Can deviation from maximal Θ23 be resolved in DUNE?

The present global analyses of the oscillation data do not agree on the octant in which the best-fit value of Θ23 lies. Further, they allow sin2 Θ23 = 0.5 at 3σ confidence level. Hence, it is imperative to question at what confidence level maximal 2-3 mixing can be ruled out. We study in detail the performance of DUNE to establish the deviation from maximal Θ23. We also discuss the impact of sin2 Θ23 - Δ m231degeneracy in establishing non-maximal Θ23 and show how this degeneracy can be resolved by exploiting the spectral shape information in neutrino and antineutrino disappearance events. We find that a 3σ (5σ) determination of non-maximal Θ23 is possible in DUNE with an exposure of 336 kt·MW·years if the true value of sin2 Θ23 < 0.465 (0.450) or sin2 Θ23 > 0.554 (0.572) for any value of true δCP and true NMO. We also study the individual contributions from appearance and disappearance channels, the impact of systematic uncertainties and marginalization over oscillation parameters, the importance of spectral analysis, and data from both neutrino and antineutrino runs, while analyzing DUNE's sensitivity for the discovery of a non-maximal Θ23.