The spin temperature of neutral hydrogen, which determines the optical depth and brightness of the 21 cm line, is determined by the competition between radiative and collisional processes. Here we examine the role of proton-hydrogen collisions in setting the spin temperature. We use recent fully quantum mechanical calculations of the relevant cross sections, which allow us to present accurate results over the entire physically relevant temperature range 1-10,000 K. For kinetic temperatures T_K>100 K, the proton-hydrogen rate coefficient exceeds that for hydrogen-hydrogen collisions by about a factor of two. However, at low temperatures (T_K < 5 K) H-p collisions become several thousand times more efficient than H-H and even more important than H-e^- collisions.

The spin temperature of neutral hydrogen, which determines the 21 cm optical depth and brightness temperature, is set by the competition between radiative and collisional processes. In the high-redshift intergalactic medium, the dominant collisions are typically those between hydrogen atoms. However, collisions with electrons couple much more efficiently to the spin state of hydrogen than do collisions with other hydrogen atoms and thus become important once the ionized fraction exceeds ~1%. Here we compute the rate at which electron-hydrogen collisions change the hydrogen spin. Previous calculations included only S-wave scattering and ignored resonances near the n=2 threshold. We provide accurate results, including all partial wave terms through the F-wave, for the de-excitation rate at temperatures T_K < 15,000 K; beyond that point, excitation to n>=2 hydrogen levels becomes significant. Accurate electron-hydrogen collision rates at higher temperatures are not necessary, because collisional excitation in this regime inevitably produces Lyman-alpha photons, which in turn dominate spin exchange when T_K > 6200 K even in the absence of radiative sources. Our rates differ from previous calculations by several percent over the temperature range of interest. We also consider some simple astrophysical examples where our spin de-excitation rates are useful.