Stellar evolution is modified if energy is lost in a “ dark channel ” similar to neutrino emission .
Comparing modified stellar evolution sequences with observations provides some of the most restrictive limits on axions and other hypothetical low-mass particles and on non-standard neutrino properties .
In particular , a putative neutrino magnetic dipole moment \mu _ { \nu } enhances the plasmon decay process , postpones helium ignition in low-mass stars , and therefore extends the red-giant branch ( RGB ) in globular clusters ( GCs ) .
The brightness of the tip of the RGB ( TRGB ) remains the most sensitive probe for \mu _ { \nu } and we revisit this argument from a modern perspective .
Based on a large set of archival observations , we provide high-precision photometry for the Galactic GC M5 ( NGC 5904 ) and carefully determine its TRGB position .
On the theoretical side , we add the extra plasmon decay rate brought about by \mu _ { \nu } to the Princeton-Goddard-PUC ( PGPUC ) stellar evolution code .
Different sources of uncertainty are critically examined .
The main source of systematic uncertainty is the bolometric correction and the main statistical uncertainty derives from the distance modulus based on main-sequence fitting .
( Other measures of distance , e.g. , the brightness of RR Lyrae stars , are influenced by the energy loss that we wish to constrain . )
The statistical uncertainty of the TRGB position relative to the brightest RGB star is less important because the RGB is well populated .
We infer an absolute I -band brightness of M _ { I } = -4.17 \pm 0.13 mag for the TRGB compared with the theoretical prediction of -3.99 \pm 0.07 mag , in reasonable agreement with each other .
A significant brightness increase caused by neutrino dipole moments is constrained such that \mu _ { \nu } < 2.6 \times 10 ^ { -12 } \mu _ { B } ( 68 % CL ) , where \mu _ { B } \equiv e / 2 m _ { e } is the Bohr magneton , and \mu _ { \nu } < 4.5 \times 10 ^ { -12 } \mu _ { B } ( 95 % CL ) .
In these results , statistical and systematic errors have been combined in quadrature .