Dark Energy Didn’t Kick In Right Away express
Dark energy is usually seen as an innate property of the cosmos, at work from the very first moments. But a new model suggests it behaved like water on the verge of boiling: long unnoticed, then suddenly kicking in at a critical moment. This scenario is called 'late activation': around 5–7 billion years ago, a shift occurred, and the universe’s expansion began to accelerate.
Three landmarks helped test the idea. The cosmic microwave background, the cooled light of the Big Bang, preserved the imprint of the young universe. Explosions of supernovae, like lighthouses, allowed distances to be measured; it was these that led Adam Riess to discover the acceleration in the 1990s. The patterns of galaxies revealed the rate of universe expansion. By combining these data, scientists essentially captured the moment the cosmos 'boiled over.'
There was a surprise:
However, this doesn't resolve the main cosmological puzzle — the Hubble tension, i.e., the discrepancy in the expansion rate measured by different methods. The mystery endures, and cracking it will require even bolder ideas.
Original abstract
as written by the authors arXiv:2603.13137
We investigate a specific emergent dark energy scenario, known as critically emergent dark energy (CEDE), in which dark energy is effectively absent in the early Universe and becomes dynamically relevant only after a critical cosmic epoch through a phase transition. We constrain this model using recent cosmological observations, including cosmic microwave background (CMB) data from \emph{Planck} 2018, baryon acoustic oscillation (BAO) measurements from SDSS and DESI DR2, and two independent Type Ia supernova compilations, PantheonPlus and Union3. Our results show that within the CEDE framework a dark energy phase transition is not ruled out. In particular, CMB-only, CMB+SDSS, and CMB+DESI datasets provide evidence for a nonzero transition scale factor and, according to standard statistical indicators such as $$\Delta\chi^2$$ and Bayesian evidence, can favor CEDE over the $$\Lambda$$CDM model. At the same time, we find that CEDE does not fully resolve the Hubble constant tension. Overall, our analysis indicates that dark energy models featuring a phase transition remain a viable and phenomenologically interesting extension of the standard cosmological framework. Upcoming high-precision cosmological surveys will be essential to further assess whether such emergent dark energy scenarios represent a genuine departure from $$\Lambda$$CDM or an effective description of current data.
If dark energy had 'woken up' too early, just a couple of billion years after the Big Bang, the galaxies and stars we know simply wouldn’t have formed, the expansion would have scattered all matter.