Research News

An Imitation "Melo Pearl"

Figure 1: The 52.49 carat "pearl" as submitted by the client. Note the grooved area towards the top. The banding that meets this feature forms a herring-bone pattern. Photograph: Adirote Sripradist.

Every now and then a client will submit a “pearl” for identification that they have been told, or believe, is from a particular mollusc. Such was the case with the 52.49 carat near-round “pearl” measuring 19.62 x 19.21 x 19.16 mm (Figure 1) that was recently submitted as a Melo “pearl”. Whilst not an obvious pearl from its external appearance it certainly could pass as one, so more detailed examination, and therefore a report, was certainly warranted, and it was understandable why the client wanted to be certain of the result.

Preliminary visual examination with a loupe and subsequently a microscope revealed distinct relatively fine banding (Figure 2) that formed a herring-bone pattern along a clear surface groove in one area. This in itself started to cause concern since it is very unusual to observe any structural banding in a pearl that can be considered “straight” unless it is somehow associated with the base of a blister pearl, when it might just be possible to get some straighter banding if there is any associated shell somehow still attached to the pearl. However, since the banding in this object appeared to run throughout the whole piece its “pearly” nature was immediately in doubt. To add even more doubt to the identity we also

Figure 2: The clear banding present throughout most of the pearl. This was very directional and this orientation shows it the best. Photomicrograph: Pantaree Lomthong. Magnified 7.5x.

noted that the colour was a little patchier than we would expect for a naturally coloured pearl of this type, so closer examination was applied and it wasn’t long before some small but distinct orange colour concentrations were observed within surface-reaching cracks (Figure 3) running off the main surface groove already described. This provided further evidence that all was not quite as it appeared. Yet we still had to gather more information to support our preliminary opinion and this meant that we had to turn to other equipment, including some of the more advanced analytical procedures, at our disposal.

Figure 3: An area within the surface groove revealed small concentrated orange colouration along fine surface-reaching fractures. These are particularly evident towards the top and bottom of this image. Photomicrograph: Nick Sturman. Magnified 90x.

After obtaining a specific gravity result of 2.84, we proceeded with our analysis by checking whether we could see any indications of a pearl-like structure using our in-house Faxitron real-time microradiographic system. This soon revealed a banded structure in the first orientation (Figure 4) and a repeat of the herring-bone structure seen via the preliminary optical examination when it was first inspected, in another orientation. From our experience most natural Melo pearls reveal one of two microradiographic structures: either very little to no useful structure; or sometimes a small void-like feature near their centres. The banded nature of this particular sample was therefore not something we would associate with a natural concentrically formed Melo pearl, or a pearl from any mollusc for that matter. The nature of the material forming this particular “pearl” was therefore concluded to be shell, so it appeared that we were handling an example of an imitation Melo “pearl” fashioned from shell.

Figure 4: The microradiographic structure in this orientation revealed the clear banding associated with shell imitations. This inner structure clearly reflected the externally visible banded nature.

What about the colour? We still had to prove whether it was naturally coloured or not. Examination under ultra-violet lighting did not provide any useful reactions for either shortwave or long-wave radiations. We therefore focused on three pieces of equipment to help determine the colour origin. Analysis with an energy-dispersive X-ray fluorescence spectrometer (EDXRF) revealed nothing out of the ordinary using a general method and confirmed that the shell used was fashioned from a saltwater mollusc owing to its low manganese and higher strontium content when a specific carbonate method was applied. A spectrum obtained after using a Perkin-Elmer 950 ultra-violet visible near-infrared spectrophotometer (UV-Vis-NIR) showed the absorption features expected for an orange coloured pearl or shell, with no features that would make us comfortable calling the colour treated, so this left us with the option of using a Renishaw inVia Raman microscope fitted with a 514 nm Ar-ion laser to determine the cause of colour more specifically. The results of this analysis showed that the shell was composed of aragonite owing to the presence of one main vibrational peak at 1086 cm-1and an accompanying doublet at around 704 & 702 cm-1, but it also proved it was not naturally coloured, since the polyenic (Salares, 1977) peaks expected for naturally coloured pearls or shell belonging to species such as Melo or Cassis amongst others, were lacking (Figures 5 and 6). Even in relatively lightly coloured pearls with natural polyenic pigments some small peaks in the regions of 1120 cm-1& 1520 cm-1should usually be visible, so when the quite saturated coloured areas of this “pearl” were examined it should have revealed these peaks quite strongly too. The fact that the peaks did not appear at all just confirmed that the colouration must be artificial.

Figure 5: Spectra of the dyed shell with the missing polyenic peaks at ~1120 cm-1& 1520 cm-1, compared to that of a naturally coloured Melo species pearl.

Figure 6: Enlarged area between 1100 cm-1and 1600 cm-1showing the polyenic peak region for both samples shown in Figure 5. The polyenic peaks in the naturally coloured Melo species pearl are clearly visible in the top spectrum and completely absent from the lower spectrum belonging to the imitation “pearl” discussed in this article.

Whilst we could not be more specific on the exact nature of the pigmentation used to colour the shell, it bore a similarity to the dye detected in another imitation pearl submitted to the AGTA laboratory in 20081.

Another feature of the imitation that made its non-pearl status more likely was that it showed clear evidence of having been shaped and polished. Whilst heavy working (i.e. filing or grinding to alter the shape) and polishing to improve the lustre are not proof in themselves of a shell origin, both of these actions are usually required to produce an imitation pearl from a piece of shell and so it makes it that much more suspicious when both are also present in addition to all the other factors that have been taken into account.

Dyed imitation Melo “pearls” are not a particular rarity in the market, but they do provide gemmologists in laboratories with something different to look at since they aren’t submitted on a particularly regular basis because most clients have a good idea whether they have an original pearl or not, and justifying the costs of testing something they suspect is an imitation is rather difficult. Therefore the only imitation pearls laboratories tend to see are those that clients truly believe are pearls, or are ones that are submitted for the sake of interest/research (Krzemnicki, 2006; Wentzell, 2006, ). As with all of the imitation Melo “pearls” mentioned in the two articles referenced above, the most likely origin of this particular dyed shell is also the Tridacna species (various species of Clam including Tridacna gigas) based on the thickness of the shell and obvious flame pattern.

1The article is no longer available via the AGTA website, but web searches can find a few references.

Krzemnicki, M.S., (2006), A Worked Shell Bead as an Imitation of a Melo Pearl
Salares, V.R., Young, N. M., Carey, P. R., Bernstein, H. J., (1977), Excited state (excitation) interactions in polyene aggregates. Resonance Raman and absorption spectroscopic evidence., J. Raman Spectrosc., 6, 282-288.
Wentzell, C.Y., (2006), Imitation Melo "Pearls", Gems & Gemology, 42, 2, 166-167.