RUBY IN DIAMOND By Henry 0.A, Meyer and Edward Gubelin The first substantiated identification of corundum It was also observed during the early micro- (var. ruby) occurring as an inclusion in natural probe analyses conducted by Meyer and Boyd diamond is presented. The ruby is assigned to the (1968, 1969, 19721, and described by Sobolev (1974) eclogitic suite of inclusions in diamond, and the and by Prinz et al. (1975), that not only did the implications of its occurrence are discussed in garnets comprise two distinct groups, but most relation to the genesis of "eclogitic" diar~londs.It is inclusions in diamond could be assigned to one concluded that diamond crystallizes from a melt over a long period of time with possible fluctuations of two suites as well: the ultramafic * or the eclo- in ambient temperature and geochemical gitic*. The members of these suites are listed in environment. table 1. Reviews of the minerals in these suites are to be found in Meyer and Tsai (1976))Gubelin et al. (19781, and Harris and Gurney (1979). It should be noted that ruby does not appear on Inclusions in gemstones can be fascinating for either list of inclusions in table 1. scientist and layman alike. Scientifically, they aid in deciphering the genesis of the mineral; gemo- RUBY INCLUSION IN DIAMOND logically, they are often distinctive and aid in During endeavors to find diamonds with unusual, identifying the host stone (Gubelin, 1953, 1974). colored mineral inclusions, one of the authors re- In 1645, John Evelyn, a diarist, reported seeing ceived two diamonds of unknown origin with red a "faire Rubie" inside a diamond that belonged to crystal inclusions that, on the basis of previous a Venetian nobleman (DeBeer, 1955). This iden- knowledge, were first tentatively identified as tification, although historically interesting as one chrome-pyrope garnets with a low CrgOycontent. of the first recorded observations of a mineral in- However, the difference in hue between the two cluded in a gemstone, has usually been consid- inclusions, one of which was claret red while the ered incorrect. Harris (1968) suggested that the other was a typical "ruby" red, was so great that inclusion to which Evelyn referred was probably additional research was pursued. a garnet, since garnets have now long been recog- The claret-red inclusion, which we freed from nized as inclusions in diamonds [see Futergend- its host diamond by breaking the latter, proved on ler, 1950, and Futergendler and Frank-Kament- analysis to be a chrome-rich pyrope garnet indeed. sky, 1961). Early in 1967, Meyer (1967) used the However, analysis with the electron microprobe electron microprobe to obtain the first chemical showed that the other inclusion, which was ex- analyses of mineral inclusions in diamond and posed at the surface of one crown face of the 0.06- discovered that two chemically distinct types of ct. brilliant-cut diamond (see figure 1)) showed garnet were present. One type consists primarily of magnesium- and chromium-rich, calcium-poor pyropes; whereas the second type is calcium and iron rich, and has virtually no chromium. This ABOUT THE AUTHORS latter group of garnets is made up of almandine Dr. Meyer is professor of mineralogy in the Department of pyropes; its members are orange-red, whereas the Geosciences, Purdue University, West Lafayette, Indiana; Dr. chromium-rich garnets (chrome-pyropes*, Meyer, GObelin is a certified gemologist in Meggen, Switzerland, and 1968) of the first group are claret colored. Similar honorary professor at the University of Slellenbosch, South Africa. results were obtained by later studies (e.g., So- Acknowledgments: The authors wish to extend their gratitude bolev and Lavrentlyev, 1969; Prinz et al., 1975; to Mr. F. Veraguth of Messrs. Voegeli and Wirz, Bienne, Meyer and Svisero, 1975). Switzerland, for the generous gift of the two diamonds containing the red inclusions discussed in this paper. Dr. Meyer acknowledges support from the National Science Foundation, grant EAR 76-22698. 'Asterisk denotes terms defined in the glossary included with this article, "7 981 Gemological Institute of America Notes and New Techniques GEMS & GEMOLOGY Fall 1981 153 TABLE 1. Minerals of the ultramafic and eclogitic inclusion suites. - -- Suite Mineral Chemical compositiona Ultramafic Olivine Enstatite Diopside Cr-pyrope garnet Mg-chromite Mg-ilrnenite Phlogopi te Sulfides Eclogitic Almandine-pyrope garnet 'Ornphacitic" clinopyroxene Coesite Rutile Kyanite Fe-chrornite Biotite Sulfides "The order of the elements indicates the order of importance. Thus, in olivine, Mg is more important than Fe. that it was in fact corundum (var. ruby). The an- or corundum. Cell dimensions were a. = 4.76 A, alytical data for this inclusion, together with the Co = 13.22 A, compared to 4.758 A and 12.991 A, data for a gem ruby from Burma, are presented in respectively, for a synthetic corundum. table 2. From the X-ray photographs it was observed Single-crystal X-ray diffraction studies further that the crystal orientation of the ruby did not substantiated this identification of ruby as an in- line up precisely with that of the host diamond clusion in diamond. The diffraction pattern of the as it would in an epitaxial relationship. However, inclusion was identical to that of a reference mby several crystallographic zones in the diamond were I GLOSSARY OF SPECIAL TERMS Chrome-mope Ultramafic A garnet that is composed mostly of pyrope A term used to describe minerals that have a low ,MggAlgSiaOiJ in which some of the A1 is re- silica content but are usually rich in magnesium, placed by Cr. The end member with no A1 and all iron, and possibly alumina. Rocks in which these Cr (MgaCr4igOta)is referred to as knorringite, minerals occur (ultramafic rocks) are thought to be common in the upper mantle of the earth, i.e., Eclogite deeper than 35 km. Peridotite, consisting of about A rock consisting predominantly erf two rnin- 90% olivine, is a good example of an ultramafic erals: a reddish iron-magnesium-calciumgarnet rock. and a pale green sodium-calcium rich pyroxene Xenolith referred to as omphacite. Eclogites are believed A fragment of rock that is foreign to its host rock. to occur at depths in the earth, but they are not Thus, in kimberlite one often finds fragments of as common as the ultramafic rocks. rocks from the deeper parts of the earth that are chemically and mineralogically unrelated to Grospydite kimberlite and were incorporated into the host A rock consisting of the minerals grossular gar- kimberlite during its ascent from a depth of about net, pyroxene, and disthene (kyanite). 150 km. 154 Notes and New Techniques GEMS & GEMOLOGY Fall 198 1 that in non-diamondiferous eclogite from the Ob- nazhennaya lzimberlite is violet-pink. The pink- ish color is due to the presence of chromium, as is the case of the corundum inclusion reported here (table 2). We have, therefore, assigned this ruby to the eclogitic suite of mineral inclusions in diamond. --1interestingly, it is usual for the inclusions in the ultramafic group to show enrichment in chro- mium; this ruby is the first member of the eclo- gitic suite to have this feature. The fact that ruby has not previously been recognized in diamond is probably a consequence of its rarity as an inclu- 1sion.- Evidence for this rarity is perhaps indicated Figure 1. Ruby crystal included in a 0.06-ct. by the rareness with which corundum eclogites brilliant-cut diamond, occur in lzimberlite. If we confine the discussion to diamonds for close to those of the inclusion, and possibly at the which genesis is associated with the chemical en- high temperatures and pressure in which the ruby- vironment in which eclogites are formed, the fol- diamond couple formed, the crystallographic ori- lowing conclusions are suggested. entation of the two minerals was better. The in- First, the generally monomineralic nature of clusion is somewhat elongated and has rounded the inclusions, their lack of impurities and chem- faces; the elongation may be subparallel to the ical zoning, and their smallness (usually less than c-crystallographic direction. 400 pm) suggest that the diamonds crystallized The small size of the ruby inclusion and the from a melt. fact that it is partly contained in its host stone Second, the similarity in chemical composi- made determination of the ultraviolet fluores- tion of garnet and pyroxene inclusions with those cence and absorption spectrum difficult. Never- of the host eclogite in which the diamond is theless, in spite of the yellowish fluorescence dis- embedded (Sobolev et al., 1972) suggests growth played by the host diamond, a faint reddish equivalent in time at least to that of the major fluorescence was obvious from the inclusion. Fur- mineral phases (garnet and pyroxene) in the thermore, a very faint absorption line was per- eclogite. ceived at 6600 A, and a slightly stronger one at Third, the zonal stratigraphy revealed by etch- 6935 A. Admittedly, the absorption data are sub- ing diamond, as well as the marked variation in ject to some error, but they are compatible with the microprobe and X-ray determinations that the inclusion is a ruby. TABLE 2. Analysis of corundum (var. ruby) inclusion in diamond compared to a ruby from Burma. DISCUSSION - -- This is the first substantiated identification of % present in the % present in the Oxide inclusion in diamond Burmese ruby corundum (var. ruby) as an inclusion in natural diamond. The occurrence of this mineral as an inclusion is not unexpected, since both ruby and sapphire have been reported in South African lzimberlite (Harger, 191 11, and corundum-bearing eclogite and grospydite xenoliths have been MgO found in lzimberlites in Siberia and Africa (So- CaO bolev, 1964; Sobolev and Kuznetsova, 1965; Rick- MnO wood et al., 1968).