End-of-the-week wrap covering the historical first ever detection of CNO neutrinos by Borexino, the ingredients required for a modified gravity theory to replace dark matter, and how to search for new light particles using the black hole mass gap.
#1 2006.15115: First Direct Experimental Evidence of CNO neutrinos by the Borexino collaboration
Stars such as our Sun are powered by nuclear fusion reactions transforming Hydrogen into Helium, with two mechanisms dominating the stage: the pp chain, and the CNO cycle, with the latter unsurprisingly associated to the elements Carbon, Nitrogen, and Oxygen (but also Fluorine). For stars such as the Sun it is the pp chain which dominates the energy production rate, whereas for heavier stars with mass >1.3 M⊙ (M⊙ means solar masses), the CNO cycle takes over. Since most of the stars in the Universe are more massive than the Sun, this means that the CNO cycle plays a fundamental role in transforming Hydrogen into Helium throughout the Universe. In the Sun, the CNO cycle contributes only about 1% to the energy production rate. Whereas pp neutrinos have been detected since the 1970s, CNO neutrinos are notoriously hard to detect, due both to the low neutrino energies involved, and the low rates. Another big problem is the Bismuth-210 (Bi210) β- decay background, although a breakthrough came in 1104.1335, where Villante et al. proposed to track the decay rate of Polonium-210 (Po210) to infer the decay rate of Bi210 (assuming the two are in equilibrium). This turned the quest for CNO neutrinos into a quest for Bi210 through Po210! As reported in this week’s paper, the Borexino collaboration finally succeeded in this task, finding for the first time evidence for CNO neutrinos at >5σ.
The CNO flux measured by Borexino is (7.0^+3.0_-2.0)×10^8 cm^-2*s^-1. Intriguingly, CNO neutrinos can be used to break the so-called composition-opacity degeneracy, which basically amounts to the fact that helioseismological observables (except for the surface Helium fraction actually) can be equally well fit if one changes the intrinsic opacity of the Sun or changes its metal content. However, this degeneracy is broken by CNO neutrinos, which directly probe the metal content of the Sun. CNO neutrinos are therefore the key towards solving the two decade-old solar composition problem (or solar modelling/opacity/metallicity problem), very well summarized by Natalie Wolchover in a Quanta Magazine piece (where I also featured). In this respect, it is intriguing (as shown e.g. in Fig. 5 of this week’s paper) that the Borexino measurement is actually mildly favoring a higher metallicity Sun, which if confirmed might turn the clock back 20 years to the old GS98 model or at least force us to seriously revise current solar models, although it should be pointed out that the Borexino error bars are rather large and even the low-metallicity AGSS09 model is consistent within ~1.7σ with the measurement. Or is this a 1.7σ tension? Unfortunately we are in the grey zone where whether you choose to label something as a consistency or a tension is rather arbitrary, perhaps we should refer to it as a 1.7σ curiosity? In any case, I have to say I am very surprised we have heard relatively little about Borexino’s CNO neutrino detection in popular science outlets (with much of the attention devoted to the XENON1T result). In my personal opinion this measurement is a milestone in physics, perhaps one of the most important ones in the last 20 years (almost if not directly comparable to the detection of gravitational waves, of the Higgs, or of neutrino oscillations). But this is my personal possibly biased view so please feel free to disagree!
#2 2007.00555: What is the price of abandoning dark matter? Cosmological constraints on alternative gravity theories by Kris Pardo and David Spergel
In the concordance cosmological model, an important ingredient is that of a collisionless cold dark matter (DM) component, required to address phenomena on both small scales (e.g. the flat rotation curves of galaxies, which first led to the need for DM) and on large scales (e.g. structure formation and the acoustic peak structure of the CMB). It is often tacitly assumed that DM has to be a particle or something similar, because it is really hard to obtain the required behavior of DM on such a wide range of scales by making use of modifications to gravity (I will briefly mention MOND in this context later). Nonetheless, this has not stopped model-builders from considering the possibility that modified gravity (MG) might still account for DM on cosmological scales. In this case, one of the first challenges is to correctly fit the observed CMB spectra in temperature and polarization. In this week’s paper, Pardo & Spergel point out an obvious but frequently overlooked fact. If your MG model wants to replace DM, it is already hard to have it fit the CMB, but that is not enough: it must explain both the CMB and the large-scale structure (LSS). In the standard paradigm, the baryon acoustic oscillations (BAO) pattern appears both in the CMB and in the LSS (e.g. in the clustering of galaxies). However, whereas BAOs have a large O(1) amplitude in the CMB, they have a much smaller O(0.04) amplitude in the LSS, with this erasure of acoustic fluctuations due to the DM-driven fluctuation growth. Any model of MG which seeks to replace DM has to be able to, beyond correctly fitting the CMB, explain why the acoustic feature is suppressed at late but not early times.
Pardo & Spergel focus on general MG models (in an agnostic way) where the dynamical equations for the growth rate of structure are linear. In this case, they can represent the effect of MG on the growth of fluctuations through a Green’s function, which is essentially the Fourier transform of the baryon perturbation transfer function. Heuristically, the real-space Green function captures the acceleration response of the MG model, and is computed relating the baryon density field at early times (as inferred from the CMB polarization field exploiting the continuity equation, which relates velocity as probed by polarization, to density) and at late times (as inferred from the galaxy 2-point correlation function). The resulting Green’s function is shown in Fig. 3 and has a rather peculiar form, in particular changing sign around r~150 Mpc (which just so happens to be the BAO scale). Therefore, it would appear like any linear MG gravity theory would have to have this type of acceleration response, and in particular contain the BAO scale in some way. Can one construct a non-baroque MG model with such an acceleration response? This is an open question, and a challenge which is left to smart model-builders.
This is a very interesting and well-written paper, emphasizing the need for MG theories replacing DM to correctly fit both the CMB and the LSS, though it is likely going to leave readers with two questions. The first, I have emphasized the fact that the MG theory has to be linear throughout this summary. What happens for non-linear theories? The claim at the end of Page 2 is that non-linear MG theories would lead to non-Gaussian features in the LSS, which are not observed. I buy this argument, but not completely. On very general grounds it is correct that non-linear behavior would lead to non-Gaussian features in the LSS, however it would be nice to see some actual numbers, e.g. the minimum level of non-Gaussianity (NG) predicted by viable MG theories replacing DM (expressed for instance as an equivalent fNL). After all, constraints on NG from the LSS are very hard to obtain, and at the moment we only have rather loose constraints on (primordial) non-Gaussianity from the LSS, which should perhaps be translatable to constraints on the type of NG predicted by these models. The second question many readers might ask is: “what about MOND”? Well, first of all I believe that this question is ill-posed to begin with, since MOND is a non-relativistic theory and does not per se have a cosmology associated to it. There are relativistic completions of MOND such as TeVeS for which the question of computing the predictions on cosmological scales is well-posed. By pure coincidence, on the same day Pardo & Spergel’s paper appeared, also 2007.00082 by Skordis & Złosnik appeared, proposing a new relativistic completion of MOND (based on a k-essence field, possibly arising from a gauged ghost condensate) which appears to fit the CMB very well, answering the question: “Can [relativistic completions of] MOND fit the CMB?”. However, this paper does not directly contradict Pardo & Spergel’s because it does not show the matter power spectrum, which would partly address the question of whether the model can fit both CMB and LSS. In this sense, I gladly await the promised upcoming reference [74] where this question should presumably be addressed.
#3 2007.00650: Missing in Axion: where are XENON1T's big black holes? by Djuna Croon, Samuel D. McDermott, and Jeremy Sakstein
Let me start off with a disclaimer: despite the appearance of “XENON1T” in the title, this is not an ambulance chasing paper, on the contrary it is a beautifully written paper pointing out a new completely unexpected discovery channel for new physics. In this sense perhaps “XENON1T” is somewhat of a clickbait (completely understandable during times like these). The point this paper is making is: “Can we use studies of black hole (BH) populations to probe new physics in the form of light particles?”. The answer is “yes”, and relies on the so-called BH mass gap (BHMG). This is a range of masses between 45 M ⊙ and 120 M⊙ where standard stellar theory predicts that BHs should not exist. The reason is that a sufficiently massive star can enter a zone where electron-positron pair production from the plasma reduces photon pressure and destabilizes the star, resulting in a so-called pulsation pair-instability supernova, which sheds lots of its mass and ends up as a lighter BH. Even heavier stars end up as pair-instability supernovae, which leave no BH remnant. The result is nicely shown in the right panel of Fig. 1, which shows the mass of the BH remnant as a function of the initial stellar mass. The final outcome of all these processes is a lower edge of the BHMG of about 45 M⊙. BHs heavier than 120 M⊙ (above the upper edge of the BHMG) can still form, but that is not relevant for this week’s paper.
The question Croon and collaborators address in this week’s paper is: can some new physics change the location of the lower edge of the BHMG? For example, a new particle coupled to electrons in the star’s plasma? Heuristically, these new particles could accelerate the pair-instability supernovae processes, resulting in less dramatic mass shedding, and hence raising the lower edge of the BHMG. One example could be an electrophilic axion which could explain the XENON1T signal (although it is always worth remarking that this explanation is in tension with stellar physics constraints). For the best-fit value of the axion-electron coupling, Croon and collaborators find that the lower edge of the BHMG could be raised up to 56 M⊙! Therefore, the observation of BHs within the traditional mass gap could be a sign of new physics, and conversely the absence of such observations could be used to constrain new physics. I really enjoyed reading this paper, which is really well written, although I do have one concern in that if a BH within the BHMG is observed, how do we know whether the lower edge of the BHMG not being as low as previously thought is attributable to new physics (such as an axion) rather than something more mundane like complicated (g)astrophysics or stellar physics? In any case, I look forward to reading more in the upcoming [10] by the same authors, which should explore many more new physics scenarios, and which from my understanding should come out very soon (if that is indeed the case one could almost be forgiven for thinking that [10] is basically ready and this week’s paper is an abridged version thereof only focusing on the axion…just in time for the XENON1T excess :) ).