This thesis addresses the problem of a form of anomalous decoherence that sheds light into the spectroscopy of blinking quantum dots. The system studied is a two-state system, interacting with an external environment that has the effect of establishing an interaction between the two states, via a coherence generating coupling, called inphasing. The collisions with the environment produce also decoherence, named dephasing. Decoherence is interpreted as the entanglement of the coherent superposition of these two states with the environment. The joint action of inphasing and dephasing generates a Markov master equation statistically equivalent to a random walker jumping from one state to the other. This model can be used to describe intermittent fluorescence, as a sequence of "light on" and "light off" states. The experiments on blinking quantum dots indicate that the sojourn times are distributed with an inverse power law. Thus, a proposal to turn the model for Poisson fluorescence intermittency into a model for non-Poisson fluorescence intermittency is made. The collision-like interaction of the two-state system with the environment is assumed to takes place at random times rather than at regular times. The time distance between one collision and the next is given by a distribution, called the subordination distribution. If the subordination distribution is exponential, a sequence of collisions yielding no persistence is turned into a sequence of "light on" and "light off" states with significant persistence. If the subordination function is an inverse power law the sequel of "light on" and "light off" states becomes equivalent to the experimental sequences. Different conditions are considered, ranging from predominant inphasing to predominant dephasing. When dephasing is predominant the sequel of "light on" and "light off" states in the time asymptotic limit becomes an inverse power law. If the predominant dephasing involves a time scale much larger than the minimum time scale accessible to the experimental observation, thereby generating persistence, the resulting distribution becomes a Mittag-Leffler function. If dephasing is predominant, in addition to the inverse power law distribution of "light off" and "light on" time duration, a strong correlation between "light on" and "light off" state is predicted.

The dynamical origin of complexity is an object of intense debate and, up to moment of writing this manuscript, no unified approach exists as to how it should be properly addressed. This research work adopts the perspective of complexity as characterized by the emergence of non-Poisson renewal processes. In particular I introduce two new complex system models, namely the two-state stochastic clocks and the integrate-and-fire stochastic neurons, and investigate its coupled dynamics in different network topologies. Based on the foundations of renewal theory, I show how complexity, as manifested by the occurrence of non-exponential distribution of events, emerges from the interaction of the units of the system. Conclusion is made on the work's applicability to explaining the dynamics of blinking nanocrystals, neuron interaction in the human brain, and synchronization processes in complex networks.