On the feasibility of production of microkelvin temperature by nuclear demagnetisation