Evolving Dark Energy Is Vacuum Energy After All

Abstract

We investigate a physically motivated model of dynamical dark energy arising from the non-perturbative topological structure of the Quantum Chromodynamics (QCD) vacuum. The model introduces no new fundamental field or propagating degree of freedom: the dark energy (DE) density emerges as a global vacuum response to an expanding spacetime. We develop the first comprehensive cosmological implementation of this QCD-DE scenario and confront it with current observations, including Planck, ACT and SPT-3G cosmic microwave background data, DESI DR2 baryon acoustic oscillation measurements, and Type Ia supernova samples from Pantheon+ and DES-Dovekie. We compare the model with $\Lambda\mathrm{CDM}$ and $w_0w_a\mathrm{CDM}$ cosmologies. The model provides an excellent fit to the data and reproduces the late-time DE evolution preferred by DESI. The model naturally predicts effective phantom crossing behaviour at intermediate redshifts ($z\sim0.67$) while avoiding the instabilities associated with phantom scalar fields. Using goodness-of-fit statistics and Bayesian model-selection tools, including Akaike and Deviance Information Criteria and Bayesian evidence estimated from Markov-Chain Monte Carlo chains, we find that the QCD-induced model is consistently favoured over $\Lambda\mathrm{CDM}$ for the full combination of early and late-time datasets. Unlike the conventional descriptions of dynamical DE, support for QCD-DE in Bayesian evidence remains more consistent across datasets, suggesting that a physically motivated departure from a cosmological constant may provide a more economical description of the expansion history preferred by current observations.