Synthetic (SiGe)-pyroxenoids are often observed to have superperiods. Wheth
er or not these superperiods grow in connection with Si-Ge ordering is a fu
ndamental question. The size difference between Ce and Si tetrahedra leads
to CdGeO3 having the pyrox-mangite structure, whereas CdSiO3 has the wollas
tonite structure. Consequently, considerable strain is expected for a pyrox
enoid with a disordered Ge-Si distribution. A Si-29 MAS-NMR study of a Cd-p
yroxmangite of nominal composition Cd-7[Ge6Si]O-7 points to considerable Si
clustering, probably as GeSiSiSiGe triples corresponding to the wollastoni
te-like units of the siebener chain. It is proposed that such ordering reli
eves strain within the tetrahedral chain. Residual strain is also relieved
by insertion of an extra pair of GeO4 tetrahedra into the pyroxene-like com
ponent of the siebener-chain units, leading to single neuner-chain units wh
ich terminate the superperiods. This growth pattern is cyclic. The presence
of several types of superperiod may reflect similar energetics for differe
nt Ge-Si-ordering patterns within the siebener chains. Ordering of Si-rich
unit cells and of unit cells having no Si is proposed as the reason for the
occurrence of the superperiods in Cd-pyroxmangite.