The normal prograde diagenetic and low-grade metamorphic sequence of dioctahedral clay minerals including illite-rich I-S and illite, as observed by TEM, proceeds from a partially disordered 1Md stacking sequence to 2M1; i.e. 1M does not normally occur as an intermediate polytype. Examples of 1M illite stacking sequences have been studied, however, from the Golden Cross gold deposit, New Zealand, the Broadlands-Ohaaki geothermal system, New Zealand, the Potsdam Sandstone, New York, and the Silverton Caldera, Colorado. Specific clay-mineral packets identified by TEM techniques as 1M illite were found to have anomalously high Mg contents. The Broadlands illite provides the most definitive data, as separate packets of 1M and 2M1 illite coexist. Average compositions for 1M and 2M1 illite are (K1.66 Ca0.04)∑1.70 (Al3.32 Fe0.31 Mg0.57 Mn0.06)∑ 4.26 (Si6.43 Al1.57) ∑8 O20 (OH)4 and (K1.57 Na0.31 Ca0.03)∑1.91 (Al3.58 Fe0.05 Mg0.29 Mn0.01)∑3.93 (Si6.70 Al1.30)∑8 O20 (OH)4, respectively. In addition, 1Md illite, which is the polytype occurring in the common 1Md to 2M1 prograde sequence, is relatively Mg poor, but coexists with Mg-rich illite in the Silverton Caldera sample.
These data confirm that 1M stacking is caused by compositional anomalies, and thus explain the lack of the 1M stacking sequence in normal diagenetic sequences in pelitic rocks, as most illite in such environments has a relatively small phengitic component. The parameter Δz, a measure of the corrugation of the oxygen sheets, may be the key parameter reflecting the polytypic state of dioctahedral and trioctahedral micaceous minerals. Such composition-determined relations may be related to the occurrence of 1M polytypism in glauconite and celadonite, both dioctahedral 2:1 clay minerals having large Mg or Fe octahedral-cation components, and in trioctahedral micas. Insofar as the 1M stacking sequence does not have the same composition as 2M1 material, these data confirm that the different varieties of illite are not polytypes, sensu stricto.