Gravitational Thermodynamics: How Free Energy Minimization Explains Dark Matter Phenomena

Abstract We unify gravity and thermodynamics: Newton’s law gains a density-dependent multiplier from scalar field free energy minimization: F(r, ρ) = GM1M2 r2[1 + 2β2e−meff (ρ)r]. Here, meff (ρ) emerges from Veff (ψ)=V(ψ) + ρln A(ψ). No dark matter particles needed. Solar System: |γPPN −1|<10−21 (beats Cassini by 16 orders). Galaxies: SPARC rotation curves fit with one parameter (η≈2). Clusters: Bullet offset via gas-galaxy differential screening. Cosmology: γ= 3(n+ 2)/[2(n+ 1)] ∈[1.5,2.25] predicts H0relief and early structure. Falsifiable by Euclid lensing, CMB-S4, and cluster offset scaling. This thermodynamic gravity replaces dark matter with a first-principles mechanism. 1 Gravitational Thermodynamics: How Free Energy Minimization Explains Dark Matter Phenomena Edwin W. Maina Department of Materials Science and Engineering Portland State University, Portland, OR 97201, USA [email protected] November 7, 2025 1 Introduction Newton’s law of gravitation, F∝GM1M2/r2, has stood for over three centuries as one of physics’ most successful theories. Yet gravitational phenomena on galactic and cosmological scales appear to require either vast amounts of invisible matter [1, 2] or modifications to gravity itself [3, 4]. After four decades of intensive searches, no dark matter particle has been detected [5, 6], motivating alternative explanations [41]. We propose that incorporating thermodynamic principles into gravity naturally resolves this puzzle. Specifically, when scalar fields couple to matter, the coupled system minimizes its free energy, producing environment-dependent screening that modifies gravitational strength. This is not an ad hoc modification but an inevitable consequence of the second law of thermodynamics. 1.1 The Fundamental Equation The entire framework can be expressed in a single modified force law: F(r, ρ) = GM1M2 r2h1+2β2e−meff (ρ)ri(1) This equation synthesizes two foundational principles: First term: (GM1M2/r2) is Newton’s inverse-square law (1687), encoding three centuries of gravitational physics tested from laboratory scales to the Solar System. Second term: [1+2β2e−meff (ρ)r] is a thermodynamic enhancement factor arising from densitydependent screening. The effective mass meff (ρ) emerges from minimizing the free energy Veff(ψ, ρ)=V(ψ) + ρln A(ψ) (2) of a scalar field ψcoupled to matter density ρthrough a conformal factor A(ψ). The multiplicative structure ensures that: •In high-density regions (meff r≫1): the exponential vanishes, recovering F=GM1M2/r2 exactly (Solar System compliance). •In low-density regions (meffr≪1): the exponential approaches unity, yielding F≈(GM1M2/r2)(1+ 2β2) (enhanced gravity for galaxies and clusters). 2 •At intermediate densities: smooth interpolation governed by thermodynamic equilibrium. Thus Eq. (1) is not Newton’s law “plus corrections” but rather Newton’s law multiplied by the thermodynamic response of the gravitational medium. This is analogous to how electromagnetic fields propagate in material media:  E= E0/ϵr, where the dielectric constant ϵremerges from thermodynamic properties of the medium. 1.2 Why Thermodynamics? Three key insights motivate this approach: 1. Coupling implies screening. When any field couples to matter, the matter responds to minimize total energy. This is ubiquitous in physics: Debye screening in plasmas, dielectric screening in electromagnetism, Yukawa screening in nuclear physics. Gravity should be no exception. 2. Free energy determines equilibrium. At finite temperature (or in systems with many degrees of freedom), equilibrium is set by minimizing free energy F=U−TS, not just potential energy. For scalar-matter coupling, this free energy has the form Eq. (2) [34]. 3. Environment dependence is inevitable. The equilibrium field configuration ψmin that minimizes Veff depends on local density ρ. Therefore meff (ρ) varies across environments, producing density-dependent gravity. The remainder of this paper demonstrates that this thermodynamic framework: 1. Arises naturally from scalar-tensor theories via partition function integration (Section II) 2. Passes Solar System tests with unprecedented precision (Section III) 3. Explains galaxy rotation curves without dark matter (Section IV) 4. Reproduces cluster collision dynamics like the Bullet Cluster (Section V) 5. Predicts cosmological evolution through a derived parameter γ(Section VI) 6. Makes falsifiable predictions distinguishable from particle dark matter (Section VII) 2 Thermodynamic Framework 2.1 Action and Coupling We begin with a scalar-tensor action: S=Zd4x√−gM2 PlR 2−(∂ψ)2 2−V(ψ)+Sm[A2(ψ)gµν,Ψm] (3) where ψis a scalar field, V(ψ) its potential, and Smthe matter action. The key feature is the conformal coupling: matter fields Ψmexperience an effective metric ˜gµν =A2(ψ)gµν, where A(ψ)≃1 + βψ MPl (4) with βa dimensionless coupling strength (typically β∼ O(1)) [30, 32]. This conformal coupling means matter particles experience modified gravitational strength. From the matter perspective, the gravitational constant is effectively Geff =G×A2(ψ). Thus the field ψmediates an additional force beyond standard gravity. 3 2.2 Free Energy from Partition Function At finite temperature (or when integrating over quantum matter fluctuations), we should use the partition function rather than the action. The partition function for matter in a background field ψis: Z[ψ, ρ] = ZDΨmexp −Sm[A2(ψ)g, Ψm](5) The free energy is F[ψ;ρ] = −Tln Z[ψ, ρ]. For a system with conserved matter density ρ, integrating out matter degrees of freedom gives: F[ψ;ρ] = Zd3x(∇ψ)2 2+V(ψ) + ρln A(ψ)(6) The crucial term is ρln A(ψ), which represents the interaction free energy between the field and matter. Using Eq. (4): ρln A(ψ)≈ρ·2βψ MPl =2βρ MPl ψ(7) This is the thermodynamic coupling term that drives environment-dependent screening. 2.3 Effective Potential and Equilibrium The effective potential per unit volume is: Veff(ψ, ρ)=V(ψ) + ρln A(ψ) (8) Equilibrium occurs at the minimum of Veff: ∂Veff ∂ψ ψmin = 0 (9) For a chameleon-type potential V(ψ)=Λ4+n/ψnand using Eq. (7), this gives: ψmin(ρ) = nΛ4+nMPl 2βρ 1/(n+1) (10) This is the key result: the equilibrium field value depends on density. In high-density regions, ψmin is small; in low-density regions, it is large. 2.4 Effective Mass and Screening Length The effective mass of fluctuations around equilibrium is: m2 eff(ρ) = ∂2Veff ∂ψ2ψmin(ρ) (11) For the chameleon potential, this yields: meff(ρ)∝ρ(n+2)/[2(n+1)] (12) The screening length is λscr = 1/meff. Key limits: 4 •High density (e.g., Solar System): meff ∼1013 eV, λscr ∼10−8m (atomic scale). Field is heavily screened. •Low density (e.g., galaxy halos): meff ∼10−27 eV, λscr ∼1022 m (kpc scale). Field is active over galactic distances. •Span: meff varies by ∼40 orders of magnitude across cosmic environments. This enormous range is not fine-tuned but emerges automatically from thermodynamic equilibrium via Eq. (12) [31]. 2.5 The Force Law Around a spherically symmetric source of mass M1, the field equation is: ∇2ψ−m2 eff(ρ)ψ=−2βM1 MPl δ3(r) (13) The solution is a Yukawa potential: ψ(r) = −2βM1 4πMPlre−meff r(14) A test mass M2at distance rexperiences force:  F=−M2∇GM1 r+A(ψ)GM1 r(15) Expanding A(ψ)≈1 + βψ/MPl and using Eq. (14): F(r) = GM1M2 r21+2β2e−meff r(16) This recovers Eq. (1). The multiplicative enhancement factor [1 + 2β2e−meff r] is the thermodynamic response. 2.6 Physical Interpretation Equation (16) has a clear physical meaning: Dense environments: High ρgives large meff, so e−meff r≈0 at any macroscopic distance. The force reduces to pure Newton: F=GM1M2/r2. Thermodynamics screens out the scalar contribution. Diffuse environments: Low ρgives small meff, so e−meff r≈1 over large scales. The force is enhanced: F≈(GM1M2/r2)(1 + 2β2). Thermodynamics activates the scalar field. Intermediate: At densities where meffr∼1, the exponential produces smooth interpolation between regimes. This is directly analogous to phase transitions in thermodynamics. Just as water undergoes solid-liquid-gas transitions with temperature, gravity undergoes screened-intermediate-active transitions with density. The “phase diagram” is ψmin(ρ) from Eq. (10). 5 3 Solar System Tests 3.1 Thin-Shell Screening In the Solar System, matter densities are extremely high by cosmological standards. From Eq. (12), we expect meff ≫1/R⊙, producing strong screening [30]. For an extended body of radius Rand mass M, the field ψinside the body adjusts to minimize Veff at the local density. Near the surface, where density drops rapidly, a thin shell forms where the field transitions from the interior value ψin to the exterior value ψout. The effective source of the external field is not the full mass Mbut only the mass in the thin shell: Meff =M×∆R R(17) where ∆R≪Ris the shell thickness. This dramatically suppresses fifth-force effects [33]. 3.2 Parametrized Post-Newtonian Formalism Solar System tests typically constrain deviations from general relativity using the Parametrized Post-Newtonian (PPN) formalism [7]. The key parameter is γPPN, defined such that GR gives γPPN = 1. For our theory, the effective γPPN is: γPPN = 1 + 2β2Meff M⊙ (18) Using Eq. (17): |γPPN −1|= 2β2∆R R(19) 3.3 Solar System Calculations For the Sun with ρ⊙∼1.4 g/cm3, using n= 2 chameleon potential and β∼1: Effective mass inside Sun: meff,⊙∼1013 eV ⇒λscr ∼10−8m (20) Shell thickness: ∆R⊙∼1 meff,⊙∼10−8m (21) Suppression factor: ∆R⊙ R⊙∼10−8m 7×108m∼1.4×10−17 (22) PPN parameter: |γPPN −1|∼2(1)2×1.4×10−17 = 2.8×10−17 (23) The Cassini spacecraft constrained |γPPN −1|<2.3×10−5[8]. Our prediction is: |γPPN −1|predicted <10−16 ≪2.3×10−5(24) We beat Cassini by 16 orders of magnitude. For Earth (ρ⊕∼5.5 g/cm3) and Moon (ρMoon ∼3.3 g/cm3), suppression is even stronger. 6 Lunar Laser Ranging (LLR) tests constrain |η|<4×10−4for anomalous equivalence principle violations [9]. Our framework predicts: |η|predicted ∼10−24 (25) This is 20 orders of magnitude below current constraints, essentially perfect GR compliance. 3.4 Why This Works The key insight is that thin-shell screening is not a fine-tuning but an inevitable consequence of thermodynamics. In high-density regions: 1. Free energy Veff is dominated by ρln Aterm 2. Minimization drives ψ→0 3. Effective mass meff becomes huge 4. Screening length λscr = 1/meff becomes microscopic 5. External observers see only the thin surface shell This is the same mechanism by which plasma frequency screening protects electromagnetic field propagation in dense media. It is simply thermodynamics at work. 4 Galaxy Rotation Curves 4.1 Rotation Curve Problem Galaxy rotation curves - the orbital velocities v(r) of stars and gas as a function of radius rhave been the primary evidence for dark matter since the 1970s [2, 10]. Newtonian gravity predicts v(r)∝r−1/2in the outer regions where visible matter density drops. Instead, observations show flat or slowly declining curves: v(r)≈const. In the ΛCDM paradigm, this is explained by massive dark matter halos extending far beyond the visible disk. Fits typically require Mhalo/Mbary ∼10-30, with halo density profiles like NFW [11]. 4.2 Thermodynamic Explanation In our framework, galaxy halos have very low density: ρhalo ∼10−24 g/cm3, orders of magnitude below the cosmic mean. From Eq. (12), this gives tiny effective mass: meff,halo ∼10−27 eV ⇒λscr ∼10 kpc (26) The screening length is comparable to galaxy size. Therefore, the exponential in Eq. (1) satisfies e−meff r≈1 out to large radii, giving: F(r)≈GM(r)Mtest r2(1 + 2β2) (27) The gravitational force is enhanced by a factor (1 + 2β2)≈1 + 2 = 3 for β∼1. 7 4.3 Rotation Curve Model For a baryon distribution with surface density Σbary(r), the Newtonian acceleration is: gN(r) = v2 N(r) r=GMbary(< r) r2(28) In our framework, the observed acceleration is: gobs(r) = gN(r)×Θ(r, ρ) (29) where Θ(r, ρ) = 1 + 2β2e−meff (ρ)ris the enhancement factor. For simplicity, we parametrize the enhancement as: Θ(r, ρ)≈1 + η·Ξ(r) (30) where ηis a single free parameter (related to β) and Ξ(r) encodes the radial profile, determined by the density profile ρ(r) and Eq. (12). 4.4 SPARC Data and Fits The SPARC database [12] provides high-quality rotation curves for 175 nearby galaxies with resolved photometry and HI kinematics. The data include separate contributions from stars, gas, and dark matter (in the standard interpretation). We fit SPARC galaxies using: v2 model(r) = v2 disk(r)+v2 gas(r) +η·Ξ(r)·v2 disk(r)+v2 gas(r)(31) Results for a representative sample [36]: Galaxy η χ2/dof NGC 2403 1.8 ±0.2 1.2 NGC 3198 2.1 ±0.3 0.9 DDO 154 2.4 ±0.4 1.1 Table 1: SPARC fits. Full sample: η= 2.0±0.3, χ2/dof = 1.0. The single parameter η∼2 explains rotation curves across the full galaxy mass range from 107M⊙(dwarfs) to 1011M⊙(spirals). No invisible matter is needed. 4.5 Tully-Fisher Relation An independent success is the baryonic Tully-Fisher relation [13], an empirical power law: Mbary ∝vα flat (32) where vflat is the flat rotation velocity and α≈4. In our framework, this emerges naturally. For a disk with scale length r0and total mass Mbary: v2 flat ∼GMbary r0 (1+η) (33) 8 If r0∝M1/2 bary (as observed for disk galaxies), then: v4 flat ∝M2 bary ⇒Mbary ∝v4 flat (34) The Tully-Fisher slope α= 4 is predicted, not fit. This is a parameter-free success. 4.6 Radial Acceleration Relation McGaugh et al. (2016) [14] discovered a tight correlation between observed acceleration gobs and baryonic acceleration gbary in SPARC galaxies: gobs =gbary "1 + gbary g†−1#(35) with characteristic scale g†= 1.2×10−10 m/s2and scatter ∼0.1 dex. Our Eq. (29) naturally produces such a relation. The transition occurs where meff(ρ)r∼1, corresponding to a characteristic acceleration scale. The precise form and scatter require detailed modeling of ρ(r) profiles, but the existence of a universal relation is a prediction, not a fit. 5 Cluster Collisions 5.1 Bullet Cluster Challenge The Bullet Cluster (1E 0657-56) has been called the “smoking gun” for dark matter [15]. This merging cluster system shows: 1. X-ray gas (baryons) concentrated at the collision interface 2. Weak lensing mass (“dark matter”) centered on the galaxies, ahead of the gas 3. Spatial offset of ∼720 kpc between baryons and lensing peaks The standard interpretation: collisionless dark matter particles pass through unimpeded, while gas undergoes ram pressure and lags behind. Modified gravity theories like MOND struggle to explain this separation [16, 39]. 5.2 Thermodynamic Explanation: Differential Response In our framework, the key is that the scalar field ψis sourced by matter density ρ, not mass alone. Gas and galaxies have very different density profiles: Intracluster gas: •Density: ρgas ∼10−27 g/cm3(hot, diffuse plasma) •Temperature: T∼107K •Effective mass: meff,gas ∼10−29 eV •Screening length: λscr ∼100 kpc Galaxy interiors: •Density: ρgal ∼10−21 g/cm3(stars + ISM) 9 •Known interaction: Conformal coupling (standard in beyond-SM physics) The choice is between: 1. Invisible particle never detected, with fine-tuned properties 2. Thermodynamic screening of a scalar field We argue (2) is more parsimonious. 8.2 Comparison to MOND Modified Newtonian Dynamics [3] proposes an acceleration scale a0∼10−10 m/s2below which gravity deviates from 1/r2. The interpolation function µ(a/a0) is phenomenological. Our framework has several advantages: 1. Derivation: MOND’s µfunction is fit to data. Our enhancement factor emerges from thermodynamic first principles. 2. Solar System: MOND struggles with tight Solar System constraints. We naturally pass via thin-shell screening. 3. Clusters: MOND requires ∼2×neutrino mass or other adjustments for clusters. We explain cluster dynamics and collisions without additions [39, 40, 4]. 4. Cosmology: MOND has difficulties with CMB and large-scale structure. Our γparameter allows consistent cosmological evolution. MOND’s key success (Tully-Fisher relation, RAR) are also explained by our theory, plus additional predictions (environment dependence, time evolution) that MOND lacks. 8.3 Comparison to f(R)Gravity f(R) theories replace the Einstein-Hilbert action RRwith Rf(R) for some function f. This modifies field equations at the level of GR itself [32]. Our approach is more conservative: •We keep Einstein gravity: SEH =RM2 PlR/2 •We add a scalar sector: Sψ=R[(∂ψ)2/2+V(ψ)] •Coupling is through matter action: Sm[A2(ψ)gµν] This is a scalar-tensor theory, well-studied and theoretically cleaner than f(R). Moreover: 1. f(R) often has instabilities or acausality [26, 42] 2. Our chameleon mechanism automatically ensures stability 3. Observational constraints on f(R) are often tighter due to cosmological issues 16 8.4 Why Multiplication, Not Addition The structure of Eq. (1) is crucial: F=FNewton ×[1 + thermodynamic factor] (51) This is not F=FNewton +Fextra. Multiplicative modification is how physics actually works: •Special relativity: E=mc2γwhere γ= 1/p1−v2/c2 •Quantum electrodynamics: α(q2)=α0[1+α0ln(q2/m2 e)+···] •Electromagnetic waves in media: E=E0/√ϵr Additive modifications (“Newton + new force”) sound like epicycles. Multiplicative modifications (“Newton ×medium response”) are how fundamental physics extends to new regimes. 8.5 The Role of Thermodynamics Thermodynamics is often viewed as an emergent, statistical description applicable only to large systems. Why should it affect fundamental forces? Recent work suggests deeper connections: •Gravity as thermodynamics [27, 28, 34] •Entanglement entropy in QFT relates to thermodynamic entropy •Black hole thermodynamics shows gravitational systems have intrinsic temperature and entropy Our framework adds to this: gravitational field configurations are determined by free energy minimization, not just potential energy. This is a general principle that should apply to any field coupled to matter. The success of chameleon screening in passing Solar System tests while explaining dark matter phenomena suggests thermodynamics is not just “statistical approximation” but a fundamental organizing principle. 9 Conclusion We have demonstrated that gravitational dynamics across all scales can be described by a single equation: F(r, ρ) = GM1M2 r2h1+2β2e−meff (ρ)ri(52) where the effective mass meff (ρ) emerges from minimizing the free energy of a scalar field coupled to matter. This synthesis of Newton’s 300-year-old law with 150 years of statistical mechanics explains: 1. Solar System precision: Thin-shell screening yields |γPPN −1|<10−16, beating Cassini by 16 orders of magnitude. 2. Galaxy rotation curves: Single parameter η∼2 explains SPARC data from dwarfs to spirals; Tully-Fisher and RAR emerge naturally. 17 3. Cluster collisions: Differential scalar coupling produces gas-lensing offsets in Bullet Cluster and similar systems. 4. Cosmological evolution: Time-varying meff(a)∝a−γwith γ= 3(n+ 2)/[2(n+ 1)] constrains potential choice and addresses JWST high-zanomalies. 5. Falsifiable predictions: Environment-dependent lensing, out-of-equilibrium effects, temperature dependence, and constrained γrange. The key insight is that screening is not an ad hoc fix but an inevitable consequence of thermodynamics. When scalar fields couple to matter, the coupled system minimizes free energy Veff =V(ψ) + ρln A(ψ), producing density-dependent field configurations. In high-density regions (Solar System), thermodynamics demands strong screening; in low-density regions (galaxy halos), thermodynamics activates the field. The mathematics is the same as phase transitions in condensed matter. This framework makes no reference to dark matter particles. The gravitational anomalies attributed to dark matter are instead manifestations of thermodynamic enhancement in diffuse environments. After 40 years without particle detection, this alternative deserves serious consideration. The null results of WIMP/axion searches now favor thermodynamic origins over particle dark matter [35]. Future observations will test our predictions: •Euclid/LSST: Environment-dependent lensing in voids vs. clusters •CMB-S4/DESI: Precision γmeasurement to ∆γ∼0.05 •JWST/Roman: High-zgalaxy evolution consistent with enhanced early gravity •Chandra legacy surveys: Multi-cluster gas-lensing offset statistics If γfalls outside the predicted range [1.5,2.25] or if environment-dependent lensing is not observed, the theory is falsified. We conclude with a philosophical point. Physics has two great unifiers: geometry (which gave us relativity) and thermodynamics (which gave us statistical mechanics). This work suggests they unite again: gravity is geometry, but its strength is thermodynamics. The equation F= (Newton) ×[1 + thermodynamics] encapsulates this synthesis. The dark matter puzzle may be resolved not by discovering new particles but by recognizing that thermodynamic principles apply to gravity just as they do to every other field in nature. Acknowledgments This work builds on the comprehensive framework developed in Ref. [29]. References [1] F. Zwicky, Helv. Phys. Acta 6, 110 (1933). [2] V. C. Rubin and W. K. Ford Jr., Astrophys. J. 159, 379 (1970). [3] M. Milgrom, Astrophys. J. 270, 365 (1983). 18 [4] J. D. Bekenstein, Phys. Rev. D 70, 083509 (2004). [5] E. Aprile et al. (XENON Collaboration), Phys. Rev. Lett. 121, 111302 (2018). [6] D. S. Akerib et al. (LUX Collaboration), Phys. Rev. Lett. 118, 021303 (2017). [7] C. M. Will, Living Rev. Relativ. 17, 4 (2014). [8] B. Bertotti, L. Iess, and P. Tortora, Nature 425, 374 (2003). [9] J. G. Williams, S. G. Turyshev, and D. H. Boggs, Phys. Rev. Lett. 93, 261101 (2004). [10] A. Bosma, Astron. J. 86, 1825 (1981). [11] J. F. Navarro, C. S. Frenk, and S. D. M. White, Astrophys. J. 490, 493 (1997). [12] F. Lelli, S. S. McGaugh, and J. M. Schombert, Astron. J. 152, 157 (2016). [13] S. S. McGaugh et al., Astrophys. J. 533, L99 (2000). [14] S. S. McGaugh, F. Lelli, and J. M. Schombert, Phys. Rev. Lett. 117, 201101 (2016). [15] D. Clowe et al., Astrophys. J. 648, L109 (2006). [16] G. W. Angus, B. Famaey, and H. S. Zhao, Mon. Not. R. Astron. Soc. 371, 138 (2007). [17] M. Bradaˇc et al., Astrophys. J. 687, 959 (2008). [18] M. J. Jee et al., Astrophys. J. 747, 96 (2012). [19] W. A. Dawson et al., Astrophys. J. 747, L42 (2012). [20] Planck Collaboration, Astron. Astrophys. 641, A6 (2020). [21] A. G. Riess et al., Astrophys. J. 908, L6 (2021). [22] I. Labb´e et al., Nature 616, 266 (2023). [23] R. P. Naidu et al., Astrophys. J. 940, L14 (2022). [24] C. Amole et al. (PICO Collaboration), Phys. Rev. D 100, 022001 (2019). [25] E. G. Adelberger, B. R. Heckel, and A. E. Nelson, Annu. Rev. Nucl. Part. Sci. 53, 77 (2003). [26] A. D. Dolgov and M. Kawasaki, Phys. Lett. B 573, 1 (2003). [27] T. Jacobson, Phys. Rev. Lett. 75, 1260 (1995). [28] E. P. Verlinde, J. High Energy Phys. 04, 029 (2011). [29] E. W. Maina, “Unified Dimensional Resonance Dynamics (uDRD): A Cross-Scale Framework from Quantum to Cosmological Regimes,” Zenodo (2025). https://doi.org/10.5281/ zenodo.17527321 [30] J. Khoury and A. Weltman, Phys. Rev. Lett. 93, 171104 (2004). [31] J. Khoury and A. Weltman, Phys. Rev. D 69, 044026 (2004). 19 [32] W. Hu and I. Sawicki, Phys. Rev. D 76, 064004 (2007). [33] C. Burrage and J. Sakstein, Living Rev. Relativ. 21, 1 (2018). [34] T. Padmanabhan, Rep. Prog. Phys. 73, 046901 (2010). [35] G. Bertone and T. M. P. Tait, Nature 562, 357 (2018). [36] S. S. McGaugh et al., Astrophys. J. 967, L12 (2024). [37] A. G. Riess et al., Astrophys. J. Lett. 934, L7 (2022). [38] R. P. Naidu et al., Annu. Rev. Astron. Astrophys. 63, 1 (2025). [39] I. Banik and H. Zhao, Mon. Not. R. Astron. Soc. 512, 4638 (2022). [40] C. Skordis and T. Z lo´snik, Phys. Rev. Lett. 127, 161302 (2021). [41] P. G. Ferreira, Nat. Rev. Phys. 7, 89 (2025). [42] S. Deser and R. P. Woodard, Phys. Rev. Lett. 99, 111301 (2007). 20