The influence of short-range correlations (SRC) on the spectral distribution of neutrons is incorporated in the solution of the gap equation for the ${}^3P_2-{}^3F_2$ coupled channel in pure neutron matter at high density. This effect is studied for three different realistic interactions. The gap in this channel is strongly suppressed by these correlations but does not vanish. For a consistent treatment we also include for the first time the effect of long-range correlations (LRC) by incorporating polarization terms in addition to the bare interaction. This allows the neutrons to exchange density and spin fluctuations governed by the strength of Landau parameters with values that are consistent with the available literature. While these LRC have an antiscreening tendency, they only slightly increase the gap in the ${}^3P_2-{}^3F_2$ coupled channel for all three realistic interactions as long as SRC are included. All three interactions generate maximum gaps around 0.1 to 0.2 MeV at most with a small dependence on the hardness of the interaction. These results are relevant for the cooling scenarios of neutron stars, in particular the young neutron star in Cassiopeia A.

Fusion reactions in the crust of an accreting neutron star are an important source of heat, and the depth at which these reactions occur is important for determining the temperature profile of the star. Fusion reactions depend strongly on the nuclear charge $Z$. Nuclei with $Z\le 6$ can fuse at low densities in a liquid ocean. However, nuclei with Z=8 or 10 may not burn until higher densities where the crust is solid and electron capture has made the nuclei neutron rich. We calculate the $S$ factor for fusion reactions of neutron rich nuclei including $^{24}$O + $^{24}$O and $^{28}$Ne + $^{28}$Ne. We use a simple barrier penetration model. The $S$ factor could be further enhanced by dynamical effects involving the neutron rich skin. This possible enhancement in $S$ should be studied in the laboratory with neutron rich radioactive beams. We model the structure of the crust with molecular dynamics simulations. We find that the crust of accreting neutron stars may contain micro-crystals or regions of phase separation. Nevertheless, the screening factors that we determine for the enhancement of the rate of thermonuclear reactions are insensitive to these features. Finally, we calculate the rate of thermonuclear $^{24}$O + $^{24}$O fusion and find that $^{24}$O should burn at densities near $10^{11}$ g/cm$^3$. The energy released from this and similar reactions may be important for the temperature profile of the star.