Direct collapse black hole formation from synchronized pairs of atomic cooling halos

High-redshift quasar observations imply that supermassive black holes (SMBHs) larger than $\sim 10^9 ~ M_\odot$ formed before $z=6$. That such large SMBHs formed so early in the Universe remains an open theoretical problem. One possibility is that gas in atomic cooling halos exposed to strong Lyman-Werner (LW) radiation forms $10^4-10^6 ~ M_\odot$ supermassive stars which quickly collapse into black holes. We propose a scenario for direct collapse black hole (DCBH) formation based on synchronized pairs of pristine atomic cooling halos. We consider halos at very small separation with one halo being a subhalo of the other. The first halo to surpass the atomic cooling threshold forms stars. Soon after these stars are formed, the other halo reaches the cooling threshold and due to its small distance from the newly formed galaxy, is exposed to the critical LW intensity required to form a DCBH. The main advantage of this scenario is that synchronization can potentially prevent photoevaporation and metal pollution in DCBH-forming halos. Since the halos reach the atomic cooling threshold at nearly the same time, the DCBH-forming halo is only exposed to ionizing radiation for a brief period. Tight synchronization could allow the DCBH to form before stars in the nearby galaxy reach the end of their lives and generate supernovae winds. We use N-body simulations to estimate the abundance of DCBHs formed in this way. The largest source of uncertainty in our estimate is the initial mass function (IMF) of metal free stars formed in atomic cooling halos. We find that even for tight synchronization, the density of DCBHs formed in this scenario could explain the SMBHs implied by $z=6$ quasar observations. Metal pollution and photoevaporation could potentially reduce the abundance of DCBHs below that required to explain the observations in other models that rely on a high LW flux.