The objective of this paper is to describe the new serpentine group mineral, guidottiite, which is analogous to cronstedtite. Guidottiite has an ideal chemical composition of (Mn2Fe3+)(SiFe3+)O5(OH)4. The sample is from the N’chwaning 2 mine, Kalahari manganese field, Republic of South Africa, and apparently forms from hydrothermal solutions. Grains are optically near opaque [average index of refraction 1.765, with variable extinction on the (001)], vitreous, and black, with perfect {001} platy cleavage. A non-separable fibrous substructure exists perpendicular to cleavage that results in a silky luster under optical examination. The average chemical analysis determined from electron microprobe based on four grains with ten analyses each resulted in a structural formula of (Mn1.86Fe0.613+Mg0.54)Σ=3.01(Si1.36Fe0.643+)Σ=2.00O5(OH)4, with calculated density of 3.236 g/cm3. Analysis from another area of the sample showed a slightly different chemical composition and resulted in a formula of (Mn1.70Fe0.963+Mg0.24Σ=2.89(Si1.26Fe0.743+)Σ=2.00O5(OH)4, with calculated density of 3.291 g/cm3. The measured density on a bulk sample (with impurities) was 3.33 g/cm3. Thermal analysis suggested a dehydroxylation temperature of 535°C, a decomposition/recrystallization temperature of 722°C, and weight loss (= H2O loss) of 9.4%. The derived Mohs hardness from nano-indentation is H = 4.25.
The sample is mostly the 2H1 polytype with minor amounts of the 2H2 polytype. Using a predominantly 2H2 crystal, which has better crystallinity, the strongest observed X-ray peaks are: 7.21 Å (Io/Io = 100%), 3.543 (50), 2.568 (39), 1.982 (26), and 2.381 (25). All Gandolfi simulations, even with three crystal remountings, showed preferred orientation effects. Transmission electron microscope (TEM) analysis showed stacking disorder within Group D serpentine polytypes. Thus, a regular alternation of the occupancy of octahedral sets within each layer along the stacking exists, but disorder of the layer displacements of 0 and ±b/3 (b defined here as the orthohexagonal cell) exists. Ordered 2H1 (no layer displacement) and 2H2 (alternating + and —b/3 displacement) domains were also frequently observed. X-ray diffraction analysis showed that even apparent single crystals contain impurity phases, presumably Mn-rich and Ca phases that were detected in the microprobe study. The single-crystal structure refinement used a well (stacking) ordered apparent 2H2 crystal with little to no streaking in the diffraction pattern. Results showed that the crystal has a random interstratification of 2H2 and 2H1. The 2H2 polytype is hexagonal, space group P63, with a = 5.5472(3), c = 14.293(2) Å, and Z = 2, and was refined to R1 = 0.072 and wR = 0.108 from 656 unique reflections. Because the two polytypes in the composite have only small differences in the lower 1:1 layer, a large displacement parameter for the basal oxygen atom results, which was constrained to B = 1.5 Å2 (Ueq = 0.0190) in the refinement. Half of the tetrahedral sites in the 2H1 upper layer superpose over half of the tetrahedral sites in the 2H2 upper layer (T1 sites only) per unit cell. This superposition produces an apparent excess of electron densities of the T1 site relative to the T2 site (T1 = 21.9 electrons, T2 = 15.8). Comparison with the microprobe data indicates that observed tetrahedral bond lengths are generally not affected by this intergrowth. Tetrahedral bond lengths indicated that the tetrahedral sites contain T1 = Si0.678 Fe0.3223+ and T2 = Si0.631Fe0.3693+. This excess of electron densities and other refinement problems associated with the guidottiite single-crystal refinement closely parallel all single-crystal cronstedtite-2H2 refinements to date, suggesting that these refinements also involve random interstratifications of 2H2 and 2H1 polytypes.