Improved Measurement of Solar Neutrinos from the Carbon-Nitrogen-Oxygen Cycle by Borexino and Its Implications for the Standard Solar Model

We present an improved measurement of the carbon-nitrogen-oxygen (CNO) solar neutrino interaction rate at Earth obtained with the complete Borexino Phase-III dataset. The measured rate, $R_{\mathrm{CNO}}={6.7}_{-0.8}^{+2.0}\text{counts}/(\mathrm{day}\times 100\mathrm{tonnes})$, allows us to exclude the absence of the CNO signal with about $7\sigma$ C.L. The correspondent CNO neutrino flux is ${6.6}_{-0.9}^{+2.0}\times {10}^8{\mathrm{cm}}^{-2}{\mathrm{s}}^{-1}$, taking into account the neutrino flavor conversion. We use the new CNO measurement to evaluate the C and N abundances in the Sun with respect to the H abundance for the first time with solar neutrinos. Our result of $N_{\mathrm{CN}}=(5.78_{-1.00}^{+1.86})\times {10}^{-4}$ displays a $\sim 2\sigma$ tension with the “low-metallicity” spectroscopic photospheric measurements. Furthermore, our result used together with the ${}^7\mathrm{Be}$ and ${}^8\mathrm{B}$ solar neutrino fluxes, also measured by Borexino, permits us to disfavor at $3.1\sigma$ C.L. the “low-metallicity” standard solar model B16-AGSS09met as an alternative to the “high-metallicity” standard solar model B16-GS98.

Figure 1

Time evolution of the ${}^{210}\mathrm{Po}$ rate in the detector, visualized in terms of cylindrical $z$ slices of 0.1 m height and radius ${\rho }^2=(x^2+y^2)<2{\mathrm{m}}^2$. The horizontal black dashed lines represent the $z$ cut used in the CNO analysis. The low polonium field centers obtained from the monthly paraboloid fits with (white) and without (red) a cubic spline along the $z$ axis are also shown.

Figure 2

(a) Spectral fit (magenta) of the Borexino Phase-III data (black points) from January 2017 to October 2021 with a suppressed contribution of the cosmogenic ${}^{11}\mathrm{C}$ background (gray dashed). CNO $\nu$’s are shown as red solid line. The rates of $pep$ $\nu$’s (green) and ${}^{210}\mathrm{Bi}$ (blue) were constrained in the fit based on independent data. The energy estimator $N_h$, in which the fit is performed, represents the number of detected photoelectrons, normalized to 2000 live channels. (b) CNO-$\nu$ rate negative log-likelihood ($-2\mathrm{\Delta }\ln \mathcal{L}$) profile obtained from the multivariate spectral fit (dashed black line) and after folding in the systematic uncertainties (black solid line). The blue, violet, and gray vertical bands show 68% confidence intervals (CI) for the low-metallicity SSM B16-AGSS09met [$(3.52\pm 0.52)\mathrm{cpd}/100\text{tonnes}$] and the high-metallicity SSM B16-GS98 [$(4.92\pm 0.78)\mathrm{cpd}/100\text{tonnes}$) predictions [10, 23], and the new Borexino result including systematic uncertainty, respectively.

Figure 3

Top: spectral shape of the events after subtraction of all known backgrounds (black dots). The gray line is the fitted Monte Carlo-based CNO shape assuming standard neutrino interaction and oscillation. Bottom: residual (Res.) of the fit, defined as $(\text{model-data})/{\sigma }_{\mathrm{data}}$, shows the data compatibility with the expected shape of recoiled electrons from CNO $\nu$’s.

Figure 4

Results of the global analysis: $1\sigma$ regions allowed by solar neutrino and KamLAND reactor data (gray) and by Borexino $\mathrm{only}+\text{KamLAND}$ (green) in the ${\mathrm{\Phi }}_{\mathrm{B}}\text{−}{\mathrm{\Phi }}_{\mathrm{Be}}$, ${\mathrm{\Phi }}_{\mathrm{B}}\text{−}{\mathrm{\Phi }}_{\mathrm{CNO}}$, and ${\mathrm{\Phi }}_{\mathrm{Be}}\text{−}{\mathrm{\Phi }}_{\mathrm{CNO}}$ planes. The $1\sigma$ predictions of high-metallicity SSM B16-GS98 model (red) and low-metallicity SSM B16-AGSS09met (blue) are also shown. The best fit values for oscillation parameters are found to be $\mathrm{\Delta }{m}_{21}^2=7.50_{-0.18}^{+0.17}\times {10}^{-5}{\mathrm{eV}}^2$ and $\tan {\theta }_{12}=0.43_{-0.02}^{+0.04}$.

Figure 5

Comparison of abundance of $(\mathrm{C}+\mathrm{N})/\mathrm{H}$ in the solar photosphere, $N_{\mathrm{CN}}$, from spectroscopy (squares) and from the Borexino measurement (circle). The gray area highlights the uncertainty due to the precision of the CNO rate measurement. The white cross marks the result of the very same study repeated after changing the reference SSM from the B16-GS98 to the B16-AGSS09met.

Figure 6

Contributions to the $N_{\mathrm{CN}}$ error budget. The first two lines indicate the uncertainties linked to the experimental measurement of solar $\nu$ fluxes, the second group to the uncertainty of the nuclear cross sections, and the third group shows the suppressed contribution of environmental parameters (inflated by $\times 10$). The last lines show the impact of diffusion and of the precision of ${}^{13}\mathrm{N}/{}^{15}\mathrm{O}$ ratio predicted by SSM, as well as the total uncertainty.