The existence of dark matter is firmly established, yet its cosmic abundance Ω_c = 0.268 ± 0.012 [1] and the nature of dark energy (Ω_Λ ≈ 0.684) remain central open problems. String theory and M-theory require, for self-consistency, 10 or 11 spacetime dimensions, leaving N = 9 or N = 10 spatial dimensions after compactification [2][3]. Several works have explored dark matter candidates from extra dimensions, such as Kaluza-Klein dark matter [4][5] and mirror matter in brane-world scenarios [6][7]. However, a direct link between the number of extra dimensions and the observed cosmic fractions is still missing.

For completeness, the dark matter problem can also, in principle, be addressed within extended theories of gravity. As discussed by Corda [8], modifications of the gravitational sector, including f(R) and scalar-tensor theories, provide an alternative framework in which effects conventionally attributed to dark matter and dark energy may be accounted for by modified gravitational dynamics. The present conjecture instead attributes the dark matter contribution to matter on other branes coupled through higher-dimensional gravity; its combinatorial estimate does not exclude extended-gravity explanations.

In this study, we adopt a simplified statistical approach based on the possible orientations of 3-branes in a higher-dimensional bulk. If, during the quantum gravity era, 3-branes of different orientations emerge with nearly equal a priori probability, and if Standard Model fields are confined to our visible 3-brane, then matter on other 3-branes that share no spatial dimensions with the visible sector will interact only through higher-dimensional gravity. Their abundance becomes a purely combinatorial quantity, independent of detailed particle physics. We call this the “Combinatorial Brane Statistics” (CBS) hypothesis.

We will show that for N = 9 and N = 10, the combinatorial factor yields a dark matter fraction of 23.8% and 29.2%, respectively, which numerically bracket the Planck value. The threefold classification of 3-branes—fully visible, fully invisible, and partially overlapping—is mutually exclusive and exhaustive, automatically satisfying Ω_total = 1 and thus consistent with the observed spatial flatness of the Universe [1]. Notably, the statistical weight of partially overlapping branes (75.0% for N = 9, 70.0% for N = 10) is of the same order as Ω_Λ, suggesting a possible geometric connection. We explore a mechanism in which gravitational contraction on partially overlapping branes projects as accelerated expansion on the visible brane, and we distinguish between overlaps of one and two spatial dimensions. This distinction naturally leads to a two-component effective dark energy, with the two-dimensional overlap providing a transient early contribution and the one-dimensional overlap driving the late-time acceleration. The resulting dark energy equation of state w(z) can evolve, offering a testable prediction for upcoming surveys.

We assume a D = (1 + N)-dimensional bulk spacetime with one time dimension and N spatial dimensions. A 3-brane is a hypersurface spanned by three out of the N spatial coordinates:

Gauge interactions are confined to the brane, while gravity propagates in the full bulk. We assume that in the extremely early universe, quantum gravitational processes produce 3-branes of all possible orientations with approximately uniform probability. The total number of 3-branes is then C(N, 3). Fixing the visible brane along dimensions {1, 2, 3}, all other 3-branes fall into three mutually exclusive classes:

Fully visible brane: all three dimensions belong to {1, 2, 3}. There is exactly 1 such brane (our own).Partially overlapping branes: one or two dimensions belong to {1, 2, 3}, the rest lie in the extra dimensions {4, …, N}.Fully invisible branes: all three dimensions are taken from {4, …, N}. They share no spatial dimensions with the visible brane.

Matter on fully invisible branes has no gauge interaction with the visible sector, only higher-dimensional gravity. It therefore behaves as cold, collisionless dark matter. The number of such branes is C(N − 3, 3). If each 3-brane carries, on average, roughly the same initial matter energy density, and if no substantial energy exchange occurs between branes, the fraction of dark matter in the total cosmic energy budget is

For the cases relevant to string/M-theory:

N = 9: C(9, 3) = 84, C(6, 3) = 20 → f_DM(9) = 20/84 ≈ 0.238.N = 10: C(10, 3) = 120, C(7, 3) = 35 → f_DM(10) = 35/120 ≈ 0.292.

The Planck 2018 measurement gives Ω_c ≈ 0.268. These two values bracket it, with relative deviations of ~11% and ~9%, without any adjustable parameters.

The three classes are exhaustive and mutually exclusive, so their statistical weights sum to exactly 1:

Table 1 lists the numbers and weights for N = 9 and N = 10. The completeness automatically yields Ω_total = 1, consistent with the observed flatness Ω_k ≈ 0.0007 ± 0.0019 [1].

Remarkably, the weight of partially overlapping branes (75.0% for N = 9, 70.0% for N = 10) is of the same order as Ω_Λ ≈ 68.4%. This numerical proximity suggests that partially overlapping branes might be the geometrical origin of dark energy.

Table 1.Combinatorial classification and statistical weights for N = 9 and N = 10.

In general relativity, tidal forces demonstrate that compression in one direction can be accompanied by expansion in orthogonal directions. We propose that matter on partially overlapping branes undergoes gravitational contraction; through the shared dimensions, this contraction is geometrically projected onto the visible brane. Along the shared directions, the projection may manifest as an attractive acceleration, while along the non-shared directions it appears as an effective expansion of space. Thus, the late-time accelerated expansion on the visible brane would not originate from a true cosmological constant but from the inter-brane geometric relationship (see Figure 1).

Figure 1. Schematic illustration of 3-brane interactions in 1 overlapping dimension. Each enclosed region represents a 3-brane. Shared edges indicate dimensions common to adjacent branes. The central region corresponds to the visible brane, surrounded by partially overlapping branes.

To illustrate the mechanism, consider a 5-dimensional bulk (N = 4) with coordinates (t, x, y, z, w). The visible Brane A is at w = 0, extended along {x, y, z}. A partially overlapping Brane B is at z = 0, extended along {x, y, w}, sharing two dimensions {x, y}. Assume a Kasner-type metric ansatz:

If gravitational contraction on Brane B causes a₄ to contract, the vacuum Einstein equations admit solutions where a₃ expands exponentially while a₄ contracts, preserving the total volume element. The visible brane thus experiences an effective Hubble parameter H_eff along its exclusive direction z, sourced entirely by the ordinary matter density ρ_(B) on Brane B:

This toy model demonstrates the core idea: contraction in overlapping extra dimensions can project as expansion in the visible non-shared dimensions. Generalization to N = 9 or 10 and inclusion of moduli stabilization remain for future work.

The partially overlapping class splits further into branes that share one spatial dimension with the visible brane and those that share two. ForN = 10, for example, there are 63 one-dimension-sharing branes and 21 two-dimension-sharing branes. Their geometric coupling to the visible brane is expected to differ in strength and distance dependence. Let the coupling strengths be κ 1 ( d ) ∝ d − α 1 and κ 2 ( d ) ∝ d − α 2 , where d ( a ) ∝ a β denotes the typical scale of the extra dimensions as a function of the cosmic scale factor a, and α₂ < α₁ is expected. The matter density on all branes redshifts as ρ ( B ) ∝ a − 3 . The effective dark energy density induced on the visible brane from the two subclasses is then

where A₁, A₂ are proportional to the respective brane numbers and their average matter density. The total effective dark energy density is

The effective equation-of-state parameter follows from the continuity equation:

Because ρ DE ( 2 ) has a larger initial coupling but decays more rapidly (α₂ < α₁), it dominates at early times and is a natural candidate for Early Dark Energy (EDE), potentially easing the Hubble tension. ρ DE ( 1 ) becomes dominant at late times, driving the present-day acceleration. Depending on the sign of β and the α-values, w(z) can exhibit either thawing (w > −1) or phantom (w < −1) evolution. Recent DESI 2024 results, which hint at w₀ ≈ −0.85 in a w₀w_a CDM parametrization, are qualitatively compatible with a thawing behavior from the one-dimension-sharing component with a small positive α₁β.

The observed macroscopic space is three-dimensional. In string theory, D3-branes are stable objects supporting four-dimensional gauge theories. The current model faces two primary theoretical challenges: the lack of a dynamical origin for the equal a priori probability hypothesis, and the absence of a full quantitative derivation of the dark energy geometric mechanism for the physically relevant cases N = 9, 10. Nevertheless, the framework offers several testable signatures.

Most importantly, the two-component dark energy derived predicts an evolving equation of state w(z). The one-dimension-overlap component produces a slowly evolving late-time w, while the two-dimension-overlap component yields a transient early dark energy signature. Upcoming surveys such as DESI, Euclid, and LSST will map w(z) with high precision. A detection of w(z) deviating from −1, especially with a past EDE-like bump, would strongly motivate a geometric dark energy picture of the type proposed here. Conversely, a persistent w = −1 at all redshifts would rule out the simplest versions of this scenario.

Traces of EDE could also be imprinted in the CMB and the measurement of H₀. A full confrontation with cosmological data, combined with a derivation of the coupling parameters from a fundamental compactification, is left for future work.

Starting from the statistical distribution of 3-brane orientations in higher dimensions, we have provided a phenomenological framework that addresses the cosmic dark matter abundance and the dark energy fraction simultaneously. The dark matter fraction is purely combinatorial, f_DM = C(N − 3, 3)/C(N, 3), giving 23.8% (N = 9) and 29.2% (N = 10) with no free parameters. The mutually exclusive and exhaustive classification naturally enforces a flat Universe. The statistical weight of partially overlapping branes is of the same order as Ω_Λ, and we explored a geometric interpretation where gravitational contraction on these branes projects as dark energy. Distinguishing between overlaps of one and two spatial dimensions leads to a two-component effective dark energy that can produce a transient early dark energy signal and a thawing late-time equation of state. While a complete dynamical foundation is still missing, the numerical coincidences and the testable predictions for w(z) make this framework a worthwhile stimulus for further research on the connection between extra dimensions and cosmological parameters.