Performance of a feedback-controlled, deterministic source of single chromium atoms

Abstract

A magneto-optical trap with feedback-controlled loading and loss rates is used to realize a deterministic source of single chromium atoms with single-atom occupation probability as high as 0.987±0.001. We present a series of measurements of the performance of this source and discuss the dependence of the probability of not having a single atom in the trap (error rate) on experimental parameters. We describe a simple analytical model that considers mean load rate, trap lifetime, stray load rate, and feedback response time, and we also present results of Monte Carlo calculations that take into account all experimental conditions. We find that the analytical model describes the behavior well for error rates as small as about 0.03, but the Monte Carlo simulations must be used to model behavior at error rates lower than this, as the occupation probability approaches unity.