Today, cathodoluminescence (CL) is a well known and expanding method among sedimentologists [Sippel (1965), Zinkernagel (1978), Marshall (1988)] and gemologists [Gaal (1976-77), Ponahlo (1985) , Sungawa (1992)] to reveal surface effects in thin slides of rock material which cannot be recorgnized by other methods used in petrographic studies or to differentiate between natural and synthetic diamonds, gemstones or artificial products. As far as jadeite is concerned, Ponahlo (1988, 1995) reported that natural untreated jadeite could easily be recognized in a mixture of very like colored minerals such as nephrite, hydrogrossular and amazonite by means of its canary green CL color and peculiar texture. The CL method was said to be applicable to any form of polished pieces, to cabochons and carved figures etc… when making use of a cold cathode device like the Luminoscope mounted on the table stand of a microscope. Later, its efficiency was improved by an attached computer-assisted photomicrospectrometer [Ponahlo (1989)]. Today, with the increased occurrence of dyed jadeite, impregnated quartz and other fake products on the market, it was a necessity to find out whether the recommended CL method could also be used for a fast and accurate recognition of dyed jadeite and impregnated quartz.
Description of samples
It was a prerogative for a study of the CL of natural and dyed jadeite to have samples at hand which were processed from one and the same original jadeite material. A short description of these samples is given in Table 1. The density figures in the same table show some differences as far as both the natural, the bleached and the dyed jadeite are concerned. But the hydrostatic method cannot be easily applied to carved pieces or necklaces. It was used for the purpose of control only.
Sample A (Figure 1) consists of two small slices taken from the natural rough jadeite (013/95). The bleached and the dyed jadeite samples (B(Figure 1) and C (Figure 2) [(0114/95) and (015/95) respectively] are said to have been produced from the same rough jadeite material. The slices of jadeite sample A fitted well into the REM and the microprobe apparatus which were also used in this study. The other two natural jadeites and the impregnated quartz cabochons listed in Table 1 were small enough for REM analyses. All specimens were cleaned in distilled water and acetone dried at 105°C. No change of color was observed. No determination of the density was carried out with the jadeite samples B because of its surface porosity.
Methods
3.1. Density Determinations
The densities of the samples were determined by the hydrostatic method by using alternatively both ethylene dibromide or a new non-poisonous liquid called “Solketal”. Making use of a set of micropipettes, once developed by Pregel in Austria, the hydrostatic method becomes independent of keeping the temperature of both the liquid and the environment constant. But it needs the determination of the density of the immersion liquid at the temperature of the test. Using an analytical balance, the mean error of the density determinations could be reduced to the figures given in Table 1.
3.2. The could cathode CL
The method was first described in PONAHLO (1988) in the German language. Its description, therefore, is briefly repeated in English.
CL is a luminescence phenomenon which is observed when electrons hit the surface of solids in vacua. Depending on the energy of the impinging electrons — usually between 1 and 30 kV — their penetration depth is not more than 1 to 3 μm. In the Luminoscope® used in this study, the electrons were generated in a cold cathode device which is a small discharge tube made of glass of 8cm in length, containing both the cathode and the anode. Like in the gas discharge lamps, electrons are created from the low pressure gas inside the tube as soon as high tension is applied between cathode and anode.
But unlike the procedure in the discharge lamp, the electrons in the cold cathode device pass the hollow anode so as to enter a large vacuum compartment where the sample rests on an x-y-movable sledge. The whole unit is well evacuated to a pressure of 100 mbar in order to have enough carrier gas ions in the gun for a continuous production and flow of electrons. As the amount of electrons, the beam current, is dependent on the gas pressure, it is compulsory to regulate the pressure by means of a gas inlet valve on top of the vacuum compartment.
A solenoid flanged in between the electron gun (discharge tube) and the compartment is an additional necessary gadget. It helps to narrow or widen the ray of electrons is, then, deflected onto the surface of the sample by means of a set of magnets. The current density of the impinging electrons can also be varied. The protection of the observer against the secondary X-rays is secured by means of lead glass by which, at the same time, facilitate the visual observation of the luminescing sample or the gemological inspection under incandescent light. The whole unit (vacuum compartment plus electron gun) rests on the table stand of a microscope that is fitted with long distance objectives so as to guaranty observation of CL effects at low and at higher magnifications.
Non-destructive Method To Distinguish Jadeite and Dyed Quartz Simulants P.61
3.2. The photomicrospectrometer
The luminescing light emitted from the sample is focused through the objective onto the entrance slit of a monochromator mounted on top of the microscope. Spectra of the emitted light can be taken in steps of 1 to 20nm and within the spectral range between 380 and 1,000nm. The width of the entrance slit may be narrowed down to 0.5nm. The diffracted light leaving the monochromator is focused again and targeted to a photomultiplier and further to a “lock-in” amplifier where the luminescent light signals are turned into electric impulses which are amplified and further processed in a computer. The whole unit and procedure is computer-driven. Spectra between 380 to 950nm are run within 25 seconds. Thus, heating up of a sample during bombard-ment is greatly reduced. When working semiquan-titatively, the distance between the front lens of the surface of the sample must remain unchanged. The photomicro-spectrometer works as a single ray unit. Therefore, and adjustment of the beam conditions is compulsory whenever the instrument is switched on. For this procedure a jadeite master stone was used.
The results
4.1. The CL Colors
Table 1, last column, contains remarks on the luminescence behavior of the jadeites A, B, C, and 6 and 7 [numbers (013), (014), (015), (018), and (017) respectively]. The last two samples 6 and 7 are natural green cabochon-jadeites not frequently encountered. Samples A and 7, although different in visual appearance, are characterized by exhibiting the same canary green CL color one observes with most jadeites investigated so far. The macro-photo of the Figure 1 show the two slices of sample A, one rough, the other polished, luminescing far more greenish than the bleached sample B with its predominantly yellow CL. Figure 2 is a macro-photo of the cabochon jadeite 7. With higher magnification the yellow CL color grades into canary green. This can be seen from the micro-photo [5]. The jadeite sample 7 also contains distinct areas which luminesce red, blue, and violet as documented in Figure 4. In nearly all cases of polished jadeite samples, whether caved pieces, steeply curved cabochons or even necklaces, the very peculiar texture of block or felted jadeite crystals is immediately made visible to the microscopist by CL without making and resort to the tin-slide technique.
Additionally, Figure 1 reveals that the yellow luminescing bleached jadeite B contains some small rose-red blotches fringing the sample. The high CL intensity of this bleached jadeite is striking.
Contrary to the behavior of the jadeites A and B, the bleached and dyed jadeite sample C (015/95) displays but a negligibly low CL. Its intensity is best described by a “subdued dark green”. There are a few spots of some yellow brighter CL. Because of the very low luminescence, no macro-photo was taken. At higher magnifications a micro-photo of one of the yellowish luminescing patches was shot which is reproduced in Figure 2. The blocky and slightly felted texture of the luminescing jadeite 7 can best be studied in Figures 4 and 5. Such a blocky texture is but another characteristic feature of natural jadeite that becomes immediately visible by the application of the CL method.
4.2. The CL Spectra
Although jadeite spectra have already been published [Ponahlo (1989)] not much is known about its general applications in gemology. Figure 6 shows some of the cL-spectra of the jadeites worked on. The findings are striking. First of all, most jadeite spectra investigated in this work are characterized by a characteristic single band. This band consists of two peaks, one at 550nm, the other at 562nm. No difference exists between the canary green band of the CL spectrum of the natural jadeite A (thick full line) and the more yellow one of the bleached sample B (long dashed line). The CL spectrum of the sample B is made of an additional band in the NIR with two peaks at 756nm and 764nm respectively. The NIR band shows an even greater intensity than the yellowish one. As expected, the bleached and treated jadeite C shows but a weak spectral response (thin short dashed line) with the meager indication of a CL band at 566nm.
Figure 7 contains the CL spectra of the other two jadeites 6 and 7, and additionally, those of the two impregnated quartz samples 4 and 5. Added for ease of comparance, is the CL spectrum of the bleached jadeite B. One immediately recognizes how well the spectra of both jadeites 6 and 7 fit that of jadeite A. Only a very small deviation of their peak wavelengths (558nm instead of 552nm) and a tail of the greenish yellow band into the red are peculiarities of the jadeite 7. The impregnated quartzes contrast markedly both in their visual CL colors and their respective CL spectra. They are dealt with below.
4.3. The CL colors and CL spectra of impregnated quartzes
In Table 1 the quartz 4 is described luminescing with dark reddish violet CL color exhibiting a very low luminescence indeed. Quartz 5 is said to display a strong reddish brown CL color with few orange patches. Both CL colors and the textural features of these luminescing two impregnated quartzes are y no means comparable to the CL colors and textual features of the jadeite regardless of origins. This follows also from Figure 8, which contains a macro-photo of the particular brownish red CL ocular of the sample quartz 5. The quartz spectra of Figure 7 reveals their striking difference to those of natural jadeite. One clearly notices that the CL spectra of the quartz samples do not contain any CL band in the greenish yellow spectral range, and no indication of a NIR band. Only the orange patches of the impregnated quartz 5 might occasionally cause a medium intense band in the red with small peaks at 626, 642 and 658nm. These findings are conclusive proofs of the diagnostic value of both CL methods, the visual inspection and the photomicro-spectrometry.
Conclusions
5.1. The 557nm band and the manganese ion.
Jadeite is a pyroxene of the chemical formula NaA1[(SiO3)2]. It differs from diopside, another pyroxene of the chemical formula CaMg[(SiO3)2] , by the replacement of divalent Ca²+ by monovalent Na+ and of divalent Mg²+ by trivalent A1³+. There is wide agreement that the A1³+ion, octahedrally coordinated, can easily be replaced by the activator ionCr³+.Jadeite as a metamorphic, high pressure, and low temperature mineral incorporates also other elements in small amounts like magnesium or calcium, iron and titanium which have different preferences for certain lattice positions in such pyroxenes. This also follows from the jadeites studied in this work as can be seen from the REM and microprobe data listed in Table 2. In particular, Ca²+ as an eight fold coordinated cation is preferentially built in the center of a distorted M2 cube that is formed by the surrounding oxygen anions in the lattice of the diopside and other pyroxenes. Since the studies of GREEN (1981) and other research workers, it is agreed that the Mn²+ion which can replace the Ca²+ion[Walter(1985)] easily acts there as an activator to excite CL. One observes in diopside a Mn²+- based CL band peaking at 574nm in the yellowish green spectral range. Coincidentally, the strongest CL band of most jadeites investigated so for is also found in the same spectral range located at 557±6nm; that is only 20nm lower than the diopsidic CL band. Therefore, it seems justified to ascribe this CL band to divalent manganese ions as activators being also centered in M2 sites in the jadeite lattice. Excitation spectra are under way to corroborate this assumption.
5.2. The NIR band and the difficulty of assignment
Since the vast research work carried out on rock samples from the moon [Geake et. al. (1971) and Telfer et. al. (1978) among many others] it became evident that tetrahedrally coordinated Fe³+ ions may also excite CL, although iron was primarily known as a most active quencher of luminescence. The Fe³+ -based CL bands are found at wavelengths above 715nm in the NIR.
In a sodium pyroxene like jadeite such a configuration of iron is rarely met with. Either an Fe³+ is replacing A1³+, or and Fe²+ is preferentially built in M1 sites normally occupied by Mg²+, In both cases iron is six fold coordinated and cannot excite luminescence. In jadeite we, therefore, may find Fe³+ ions in tetrahedral configuration only if they have replaced some of the Si4+ positions of the silica tetrahedral chains favoring charge compensation. Although there exist no such proof in this work, the author is inclined to tentatively ascribe the broad NIR band in the spectra of excited jadeites (with peaks located between 745 and 760 nm) to such a Fe³+ activator by assuming the above mentioned replacement and charge compensations of this ion. At present the problem is under discussion and a solution must be left to further studies which are under way.
Summary
The present study has confirmed the efficiency of the CL method to differentiate non-destructively between natural untreated jadeite and its bleached and dyed products. It has also brought evidence for the striking difference between the CL of a natural jadeite and an impregnated quartz, however careful the latter may be prepared. Compared to other equipment like photometers for IRS, a modern CL apparatus with an attached photomicrospectrometer is a low cost device. It can be used for the solution of a large variety of gemological problems. The solution of one of these problems has been presented in this work. Clearly, the research on the CL of the jade group is just in its upward move. Plenty of CL
Data on jadeite and its substitutes have been obtained by the author within the last years. These data are critically revised at present. It is hoped that conclusive studies of available samples by means of excitation spectra, REM, microprobe and EDPXRF trace analyses, eventually, will help to shed much more light on the safe solution of recognition of jade materials. There is no doubt that such results — like those presented in this work — will widen the application of CL to differentiate between most of jadeites, their substitutes, non-jadeite minerals and the many treated products.
Acknowledgments
The present study has been carried out at the department for Mineralogy and Petrography of the Museum for Natural History in Vienna, Austria, director Prof. Kurat, to whom the author wants to express his gratitude for support. The REM analyses have been carried out by Dr. Walter, and the microprobe analyses by Dr. Brandstatter, assistant manager in the department. It deserves special mentioning that it was a pleasure to cooperate and discuss relevant problems with these research workers. Considering the difficulty to obtain reliable jadeite gem material for such studies the author is much obliged to Mr.ken Scarratt, director of the AIGS, the Asian Institute for Gemological Sciences in Bangkok, who generously has made available the samples listed in Table 1.