REVIEW 2 major objections 5 minor 27 references
Ultralight dark photon as a model for early universe dark matter
T0 review · 2 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read An ultralight dark photon that acts as radiation in the early universe and then as dark matter shrinks the BAO sound horizon by about 6 percent and raises the inferred Hubble constant to 73 km/s/Mpc, resolving the Hubble tension.
desk verdict The vector-field equation-of-state transition is correct, but the cosmological application misnormalizes the early radiation density and the claimed Hubble-tension resolution does not survive energy conservation. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing mechanism is the homogeneous solution of the Maxwell–Proca equation $\ddot{A} + \frac{1}{2t}\dot{A} + m^2 A = 0$ in the radiation-dominated background $a(t) = \left(2\sqrt{\Omega_r}H_0 t\right)^{1/2}$, whose closed form involves the Bessel functions $J_{\pm 1/4}$ and $Y_{\pm 1/4}$. Its role is to turn the effective equation of state into a step function, $w = 1/3$ for $t < m^{-1}$ and $w = 0$ for $t > m^{-1}$, converting the vector field from a radiation component into cold dark matter without any tuned potential. Isotropy is preserved by averaging over a triplet of mutually orthogonal fields of equal mass, which diagonalizes the stress-energy tensor. This solution is inserted into the modified Hubble function $E(a) = \sqrt{(\Omega_r + \Omega_A)/a^4 + (\Omega_m - \Omega_A)/a^3 + \Omega_\Lambda}$ before the transition scale $a_1$ and the ordinary $\Lambda$CDM form afterward; the sound-horizon integral of $1/E(a)$ is the quantity whose 6 percent reduction is the paper's main numerical result.
What would settle it
A direct calculation would settle the central claim: impose continuity of the vector-field energy density at the transition scale $a_1 = \left(2\sqrt{\Omega_r}H_0/m\right)^{1/2}$, so that $\rho_A(a) = \Omega_A a_1/a^4$ for $a < a_1$ rather than $\Omega_A/a^4$, and re-evaluate the sound-horizon integral. Since $a_1$ is of order $10^{-4}$, the early-radiation enhancement shrinks by roughly $10^4$, the claimed 6 percent reduction of $r_s$ disappears, and the inferred $H_0$ falls back toward the standard $\Lambda$CDM value. Observationally, because big-bang nucleosynthesis lies well before the transition for most of the quoted masses, the paper's $\Omega_A/a^4$ branch would contribute $\Delta N_{\rm eff} \approx \frac{8}{7}\frac{\Omega_A}{\Omega_\gamma}\left(\frac{11}{4}\right)^{4/3}$ of order 10 to 100, whereas measurements of light-element abundances and the CMB require $N_{\rm eff} \approx 3$.
Extended reading notes
Core claim
On its own terms, the paper's central claim is that a massive vector field in the radiation-dominated expanding universe has the required two-stage history automatically: the Bessel-field solution $A(t) \propto (mt)^{1/4}\left[c_1 J_{1/4}(mt) + c_2 Y_{1/4}(mt)\right]$ yields an energy density $\rho \propto a^{-4}$ for $t \ll m^{-1}$ and $\rho \propto a^{-3}$ for $t \gg m^{-1}$, so no extra dynamics is needed to make the component hot early and cold late. The paper then claims that replacing a fraction $\Omega_A$ of the cold dark matter with a triplet of mutually orthogonal such fields preserves isotropy and contributes the term $\Omega_A/a^4$ to the early radiation density, lowering the sound horizon from about 145 Mpc by roughly 6 percent. Through the BAO constraint $c/(r_s H_0) = 29.63$, this raises the inferred Hubble constant to $H_0 = 73$ km/s/Mpc for masses $m$ in $10^{-27}$–$10^{-25}$ eV and densities $\Omega_A$ in $10^{-5}$–$10^{-2}$, matching local supernova and lensing distance measurements and thereby resolving the Hubble tension while leaving the late-time expansion history of $\Lambda$CDM intact.
Load-bearing premise
The load-bearing premise is that one density parameter $\Omega_A$ can describe the vector field on both sides of the transition: its energy density enters as $\Omega_A/a^4$ while radiation-like and as $\Omega_A/a^3$ while matter-like, which makes the density jump by a factor of order $1/a_1 \approx 10^4$ at $t = m^{-1}$; requiring energy conservation across the transition would instead reduce the early radiation-like term by that same factor of about $10^4$.
Editorial extensions
If this is right
- The Hubble tension would be resolved without touching dark energy: standard ΛCDM parameters, plus a small dark-matter admixture that was briefly radiation-like, reproduce $H_0 = 73$ km/s/Mpc.
- The baryon-acoustic-oscillation standard ruler shrinks by about 6 percent, from roughly 145 Mpc to 137 Mpc, which is exactly the shift needed to bring CMB-calibrated BAO distances into line with the local distance ladder.
- The dark sector must contain a triplet of mutually orthogonal ultralight vector fields of equal mass in the $10^{-27}$ to $10^{-25}$ eV range, a structural prediction that distinguishes the scenario from axion-based alternatives.
- Late-time cosmology is essentially unchanged, since $\Omega_A \le 4.4 \times 10^{-3}$ lies below current uncertainties in the dark-matter density, so the late-universe successes of ΛCDM survive.
Reading between the lines
- A consequence the paper leaves implicit: for most of the quoted mass range the radiation-like branch extends through big-bang nucleosynthesis, adding an effective number of relativistic species of order tens, so existing bounds on $N_{\rm eff}$ would test — and likely exclude — the parameter region independently of the sound-horizon fit.
- The sharp step in the equation of state should imprint a scale-dependent signature in the matter power spectrum: modes entering the horizon before the transition expand faster, so a tilt or cutoff near the horizon scale at $z \approx 3000$ is a concrete target for small-scale structure surveys.
- Unlike scalar early-dark-energy models, which shift the dark-energy equation of state near matter-radiation equality, this vector mechanism places the extra density at a time set by the field mass; the two proposals differ in the timing and shape of the extra early density and can be told apart with CMB lensing and BAO data.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. This paper studies a homogeneous massive vector field in a flat Friedmann universe. Solving the Proca equation, it shows that for mt ≪ 1 the field has equation of state w = 1/3 and energy density scaling as a^{-4}, while for mt ≫ 1 it scales as a^{-3} and behaves as cold dark matter. The authors then extend ΛCDM by replacing a fraction ΩA of cold dark matter with this field, approximating the equation-of-state change by a step at t = m^{-1}. They claim that for m ∼ 10^{-27}–10^{-25} eV and ΩA ∼ 10^{-5}–10^{-2} the extra radiation-like component reduces the BAO sound horizon by about 6%, and using the BAO constraint c/(r_s H0) = 29.63 they infer H0 = 73 km/s/Mpc, thereby resolving the Hubble tension.
Significance. The analytical vector-field solution in Section II is elegant and appears correct, and the idea of a field that is naturally radiation-like at early times and cold-dark-matter-like later is attractive. The paper is also transparent that the model has two free parameters and that the H0 value is obtained by fitting them. If the cosmological application were sound, this would be a worthwhile contribution to the Hubble-tension literature. However, the central quantitative claim rests on a normalization error in the modified Friedmann equation; once that is corrected the claimed 6% effect disappears. As written, the paper does not demonstrate a resolution of the Hubble tension.
major comments (2)
- [III.B, Eqs. (25)-(26)] Equations (25) and (26) treat ΩA both as the present-day density parameter of the vector field and as the coefficient of a^{-4} in the early radiation-like branch. For a component whose density today is ρ_crit,0 ΩA and which makes a step transition from w = 1/3 at a < a1 to w = 0 at a > a1, continuity at a = a1 requires ρ_A(a) = ρ_crit,0 ΩA a1 / a^4 for a < a1. The paper instead uses ρ_A = ρ_crit,0 ΩA / a^4 in the first branch of Eq. (26), which overestimates the early density by a factor 1/a1. With the quoted masses, a1 = (2√Ωr H0 / m)^{1/2} lies in the range ∼2×10^{-5}–2×10^{-4}, so the early contribution is overestimated by four to five orders of magnitude. This is the term responsible for the claimed ∼6% reduction of r_s, and correcting the normalization removes the effect.
- [III.B and Fig. 1] The paper does not predict H0; it scans the two free parameters m and ΩA over a region and reports that the value H0 = 73 km/s/Mpc is attained. This is a legitimate fitting procedure if presented as such, but the conclusion that the model resolves the Hubble tension requires demonstrating consistency with the full CMB, BAO, and local-distance datasets. No such fit is presented; only the BAO product constraint and the local H0 value are used to select the parameter region. The abstract and conclusions should be moderated accordingly.
minor comments (5)
- [Eq. (27)] Equation (27) does not algebraically follow from Eq. (26): with E(a) as in Eq. (26), the integrand 1/(a^2 E) equals 1/sqrt(Ωr + ΩA + (Ωm − ΩA)a + ΩΛ a^4) in the first branch and 1/sqrt(Ωr + Ωm a + ΩΛ a^4) in the second, whereas Eq. (27) writes ΩΛ a^2 in both denominators. This term is negligible before recombination, so it is a typo rather than the source of the main error, but it should be corrected.
- [Introduction vs III.B] The abstract and introduction state that the early radiation-like behavior applies for z > 3000, while Section III.B uses z > 3600; the text should use a single consistent boundary.
- [Conclusions] The conclusions quote the parameter range ΩA ∼ 10^{-5}–10^{-2}, which is broader than the interval 9.0×10^{-5} < ΩA < 4.4×10^{-3} stated in Eq. (28); the authors should specify which range is actually used.
- [Fig. 1] Figure 1 is referenced but not visible in the manuscript text; the published version should include axes, labels, and the shaded region described in the caption.
- [Conclusions] The paper calls the field a 'dark photon' but does not specify its kinetic mixing or coupling to visible matter; the discussion should clarify whether the experimental constraints cited in Refs. [20–24] apply to the ultralight mass range considered here.
Circularity Check
The claimed H0 = 73 km/s/Mpc outcome is a parameter fit (the shaded region is selected to produce that value), not an independent prediction; the Section II vector-field derivation is self-contained.
-
fitted input called prediction
[Section III.B, after Eq. (27) and Fig. 1; see also Abstract and Introduction.]
"We can now use the relation (24) to find the value of the Hubble constant H0 via the known value for the sound horizon rs obtained within the modified ΛCDM model. We find that the Hubble constant gets the value H0 = 73 km s−1 Mpc−1 when the parameters m and ΩA lie within the shaded region on the graph in Fig. 1."
The two free parameters, m and ΩA, are not derived from first principles; they are scanned, and the shaded region is defined as the locus where the model's sound horizon rs (Eq. (27)) combined with the BAO constraint (24) yields H0 = 73 km/s/Mpc, which is precisely the external local-measurement value the paper aims to reproduce. Hence, the statement that 'H0 = 73 when the parameters lie in the shaded region' is true by construction: the shaded region is selected by requiring that H0 equal the target.
full rationale
The core field-theory result—that a massive vector field in a radiation-dominated universe has energy density scaling as a^{-4} for mt << 1 and a^{-3} for mt >> 1, with an equation of state w=1/3 turning to w=0—is derived explicitly in Section II from the Proca equation and the stress-energy tensor. That part is self-contained and does not reduce to its inputs. The circularity is confined to the cosmological application: the model has two free parameters (m, ΩA), and the paper selects the parameter region where these parameters make the computed H0 equal to the externally assumed 73 km/s/Mpc. Since that region is defined by the target value, reporting H0 = 73 in that region is a restatement of the fitting condition rather than a prediction. This warrants a score of 6: a central 'prediction' reduces by construction to a parameter fit. I do not score the energy-density discontinuity at a=a1 between Eq. (26) and Eq. (27) as circularity: that is a physical consistency flaw (the early radiation-like density should be ΩA a1/a^4 for continuity), but it is not a self-referential reduction. There is also no load-bearing self-citation or imported uniqueness theorem; the few self-citations (Refs. [15], [24]) are not used to force the model's central choice.
Assumptions & free parameters
free parameters (2)
- Omega_A (vector-field density parameter) =
range 9.0e-5 to 4.4e-3 (Eq. 28); a subregion is selected for H0=73
- m (vector-field mass) =
range about 1e-27 to 1e-25 eV
assumptions (5)
- domain assumption Radiation-dominated scale factor a(t) = (2*sqrt(Omega_r)*H0*t)^(1/2) holds throughout the solution's regime
- domain assumption The homogeneous vector field is produced non-thermally after inflation with generic initial conditions c1+c2 not equal to 0
- domain assumption Isotropy is preserved by a triplet of mutually orthogonal vector fields with equal mass and magnitude
- ad hoc to paper The equation-of-state transition can be approximated by a step function in time and inserted directly into the Friedmann equation with the same Omega_A on both sides
- domain assumption The vector field did not thermalize with the plasma and interacts very weakly with visible matter
invented entities (1)
-
Ultralight massive vector field (dark photon) with m~1e-27 to 1e-25 eV
Cite this review
Pith. "Pith review of Ultralight dark photon as a model for early universe dark matter." pith.science (2026). https://pith.science/paper/ATTKU6NP
@misc{pith2026190809432,
author = {Pith},
title = {Pith review of: Ultralight dark photon as a model for early universe dark matter},
year = {2026},
howpublished = {\url{https://pith.science/paper/ATTKU6NP}},
note = {Machine review of arXiv:1908.09432}
}
abstract
Dark photon is a massive vector field which interacts only with the physical photon through the kinetic mixing. This coupling is assumed to be weak so that the dark photon becomes almost unobservable in processes with elementary particles, but can serve as a dark matter particle. We argue that in very early Universe ($z>3000$) this vector field may have the equation of state of radiation ($w=1/3$) but later behaves as cold dark matter ($w=0$). This may slightly change the expansion rate of the Universe at early time and reduce the value of the sound horizon of baryon acoustic oscillations (standard ruler). As a result, in this model the value of the Hubble constant appears to be larger than that in the standard $\Lambda$CDM model. In particular, it is sufficient to have the dark photon mass of order $m\sim 10^{-27}-10^{-25}$ eV to fit the value of the Hubble constant to $H_0 = 73$ km$\cdot$s$^{-1}$Mpc$^{-1}$ thus resolving the Hubble tension.
Figures
Reference graph
Works this paper leans on
- [1]
-
[2]
(18) 1 More generally, the solution of Eq
1 a2√ mt for mt≫ 1. (18) 1 More generally, the solution of Eq. (14) may be written as (mt) 1 4 ( A1J 1 4 (mt) + A2Y 1 4 (mt) ) with two arbitrary constant vectors A1 and A2. Thus, taking into account Eq. (13), we conclude that the energy density of the homogeneous massive vector field in the early Universe scales as radiation fort≪m−1 and as cold dark matt...
-
[3]
We assume that this case is not realized in Nature since it requires special initial conditions. We stress that the massive vector field solution (15) is not universal as it applies only in the radiation- dominated epoch. At later times, e.g., in the matter- dominated epoch, the behavior of the vector field may change, but its density should be negligible c...
-
[4]
during the time t<m −1, but for t>m −1 it changes to the cold dark matter ( ¯w = 0). We consider an implication of the massive-vector-field extended ΛCDM model to resolve the Hubble tension. As was argued in [2], this tension may be naturally resolved by increasing the number of relativistic particle species during the radiation-dominated epoch. Since the ...
-
[5]
A. G. Riess et al. , Astrophys. J. 826, no.1, 56 (2016), arXiv:1604.01424 [astro-ph.CO]
arXiv 2016
-
[6]
A. G. Riess et al. , Astrophys. J. 861, no. 2, 126 (2018), arXiv:1804.10655 [astro-ph.CO]
arXiv 2018
-
[7]
V. Bonvin et al. , Mon. Not. Roy. Astron. Soc. 465, no. 4, 4914 (2017), arXiv:1607.01790 [astro-ph.CO]
work page Pith review arXiv 2017
-
[8]
S. Birrer et al. , Mon. Not. Roy. Astron. Soc. 484, 4726 (2019), [arXiv:1809.01274 [astro-ph.CO]]. 6
arXiv 2019
Show all 27 references
-
[9]
Holdom, Phys
B. Holdom, Phys. Lett. B 166, 196 (1986)
1986
-
[10]
Galison and A
P. Galison and A. Manohar, Phys. Lett. B 136, 279 (1984)
1984
-
[11]
Poulin, T
V. Poulin, T. L. Smith, T. Karwal and M. Kamionkowski, Phys. Rev. Lett. 122, no. 22, 221301 (2019), arXiv:1811.04083 [astro-ph.CO]
2019 arXiv
-
[12]
Poulin, T
V. Poulin, T. L. Smith, D. Grin, T. Karwal and M. Kamionkowski, Phys. Rev. D 98, no. 8, 083525 (2018), arXiv:1806.10608 [astro-ph.CO]
2018 arXiv
-
[13]
Golovnev, V
A. Golovnev, V. Mukhanov and V. Vanchurin, JCAP 0806, 009 (2008), arXiv:0802.2068 [astro-ph]
2008 arXiv
-
[14]
P. W. Graham, J. Mardon and S. Rajendran, Phys. Rev. D 93, no. 10, 103520 (2016), arXiv:1504.02102 [hep-ph]
2016 arXiv
-
[15]
A. E. Nelson and J. Scholtz, Phys. Rev. D 84, 103501 (2011), arXiv:1105.2812 [hep-ph]
2011 arXiv
-
[16]
Arias, D
P. Arias, D. Cadamuro, M. Goodsell, J. Jaeckel, J. Re- dondo and A. Ringwald, JCAP 1206, 013 (2012), arXiv:1201.5902 [hep-ph]
2012 arXiv
-
[17]
Dimopoulos, Phys
K. Dimopoulos, Phys. Rev. D 74, 083502 (2006), hep- ph/0607229
2006
-
[18]
D. D. Ryutov, D. Budker and V. V. Flambaum, Astro- phys. J. 871, no. 2, 218 (2019), arXiv:1708.09514 [astro- ph.GA]
2019 arXiv
-
[19]
Tanabashi et al
M. Tanabashi et al. (Particle Data Group), Phys. Rev. D 98, 030001 (2018)
2018
-
[20]
C. P. Ma and E. Bertschinger, Astrophys. J. 455, 7 (1995), astro-ph/9506072
1995 arXiv
-
[21]
Komatsu et al
E. Komatsu et al. [WMAP Collaboration], Astrophys. J. Suppl. 180, 330 (2009), arXiv:0803.0547 [astro-ph]
2009 arXiv
-
[22]
Aubourg et al
´E. Aubourg et al. , Phys. Rev. D 92, no. 12, 123516 (2015), arXiv:1411.1074 [astro-ph.CO]
2015 arXiv
-
[23]
Raggi and V
M. Raggi and V. Kozhuharov, Riv. Nuovo Cim. 38, no. 10, 449 (2015)
2015
-
[24]
I. G. Irastorza and J. Redondo, Prog. Part. Nucl. Phys. 102, 89 (2018), arXiv:1801.08127 [hep-ph]
2018 arXiv
-
[25]
Jaeckel, Frascati Phys
J. Jaeckel, Frascati Phys. Ser. 56, 172 (2012), arXiv:1303.1821 [hep-ph]
2012 arXiv
-
[26]
Arik et al
E. Arik et al . [CAST Collaboration], JCAP 0902, 008 (2009), arXiv:0810.4482 [hep-ex]
2009 arXiv
-
[27]
V. V. Flambaum, I. B. Samsonov and H. B. Tran Tan, Phys. Rev. D99, no. 11, 115019 (2019), arXiv:1904.02271 [hep-ph]
2019 arXiv
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.