We describe the fate of incident C-12, N-14, and O-16 in accreting neutron star atmospheres. When the accreting material is stopped by Coulomb collisions with atmospheric electrons, all incoming elements heavier than helium thermalize at higher altitudes in the atmosphere than the accreting protons. The incoming protons and helium then destroy the elements via nuclear spallation reactions. A small fraction of the nuclear reactions cause nuclear excitation and subsequent gamma-ray emission. The probability for a nucleus to survive this bombardment depends on how long it spends in the hazardous region of the atmosphere. For typical accretion rates, the nucleus resides in the hazardous region for many destruction times, and therefore has a small survival probability. We calculate the fractions of incident C-12, N-14, and O-16 that survive proton bombardment as a function of the accretion rate, and the mass and radius of the neutron star. The subsequent paucity of CNO nuclei decreases hydrogen-burning rates in the deep regions of the atmosphere, thereby reducing the amount of helium available for the unstable nuclear flashes that cause type I X-ray bursts. X-ray bursts still occur, but they are predominantly of the mixed hydrogen helium type first outlined by Taam & Picklum. We also determined the gamma-ray line emission from this collisional deceleration scenario. For accretion of material with solar abundances of CNO elements, the flux in the 4.438 and 6.129 MeV gamma-ray lines is always far below the observational limits of the Gamma-Ray Observatory (GRO). We outline other gamma-ray emission processes that might be observable to GRO. Many previously unaddressed properties of an accreting neutron star atmosphere prove important when solving the heavy-element diffusion problem. Specifically, we show that the incident electrons are crucial in determining the macroscopic electric field present in the atmosphere. We briefly discuss how our results are modified in the high magnetic field environments (B greater-than-or-similar-to 10(12) G) of X-ray pulsars. Alternative stopping scenarios such as collisionless shocks and disk accretion are also discussed.
