COBALT GLASS AS A LAPIS LAZULI IMITATION By George Bosshart A ~lecl<laceof round beads offered as "blue quartz from India" was analyzed by gemological and addition~~l advanced techniques. The violet-blue ornamental material, which resembled fine-q~ralitylapis lazuli, turned OLJ~LO be a nontransparent cobalt glass, unlil<e any glass observed before as a gem substit~zte.The characteristic color irregularities of lapis (whjtein blue) had been imjtated by white crystdlites of low- crjstobalite .ir~cludeclin the deep blue glass. The gemological world is accustomed to seeing gemstones from new localities, as well as new or improved synthetic crystals. With this in mind, it is not surprising that novel gem imitations are also encountered. One recent example is 'lopalitellla 400 500 600 700 convincing yet inexpensive plastic imitation of Wavelength A (nm) white opal manufactured in Japan. This article de- scribes another gem substitute that recently ap- Figure 1. Absorptio~ispectrum of a cobalt glass peared in the inarlzetplace. imitating lapis lazuli recorded ~hrougha chjp of Hearing of an "intense blue quartz from India1' approximately 2.44 inm thickness in the range of was intriguing enough to arouse the author's sus- 820 nm 10 300 nm, ot room temperature (Pye picion when a neclzlace of spherical opaque Ui7jcam SP8-100 Spectrophotometer). violet-blue 8-mm beads was submitted to the SSEF laboratory for identification. Because blue quartz in nature is normally gray-blue as a result of the refractive index of the tested material (1.508)does presence of Ti02(Deer et al.! 1975! p. 2071 or tour- not differ marlzedly from that of lapis (approxi- maline fibers (Stalder) 1967)! this particular iden- mately 1.50)/ its specific gravity of 2.453 is tification could be immediately rejected. Al- significantly lower than the average for lapis (ap- though synthetic cobalt-colored quartz exists, proximately 2.80). The absorption spectrum [fig- thus far it has been produced only in a transparent ure 1)differs from that of a blue-filter glass only by form. The beads of the neclzlace we examined re- its slightly stroEger iron pealzs and by a shift in the sembled more closely a fine lapis lazuli! with the ultraviolet absorption edge from approximately characteristic color irregularities of lapis! yet they displayed a tinge of violet exceeding that of top- quality lapis and they contained no pyrite grains. ABOUT THE AUTHOR Accordinglyl a series of gemological and other Mr. Bossharl, a mineralogist and gemolog;st, is laboratory direc- tests were conducted to determine the precise na- tor of the Swiss Foundation for the Research of Gemstones ture of this unusual material. (SSEF), Zurich. Acknowledgments: The author wishes to /hank Dr, W. B. Stern, of RESULTS OF GEMOLOGICAL TESTING lhe Geochemical Laboratory of /he University of Easel, for the The properties compiled in table 1 clearly indicate XRF-EDS data; and Dr. H. A. Hanni, also of the Geochemical Laboratory of the University of Easel and of the SSEF Zurich, /or that the material is not lapis lazuli or any other providing the X-ray diffraction determination in addition to his natural material! but rather a man-made cobalt- helpful suggest;ons. colored substancel apparently a glass. While the 0 1984 Gemolog~calInstitute of America 228 Notes and New Techniq~~es GElMS & GEMOLOGY Winter 1983 Apart from the band at 490 nm (very weak], no TABLE I. Properties of a cobalt glass imitating lapis lazuli other faint iron bands, recorded by the spectrome- Property Description ter, were detected with the spectroscope. The photographs in figures 2 and 3, talzen in Color Violetish blue of strong saturation reflected light, show bands and aggregates of white DIN 61 64 color indicesa 15% : 6 : 4 (hue, saturation, inclusions that are essentially of two types. One is darkness) a flat dendritic or fernlike array (similar to that in Degree of transparency Opaque to semitranslucent (in thin sections, translucent to figure 9, "metajade," of Hobbs, 1982). The other transparent) consists of planes in radiating to stellate patterns Absorplion (recorded at room Strong bands at 642, 592, similar to coral septa, with the planes perpendicu- temperature) 535 nm (cobalt); faint bands at lar to the bead surface, indicating that the glass 490, 438, 378 nm (iron) was annealed. In contrast to the macroscopic ap- U,V, fluorescence Long-wave, extremely weak Short-wave: absent pearance of the material, the inclusion patterns Refractive index, nD 1.508 on a section (spot seen under magnification are completely different read~ngsslightly lower) from the aggregates of small blue, white, and fre- Optical character Isotropic (in thin sections: quently metallic yellow grains commoi~lyseen in anomalous extinction) lapis lazuli. In figure 3, shallow depressions on the Vitreous (slightly silky sheen on inclusions) spherically molded glass can be recognized, and are Apparent porosity Nonporous in part filled with a white, grainy material that Specific gravity (4OC) 2,453 (one specimen) evidently had never melted. However, true bub- Surface Smooth, spherically molded bles or swirls were not detected, although the glass Surface of fractures Concho~dalto almost flat, w~th was observed with the microscope to be fairly fine structure Luster of fractures Waxy to vitreous transparent around the white inclusions. The in- Streak* scritch Both white clusions themselves ranged in size from a few Mohs hardness Approximately 5Y2 micrometers for the tiny white grains that form External characteristics Regular c~rcularshrinkages the two types of aggregates to approximately 3 min around drillholes, few subspher- for the longest septa and several milliineters for ical depressions (molding marks?), and several filled the ferns. angular cavities on bead surfaces CHEMICAL AND X-RAY Internal characteristics White crystallites of micrometer DIFFRACTION DATA size forming dendritic and large radiating to stellate patterns, In According to Bannister's diagram for conventional most casessurrounded by traris- glasses (Webster, 1975, p. 3861, a calciuin or even a parent blue areas and emanal- borosilicate glass could account for the refractive ing from a grainy center index and specific gravity determined, but no ref- Reaction to heat None to thermal lest tip ; Reaction to ferrornagnetism None erence to this particular lapis-imitation glass was Reaction to diluted HCI None found in the gemological literature (Crowning- Chemical elements Si; Ca, Ti, Mn, Fe, Co, Cu, Zn, As shield, 1974; Farn, 1977; Schiffmann, 1976; (as detected by energy- Webster, 1975; and footnote below*; although dispersive X-ray fluorescence) - -- - -- - aV/esf German color chaff sysfemon fhe bask of C.I.E. ~lluminanls. *A very olcl, ifnot the oldest, artificiallapis-like materialdates back to pre-Christian times, when Egyptians sintered calcite, quortz, malachite, and azuri!e to create a brilliant blue sub- stunce that is now called "Egyptian bl~~e."In ancient Egypt 290 nm to 320 nm (also the result of trace amounts this ~naterialwas used{or scarabs, to ornament royal tolnbs, of Fe], It must be stressed that the peak positions and, inpowdered forn~,as a pigtnent and cosn~etic,The chemi- and intensities visually observed with the spectro- cal co~npositionof "Egyptian blue" is close to, and its cryst(11- line structure identical with, the mineral cuprorivaite, CaCu scope partially deviate from the recorded spectro- [Si40jo] (G. Bayer, pcrsonal communication). The prodz1ction meter data provided in figure 1. With the spectro- of this material was made particularly sz~ccessfulthrough the scope, the bands were seen to be centered at about application of leadoxide or all<alifluxes. Ifless Ca and Cu, but more all~alis,were used, a well-melted transparent-to-o~~aque 660 nm (strong), 585 nm (medium, narrow), and glass (colored by Cu alone, or by Cu, Co, and FE)would result 530 nm (inedium strong, very wide, asymmetric]. (Bayer and Wiedem(lnn, 1976). Notes and New Tech~~iq~~es GEMS & GEMOLOGY Winter 1983 229 Figure 2. White bands and radiating septa of low-cristobalite in a devitrified, opaque cobalt- glass bead imitatinglapis lazuli. Section through bead in reflected light; magnified 6 x (Wild M8/MPS55). Figure 3. Dendritic and radiating patterns of Nassau, (19801, reported a pyrite-lapis imitation white low-cristobalite exsolutions and essen- made of another blue specialty glass that contains tially transparent blue areas in a cobalt-glass copper crystals, similar to a "goldstone"]. Chemi- bead imitating lapis lazuli. Reflected light, mag- cal data, nowadays readily available through nified 13 x (Wild M8/MPS55). nondestructive energy-dispersive X-ray fluores- cence (XRF-EDS; Stern and Hanni, 1982),were cer- tainly of interest in this case. Figure 4 exhibits no successfully masked by a filter, the intensity of the fewer than eight metallic element signals in addi- Si peak was greatly reduced and an additional peak tion to the strong silicon peak. When the AgL due to potassium was resolved, providing for the series produced by the silver tube radiation was identification of at least nine metal oxides ad- COUNTS Figure 4. Unfiltered energy spectrum of a specialty cobalt glass imitating lapis lazuli; 249 SEC ENERGY (KEU) counting time 249 seconds BL GLFISS,21-.02 K2 MOF,1/10/82 (Tracor Northern 1 710 X-ray GEOCHEH. LRB/UNIU. BSL Fluorescence Spectrometer). Notes and New Techniques GEMS & GEMOLOGY Winter 1983 mixed to the Si02glass. Although boron cannot be CONCLUSION detected by XRF-EDS analysis, the fact that the The cobalt glass described in this article is the best beads examined had a Mohs hardness of less than 6 glass imitation of lapis lazuli that this author has (which is low for a borosilicate glass] would sug- seen to date (see also Webster, 1975, p. 221). Al- gest that boron is not present in this material in though lapis lazuli was immediately eliminated as significant amounts.