Although
the majority of computational predictions of the properties
of diamond nanoparticles (nanodiamonds) are based on sets of exclusively
single-crystal structures, most experimental observations contain
twins. The influence of twins is difficult to characterize computationally
since they are metastable and their relative stability will depend
on the thermochemical conditions. The properties of entire samples
of nanodiamonds will also depend on how many twins there actually
are. In this study, we have used a combination of electronic structure
simulations, ab initio thermodynamics, and a simple
statistical method called ensemble filtering to investigate the probability
of twinning in nanodiamond and their impact on charge transfer properties
such as the ionization potential, the electron affinity, and the electronic
band gap. We find that, provided some degree of control can be exercised
over the surfaces, increasing the number of twinned particles in samples
could shift the selectivity of the electron affinity and the band
gap and could greatly improve the quality of samples by decreasing
the bandwidth, improving the sensitivity and specificity.