Ab-initio electronic structure methods are overwhelmingly focused on ground-state electronic behavior. Multiple computationally tractable approaches have been developed to high degrees of refinement for the determination of electronic behavior at T=0 K, such as density functional, coupled cluster, and quantum Monte Carlo methods. There are, however, important systems whose behavior depends intrinsically on electronic excitation, such as photovoltaic materials, photochemical processes, warm-dense matter, and planetary interiors. Those require accurate electronic structure approaches not tied to the ground state. Available methods range from treatment of a handful of user-selected states to ensemble-averages. Pure-state descriptions maintain coherence information for selected states. Progress has been made on ensembles for excited states but thermal ensembles must include myriad thermally populated many-body states at the cost of sacrificing coherence information. Out-of-equilibrium methods face both challenges.

The two ends of this methodological spectrum typically are addressed by different communities. Pure states usually are the domain of quantum chemistry. It has developed both deterministic techniques, e.g. equation-of-motion coupled cluster [1,2] and linear-response coupled cluster [3], as well as various stochastic techniques, including excited state quantum Monte Carlo (QMC) [4], full configuration interaction QMC [5], and auxiliary-field QMC [6]. At the other end, the warm-dense matter community (mostly physics) has driven ensemble-average methods, that also are roughly categorizable as deterministic and stochastic. Examples for the former are thermal density functional theory [7], finite-temperature coupled cluster [8-10], and finite-temperature Green's function methods [11]. The stochastic techniques include density matrix QMC [12], various path integral Monte Carlo methods [13,14], auxiliary field QMC [15], and finite-temperature full configuration interaction QMC [16].

The goal of this CECAM workshop is to bring these communities into closer, more productive contact. The two speciality communities face many common challenges but are separated by vocabulary, funding, and methodological heritages. An example is the finite-size behavior of excited states. Conversely, the methods may be more broadly applicable than originally conceived. For instance, the coupled-cluster method, originally for pure states, recently was generalized to thermal ensembles [8-10]. This workshop also aims at identifying new potential applications and open physical and methodological questions across the two specialities. For example, do the metal-insulator and structural transitions coincide in the hydrogen liquid/liquid phase transition? Note that though attosecond time-dependent DFT has some overlap with the goals of this workshop, for the sake of focus, we must put it outside the workshop scope.