Synthesis of pigeonites for spectroscopic studies

Introduction

Pigeonite, i.e., clinopyroxene containing approximately 10% Wo [where the amount of Wo (at%) = 100Ca/(Ca+Mg+Fe²⁺)], is reported from terrestrial, lunar, and planetary samples (e.g., Basaltic Volcanism Study Project 1981, p. 65, 184, and 224). It forms in various igneous and high-temperature metamorphic rocks but survives as a discrete phase almost exclusively in rapidly cooled, fine-grained lavas. This is because pigeonite has a minimum thermal stability temperature below which, upon slow cooling, it breaks down (“inverts”) to an intergrowth of augite lamellae in an orthopyroxene host—the so-called “inverted pigeonite” texture. Since pigeonite does not survive as a discrete phase in plutonic rocks, and it is tedious to separate pigeonite grains from the fine-grained lavas in which they do survive, we lack natural pigeonite for measurements that require bulk samples. To remedy this, we have synthesized eight samples of pigeonite to serve as standards for spectroscopic and other studies.

Synthesis of pigeonite is complicated by several characteristics of phase equilibria in the pyroxene quadrilateral. Pigeonite synthesis must take place within its stability field, that is, above the temperature of its minimum thermal stability (e.g., Fig. 6 of Davidson and Lindsley 1985) but below the solidus temperature in the pure Ca₂Si₂O₆-Mg₂Si₂O₆-Fe₂Si₂O₆ system (Fig. 4b of Huebner and Turnock 1980; our Fig. 1). Thus, for a given X [X = 100Fe²⁺/(Fe²⁺+Mg) molar], there is a relatively narrow temperature window for successful synthesis. Furthermore, as both the minimum pigeonite stability and solidus temperatures decrease with increasing X, this window shifts with composition. Much early work concentrated on making pigeonite in the Fe-free (Kushiro 1969; Yang 1973; Schweitzer 1982; Tribaudino et al. 2002) and Mg-free (Lindsley 1981) joins. Turnock et al. (1973) reported the synthesis of pigeonite at a wide variety of X and listed several techniques for synthesis. They used their Method 1 (controlled-gas, quench furnace, 1 atm) to make pigeonites with X ranging from 20 to 60, and their Method 5 (hydrothermal synthesis at 2 GPa) for pigeonite with X = 75 (because pigeonite of that composition lies in the “Forbidden Zone” and is not stable at 1 bar, e.g., Lindsley 1983). Neither method, however, is well suited for making gram-quantities of pigeonite, so we used a modification of their Method 2: synthesis in sealed, evacuated silica-glass tubes, to allow us to process 1 to 3 g of material in each tube. We aimed to make ~5 g of each composition, with Wo values of 8 and 10, and X ranging from 20 (few natural pigeonites are more magnesian than this) to 60 (near the edge of the “Forbidden Zone” and thus near the limit for synthesis at low pressure).

Figure 1

Solidus diagram for Ca-Mg-Fe²⁺ pyroxenes at 1 atm (adopted from Huebner and Turnock 1980; their Fig. 4b), showing the approximate pigeonite field and the target compositions for our work (open circles). Wo (mol%) = Ca₂Si₂O₆, or 100Ca/(Ca+Mg+Fe²⁺); X = 100Fe²⁺/(Mg+Fe²⁺) (at%); En = Mg₂Si₂O₆; Fs = Fe₂Si₂O₆; Di = CaMgSi₂O₆; Hd = CaFeSi₂O₆; Opx = orthopyroxene; Ol = olivine; S = silica phase (tridymite, cristobalite, or quartz). We omit labels for the complex region near Hd and Fs. Note that olivine has higher X than the coexisting pigeonite, as shown by the composition of olivine coexisting with pigeonite at the solidus.

Synthesis details

Final synthesis

Pre-reacted materials were packed into iron capsules machined to 2 to 4 cm in length with snugly fitting lids from medium-purity (99.6 to 99.95%)¼ inch (6.35 mm) OD rod. The loaded capsules were placed in 7 mm ID silica-glass tubes, dried under vacuum at 800 °C for 10 min using an iron “oxygen getter” (Fe⁰ sponge) placed in the tubes at ~600 °C to prevent oxidation, and the tubes were then sealed under vacuum. We used two heating approaches for the synthesis. The first involved holding the sample at a fixed temperature below the solidus for direct synthesis in the solid state. The second approach involved partially melting the sample at ~25–50 °C above the solidus temperature, then cooling it slowly to below the solidus temperature, and finally holding the sample at that lower temperature, with the goal of using the presence of melt to speed up the reaction and to grow larger crystals. Unfortunately, most samples treated using this second approach were strongly zoned, with low-Ca, high-Mg cores mantled by rims close to the target composition. Only one sample treated in this way (Wo₁₀X = 60 T) was sufficiently homogeneous to be acceptable (Fig. 2). Samples are named by the target composition plus a letter (A, B, etc.) designating the sequence of that sample in the synthesis attempts for that composition.

Figure 2

Portion of the pyroxene quadrilateral showing target compositions (open circles) and microprobe analyses (solid points; Supplemental^[1] Table S2) of the samples. All analyses with close to (Mg+Fe+Ca)₂Si₂O₆ stoichiometry [Si = 1.98 to 2.02; Ca+Mg+Fe(+Mn) = 1.98–2.02] are plotted. MnO (generally <O.6 and more commonly <0.3 wt%) was added to FeO to form Fs component. Black symbols and text refer to Wo₁₀ samples; gray to Wo₈. We deliberately searched for and report outlier compositions to show diversity of grain compositions. For this reason the bulk compositions of the samples are best represented by the clusters of analyses near the target compositions, not an average of the points shown. See text for interpretation of outlier points. Components as in caption to Figure 1.

All but one sample (Wo₈X = 40 E) underwent more than one stage of “final” synthesis: the products of previous attempts were re-heated (usually with intermediate grinding) to produce more nearly homogeneous materials. Details are in Supplemental^[1] Table S1. Especially for more magnesian compositions that required higher temperatures, run durations were nerve-wracking compromises between the desire to maximize reaction time and the knowledge that the enclosing silica-glass tubes would eventually fail because of softening and/or devitrification. Because of the recycling and associated losses of material (when the silica-glass tubes failed and samples within them became oxidized), the samples reported here fall short of the 5 g target by varying amounts.

Implications

Because bulk samples of natural pigeonites are not available, synthetic samples are needed for various studies. However, synthesis of gram-quantities of pigeonite remains difficult. The eight samples reported here are products of the most successful of more than 125 synthesis attempts over several years. Potential users of these samples should be aware of the within-sample variations; the samples should be used for studies on bulk samples. As any individual grain may deviate considerably from the nominal composition, we do not recommend using individual crystals unless each crystal has been analyzed by microprobe.

Those who wish to obtain samples for specific studies should contact T.D. Glotch, (timothy.glotch@stonybrook.edu), stating the composition(s) desired, which measurements will be made; optimum and minimum quantities needed; and whether the sample(s) can be returned undamaged and uncontaminated after use. Requesters should take into account: (1) the widely varying amounts available for each sample, (2) that some samples are more nearly uniform in composition than others, and (3) that there are variations in grain size. Preference will be given to those who will use nondestructive techniques. Please note that the sample names include both the target compositions and the letter (e.g., A, B, C) that uniquely designates the particular sample. (For example, the sample name Wo₁₀X = 60 T implies that there are at least 19 other samples (A…S) of that target composition! The T is essential for identifying that sample.) The letter designator must be retained in all reports that use the sample.