Light Particle Solution to the Cosmic Lithium Problem

We point out that the cosmological abundance of ${}^7\mathrm{Li}$ can be reduced down to observed values if during its formation, big bang nucleosynthesis is modified by the presence of light electrically neutral particles $X$ that have substantial interactions with nucleons. We find that the lithium problem can be solved without affecting the precisely measured abundances of deuterium and helium if the following conditions are satisfied: the mass (energy) and lifetimes of such particles are bounded by $1.6\mathrm{MeV}\le m_X(E_X)\le 20\mathrm{MeV}$ and $\mathrm{few}100\mathrm{s}\lesssim {\tau }_X\lesssim {10}^4\mathrm{s}$, and the abundance times the absorption cross section by either deuterium or ${}^7\mathrm{Be}$ are comparable to the Hubble rate, $n_X{\sigma }_{\mathrm{abs}}v\sim H$, at the time of ${}^7\mathrm{Be}$ formation. We include $X$-initiated reactions into the primordial nucleosynthesis framework, observe that it leads to a substantial reduction of the freeze-out abundances of ${}^7\mathrm{Li}+{}^7\mathrm{Be}$, and find specific model realizations of this scenario. Concentrating on the axionlike-particle case, $X=a$, we show that all these conditions can be satisfied if the coupling to $d$ quarks is in the range of $f_d^{-1}\sim {\mathrm{TeV}}^{-1}$, which can be probed at intensity frontier experiments.

Figure 1

Spallation of a nucleus due to absorption of a bosonic state $X$.

Figure 2

The contours of light element abundances as a function of the two reaction rates R1 and R2 in Scenario A, for ${\tau }_X\gg t_{\mathrm{BBN}}$ (top panel), and ${\tau }_X={10}^3\mathrm{s}$ (lower panel); ${\sigma }_D$ is constant along the dotted lines. Inside the shaded regions, the lithium problem is solved.

Figure 3

Temperature evolution of elemental abundances, with BBN modified by R2, initiated by $X$ with ${\tau }_X={10}^3\mathrm{s}$ and $(n_X/n_b){\sigma }_{\mathrm{D}}v=5\times {10}^{-32}{\mathrm{cm}}^2$. The temporary increase in $n$ leads to the suppression of ${}^7\mathrm{Be}$ but does not affect $(\mathrm{D}/\mathrm{H}{)}_{\mathrm{BBN}}$. The dotted lines correspond to the prediction of standard BBN.

Figure 4

Lithium solution by ALPs that are injected from a progenitor state $X_p$ with mass $m_{X_p}=10\mathrm{MeV}$. The LSND sensitivity line is fixed, but all other contours can move vertically by adjusting the $X_p$ initial abundance $n_{X_p}/n_b$.