The study, published in The European Physical Journal C by Adel Awad, Dmitry Chirkov, Alexey Golovnev and Alexey Toporensky, examines the Bianchi IX cosmology, the mathematically richest of the homogeneous but anisotropic universe models, whose spatial sections are three-dimensional spheres. The authors set out to explain, from first principles, a puzzling phenomenon that had previously been observed only in numerical simulations: the volume of such a universe does not monotonically decrease toward the singularity, but instead oscillates, repeatedly flipping between contraction and expansion. Their analysis reveals that this behavior is not a numerical artifact but a direct and inevitable consequence of the mathematics of phantom-dominated dynamics.

To appreciate the result, it helps to recall the standard BKL framework. In the vacuum Bianchi IX universe, the approach to the singularity can be described as a sequence of Bianchi I Kasner epochs, in which each scale factor grows or shrinks as a power of time, a ∝ t^p1, b ∝ t^p2, c ∝ t^p3, with the exponents satisfying p1 + p2 + p3 = 1 and p1² + p2² + p3² = 1. These constraints force at least one exponent to be negative but never allow any exponent to fall below −1/3. When one direction becomes dominant, curvature terms act like an exponential potential wall, and the universe bounces into a new Kasner epoch with new exponents. This discrete transformation, called the Mixmaster map, has been shown to be chaotic, and the BKL conjecture holds that this oscillatory regime is the generic behavior of spacetime near a singularity.

Matter, remarkably, barely matters in this regime. As the original BKL authors already understood, ordinary matter becomes negligible near the singularity, with one crucial exception: a stiff fluid with equation-of-state parameter w = 1, whose most natural representation is a massless scalar field. A standard massless scalar field actually kills the chaos entirely, replacing the oscillations with monotonic Kasner-like behavior. But the new work considers a massless scalar field with the wrong sign of its kinetic term, the so-called phantom field, whose energy density is negative. This single sign flip rewrites the rules of the game in profound ways.

The technical heart of the analysis lies in the constraint equation governing the Kasner exponents. With a phantom field, the constraint becomes p1² + p2² + p3² = 1 + q², where q is the conserved scalar momentum, a quantity protected by the shift symmetry of the massless field. Because the right-hand side is now larger than one, the exponents are no longer bounded. Two consequences follow immediately. First, two of the three Kasner indices can be negative simultaneously, meaning that two spatial dimensions grow while approaching the singularity in the past, a situation impossible in vacuum. Second, individual exponents can take arbitrarily large values, positive or negative, as q grows.

The possibility of two negative indices has a striking consequence for the Mixmaster map itself. In the standard picture, only one scale factor can trigger a bounce, so the exponents alone determine the next epoch. With two negative indices, either of the two corresponding dimensions can drive the bounce, and which one actually does depends on the actual sizes of the scale factors, information that is not encoded in the exponents at all. The authors show that when the two dimensions have very different sizes, the usual bounce formulas apply, but with the roles of the indices assigned according to which dimension dominates. When the two scale factors are nearly equal, the exponential potential walls disappear altogether, and the system undergoes a smooth, continuous evolution described by Bessel functions, an analog of the small oscillations near the Taub solution known from the vacuum case, but now accompanied by a reversal of the volume expansion rate.

The most dramatic finding concerns the volume of the universe. After each bounce, the rate of change of the volume in logarithmic time is proportional to 1 + 2p1, where p1 is the index of the bouncing dimension. In vacuum, p1 is always greater than −1/3, so this factor stays positive and the volume keeps shrinking toward the singularity. But with a phantom field, each bounce with a mildly negative index increases q², and once q² exceeds 3/8, the smallest index can drop below −1/2. At that point the denominator of the Mixmaster map changes sign, and the volume, followed backward in time, switches from contraction to expansion. The authors’ numerical integrations confirm this: after a finite sequence of bounces, the volume derivative flips sign, and the universe that was heading toward a singularity instead begins to re-expand.

This mechanism explains the volume oscillations observed in earlier numerical studies of phantom-dominated Bianchi IX cosmologies. The expansion phase cannot last forever, however. The Bianchi IX universe has the topology of a sphere, and its positive isotropic curvature falls off only as the inverse square of the mean scale factor, much more slowly than the anisotropies and the phantom energy, which both scale as the inverse sixth power. At large volumes, the curvature inevitably dominates and forces recollapse, a feature already known for vacuum Bianchi IX from the work of Lin and Wald. The result is a universe whose volume oscillates between finite bounds, bouncing off a curvature-driven maximum at large sizes and a phantom-driven minimum at small ones.

The oscillating volume has a deep implication for the BKL conjecture itself. Because volume maxima are unavoidable and the Mixmaster approximation breaks down near them, the map cannot be iterated infinitely. In the phantom case, the isotropic curvature term, though negligible near the singularity, always retains a finite contribution as the volume never approaches zero. The authors therefore conclude that the chaotic Mixmaster regime, rather than being a global attractor of the dynamics, becomes a transient phenomenon realized only with finite accuracy over a finite number of bounces. This challenges the expectation, still lacking a rigorous proof even in the vacuum case, that BKL oscillations describe the generic approach to a cosmological singularity.

The work also uncovers a genuinely new exact solution of the Bianchi I equations, possible only with phantom matter: an exponentially expanding or contracting universe of constant volume rate, in which the negative phantom energy exactly balances the anisotropic kinetic energy. Formally, such solutions correspond to infinite Kasner exponents, illuminating why the Mixmaster map can produce arbitrarily large indices. While a strict mathematical proof that phantom-filled Bianchi IX universes never hit a singularity remains to be constructed, the connection between the condition p1 < −1/2 and volume bounces makes that conjecture eminently reasonable. If phantom fields, which appear naturally in dark energy models and modified gravity theories such as Horndeski gravity, played a role in the early universe, the violent chaotic singularity foreseen by classical general relativity may be replaced by an eternal cosmic heartbeat, a universe that forever oscillates between expansion and collapse without ever reaching an end.

Subject of Research: Bianchi IX cosmological dynamics with a phantom scalar field and its effect on BKL oscillations near the singularity

Article Title: Bianchi IX dynamics with a phantom field

Article References: Awad, A., Chirkov, D., Golovnev, A., & Toporensky, A. (2026). Bianchi IX dynamics with a phantom field. The European Physical Journal C, 86(9), Article 1106. https://doi.org/10.1140/epjc/s10052-026-16386-5

Image Credits: AI Generated

DOI: 10.1140/epjc/s10052-026-16386-5

Keywords: Bianchi IX, BKL oscillations, phantom field, Mixmaster map, Kasner exponents, cosmological singularity, anisotropic cosmology, general relativity, scalar field, volume oscillations, dark energy, theoretical physics