic Expansion Flipped From Deceleration 9.5 Billion Years Ago

For decades, cosmologists have treated dark energy—the mysterious force accelerating the universe's expansion—as an unchanging constant. But a new theoretical model suggests this cosmic engine may have undergone a dramatic reversal. According to an international team of physicists, dark energy might have actually been negative in the deep past, acting as an attractive force before flipping to positive roughly 9.5 billion years ago.[1][4]

The findings, published in the journal Physical Review D, challenge the standard cosmological model, known as Lambda-CDM. In the standard view, dark energy maintains a constant, positive density throughout cosmic history, driving a steady acceleration. However, recent observations from the Dark Energy Spectroscopic Instrument (DESI)—which is mapping millions of galaxies in three dimensions—hinted in early 2025 that dark energy's influence might be evolving over time.[1][3][5]

To find a physical mechanism for this potential evolution, a team led by Professor Özgür Akarsu at Istanbul Technical University turned to an alternative to Albert Einstein's general relativity. The framework, called "teleparallel f(T) gravity," approaches the geometry of the universe differently. While Einstein's theory describes gravity through the curvature of spacetime, teleparallel gravity explains it through a mathematical property called torsion, or twisting.[1][3]

The researchers deliberately explored a previously dismissed branch of this torsional theory. Earlier studies of teleparallel gravity assumed that dark energy's density had to remain strictly positive, mirroring the assumptions of general relativity. Akarsu's team relaxed that constraint to see what the mathematics would allow. "The sign change emerged naturally from the theory itself," Akarsu noted, explaining that the model did not require adding any new, independently adjustable parameters.[1][3]

A Cosmic Reversal

In this newly analyzed branch of the theory, the effective density of dark energy is negative during the universe's first few billion years. This negative energy would have added to the gravitational pull of ordinary matter and dark matter. As a result, the universe's expansion would have decelerated much more sharply than standard models predict during its early epochs.[1][4][5]

As the universe expanded and its matter diluted, the model dictates that this dark energy density reached zero at a redshift of approximately 1.5. In cosmological terms, this transition point corresponds to roughly 9.5 billion years ago, when the universe was just over 4 billion years old. From that moment forward, the density became positive, triggering the era of accelerated expansion that continues today.[1][3][4]

The sign-switching behavior emerges elegantly without introducing new variables to the standard cosmological model. The value governing the branch depends entirely on the universe's present-day matter density. This means the model operates with the same six cosmological parameters as the standard Lambda-CDM framework, making it a highly economical solution rather than a complex mathematical patch.[1][4]

Despite its unusual behavior on cosmic scales, the paper demonstrates that this branch remains consistent with the precise tests of gravity carried out in the solar system. In dense environments like our local planetary neighborhood, the theory's deviations from Einstein's general relativity are naturally suppressed. This screening mechanism allows the model to pass established observational checks while rewriting the universe's largest scales.[5]

Resolving the Hubble Tension

Beyond matching the evolving dark energy hints from DESI, the sign-switching model offers a potential solution to one of cosmology's most stubborn problems: the Hubble tension. This tension arises because measurements of the universe's current expansion rate differ significantly depending on the observational method used, creating a crisis in modern physics.[1][4]

When astronomers measure the expansion locally using pulsating stars and supernovae, they find a faster rate of about 73 kilometers per second per megaparsec. However, when they calculate the rate based on the ancient cosmic microwave background—the fading afterglow of the Big Bang—they get a slower rate of about 67 kilometers per second per megaparsec. The standard model cannot reconcile these two highly precise measurements.[1]

By allowing dark energy to change signs, the new teleparallel model can accommodate the higher expansion rates measured locally while remaining perfectly consistent with the ancient light left over from the Big Bang. The team found that the same present-day matter density can support this dynamic history, bridging the gap between the early and late universe.[1][4]

The research team brought together expertise from multiple institutions to develop this framework. The work was carried out in Akarsu's research group at Istanbul Technical University with postdoctoral researcher Dr. N. Merve Uzun and doctoral student Bilal Bulduk. They collaborated with Professor Nihan Katırcı of Doğuş University and Professor Antonio De Felice of Kyoto University's Yukawa Institute for Theoretical Physics.[1][3][5]

Mapping the Cosmic Web

The Dark Energy Spectroscopic Instrument, which prompted this theoretical reevaluation, represents a massive leap in observational cosmology. Installed on the Mayall Telescope at Kitt Peak National Observatory in Arizona, DESI uses 5,000 robotic positioners to capture the light from thousands of galaxies simultaneously. This allows it to build a three-dimensional map of the universe with unprecedented precision.[3]

DESI measures the expansion history of the universe by tracking baryon acoustic oscillations—subtle ripples in the distribution of galaxies that serve as a standard cosmic ruler. By observing how the apparent size of these ripples changes at different distances, astronomers can reconstruct how the expansion rate has varied over the past 11 billion years.[3]

When the DESI collaboration released its first three years of data in early 2025, the results sent shockwaves through the physics community. The data suggested that dark energy might be weakening over time, rather than remaining a strict cosmological constant. This unexpected hint of dynamic behavior provided the exact observational opening that the sign-switching teleparallel model required.[3][4]

AI

The Istanbul-led team had actually submitted their initial research on the sign-changing dark energy model in October 2024, several months before the DESI announcement. The subsequent release of the observational data provided a timely and independent validation of the theoretical path they had chosen to explore.[3]

The Next Observational Tests

Akarsu and his collaborators had already been developing a phenomenological scenario known as the sign-switching Lambda-CDM model for several years. Their previous analyses showed that a cosmological constant switching from negative to positive could ease several long-standing disagreements. This new torsional gravity model finally offers a rigorous physical basis for that earlier scenario.[1][4]

The ultimate test for this gravity model will come from future observational data. While it currently aligns with DESI's initial three-year findings, the collaboration is expected to release its full five-year dataset in 2027. That release will map over 47 million galaxies and quasars, providing the statistical power needed to definitively test whether dark energy is truly a cosmic shapeshifter.[1][3]

Comprehensive testing via structure growth, redshift-space distortions, and weak gravitational lensing will determine whether the alternative model holds up against empirical data. Neither the proposed dark energy reversal nor a recent cosmic slowdown has been established as an observational fact, but the theoretical groundwork is now firmly in place for astronomers to test.[3][5]

"We identified an alternative path for the Universe within an existing theory," Akarsu explained regarding the team's findings. "The next step is to test it rigorously against the full observational data. Initial signs look promising, and we remain particularly curious about potential implications for the Hubble tension."[3]

If the upcoming DESI data confirms that dark energy is indeed evolving, the teleparallel gravity framework could become a leading candidate to replace the standard model. For now, the research serves as a reminder that even established theories can hide unexplored possibilities when long-held assumptions are deliberately relaxed.[1][4]