The present route to grain boundary engineering (GBE) is usually based on multiple annealing twinning which can only be applied to a certain subset of materials, namely those that twin prolifically. A more general approach has been highlighted recently, following experimental evidence that certain boundary planes in iron bicrystals are 'special', and that this classification is not based on misorientation. It was suggested that, under suitable conditions, individual interfaces could reorient the most energetically advantageous orientations. This approach concurs with a similar concept of 'grain boundary plane engineering', proposed previously. In the present article we explore this concept and report the effect of long duration, low temperature annealing on the distribution of boundary misorientation and planes in copper. The new findings give support to the possibility of grain boundary structure optimisation via controlled annealing. To have established that grain boundary plane reorientation is feasible opens up new avenues and challenges in the field of grain boundary research. This could have significant impact both scientifically in terms of understanding grain boundary structure and technologically in the field of GBE.