Help|Log In|Register|
Home PageMindat NewsThe Mindat ManualHistory of MindatCopyright StatusManagement TeamContact UsAdvertise on Mindat
Donate to MindatSponsor a PageSponsored PagesTop Available PagesMindat AdvertisersAdvertise on Mindat
Minerals by PropertiesMinerals by ChemistryAdvanced Locality SearchRandom MineralSearch by minIDLocalities Near MeSearch ArticlesSearch GlossaryMore Search Options
Search For:
Mineral Name:
Locality Name:
The Mindat ManualAdd a New PhotoRate PhotosLocality Edit ReportCoordinate Completion ReportAdd Glossary Item
StatisticsThe ElementsMember ListBooks & MagazinesMineral Shows & EventsThe Mindat DirectoryHow to Link to MindatDevice Settings
Photo SearchPhoto GalleriesNew Photos TodayNew Photos YesterdayMembers' Photo GalleriesPast Photo of the Day Gallery

My Specimen of Diamond in Conglomerate From Brazil: Real or Fake?

Last Updated: 7th Oct 2017

By Norman King


Diamonds in conglomerate matrix have been found occasionally in Brazil, primarily the states of Minas Gerais and Bahia. Many people have urged caution with respect to their legitimacy, however, suspecting that at least some of those specimens were faked-–that loose diamonds were glued to the conglomerate. This topic was discussed on the Mindat Messageboard between June 2009 and January 2011 (see References, Part B. Websites). I recently acquired a specimen of diamond in conglomerate from Jack Crowley of The Crystal Mine, who stated he “had it on good authority” that it was almost certainly a fake, referring to the on-line Mindat discussion. He offered the specimen (Figure 1) at a reasonable price to begin, and later reduced the price at least twice, apparently having succeeded in convincing nearly all prospective customers not to buy it. As for me, I thought it would certainly make a good conversation piece, and it might actually turn out to be a legitimate work of Mother Nature rather than Man. Jack reported that he acquired the specimen from Miner’s Lunch Box (Scott Werschky), who in turn had acquired it from Jewel Tunnel Imports (Rock Currier). This was the only such specimen I have seen where the dealer suggested it was not a legitimate natural specimen, or even that it had been “repaired.” Maybe Jack was being overly cautious about it, but he surely stands as an example of the highest degree of integrity among dealers. In fact, I concluded that the specimen is indeed natural, and learned a lot about alluvial and conglomeratic diamonds during my examination of the issue. I suggest several observations that most anyone can make to recognize the real thing.

Figure 1. Hand specimen of diamond in conglomerate from Bahia, Brazil. Arrow shows location of the diamond crystal. Specimen dimensions are 10.0 x 6.7 x 3.6 cm. The forward-facing surface has been weathered to varying degrees (strongest at front). A freshly-broken, unweathered surface is on the right. More intensely-weathered matrix has lighter color than unweathered matrix.

The Issue of Glued-On Diamonds

Contributors to the Mindat discussion stated that some (but not all) diamonds in conglomerate were likely glued to conglomerate matrix to fool prospective buyers into paying a higher price than a loose rough diamond would command. (Apparently this is also a concern for diamonds in kimberlite matrix.) Mindat discussants reported having seen no fluorescence or signs of chemical glues on the specimens of diamond in conglomerate they had examined, however. Some guessed that water-based glues may have been used, and others suggested that the “glue” was actually clay. One observed a diamond that had been embedded in clay fall out when the specimen was wetted. A detailed “post-mortem” on a faked specimen of diamond in conglomerate would surely be helpful here. (What happened to those specimens?) We need to know what the conglomerate looks like at the place of diamond attachment. Were any other large grains attached either to the diamond or to adjacent matrix? Was matrix surrounding the diamond different from matrix elsewhere in the conglomerate?

Some specimens pictured on the Internet involve a diamond resting on a rather flat side, making contact with a flat matrix surface–-a situation begging for skepticism. Several such specimens have diamonds nestled tightly in niches formed by fortuitous orientation of other kinds of large grains. Those diamonds are essentially on matrix, rather than in it, and they could easily have been artificially attached. I have not examined those specimens personally, however, and therefore cannot express an opinion on them. If genuine, such specimens would suggest certain events. For example, collectors probably used hammers and chisels to loosen the sample from an outcrop, apparently doing so without damaging the diamond or knocking it off the matrix. Or, perhaps a diamond was found in conglomerate tailings, still sticking up from the matrix after undoubtedly rough handling by garimpeiros, perhaps including pushing or dumping big chunks of conglomerate into piles using a “Bobcat” or similar device.

The diamond in my specimen is mostly within the conglomerate, with relatively little of it exposed at the surface (Figure 2). A few grains adjacent to the diamond are “leaning” against it, partially covering the opening through which it protrudes. Manufacturing such a “pocket” for insertion of the crystal would have required a high level of skill, including digging a hole for it in the conglomerate, probably having to remove other large grains where the diamond was going to be, and then reattaching a few of them near the surface to partially cover the diamond around its edges. When this work was completed, clayey material must have been packed into the hole to fill any remaining spaces. Of course, in addition to coming up with the diamond, someone had to find a suitable piece of Bahian conglomerate to serve as the diamond’s new “home.” Assuming this scenario is correct, the specimen was then probably carried around for some time by the merchant before he/she found a buyer willing to pay more for diamonds in lumps of rock than for loose diamonds. In fact, this diamond could probably have commanded a good price by itself in the local diamond markets of Brazil. Economic realities of today’s diamond trade would more likely encourage most garimpeiros and middlemen in Brazil to prefer a quick, unencumbered profit from the sale of their loose diamonds than overseeing the manufacture of forgeries good enough to fool prospective purchasers. And how many collectors are looking for one of these not-very-showy specimens of diamond in conglomerate on which to spend their hard-earned cash?

Figure 2. Close-up of diamond in conglomerate. Exposed portion of diamond crystal is 6.6 mm, maximum. A bit more may be covered by matrix. This crystal probably weighs about 1.5 carats.

Locality Cited as the Source of My Specimen

My specimen was reported to be from the area of the Rio Formosa in the state of Bahia. Diamonds in conglomerate had previously been reported from the Rio Formosa and also the Rio Formosa do Norte. I searched on-line maps and Google Earth, finding neither a Rio Formosa nor Rio Formosa do Norte in the state. Bahia does include the Rio Formoso (ending in “o”), so Rio Formoso is probably the correct spelling, and is actually the spelling required by proper syntax. This river begins near the southwestern corner of the state, flowing for just over 300 km before feeding into the larger São Francisco River. Locations of the Rio Formoso and São Francisco Rivers are shown in Figure 3. A specimen of diamond in conglomerate in the Hunterian Museum has also been attributed to the Rio Formoso (spelled correctly; see References, Part B. Websites). The Rio Formoso, however, does not flow across terrain where diamond-bearing conglomerates are exposed. All well-documented conglomeratic and placer diamonds from Bahia were mined or collected in the Chapada Diamantina (Diamond Plateau) region in the central part of the state, or the Espinhaço Range (pronounced ess-pin-yah-so), a narrow, roughly N-S trending range of mountains just west of the Chapada Diamantina. The diamonds were eroded from exposures of the Espinhaço Supergroup of Proterozoic age that is exposed in those areas.

Figure 3. Simplified relief map of Bahia (colored) and surrounding states in Brazil. Rio Formoso and São Francisco River are shown, along with Chapada Diamantina and Espinhaço Range. Red outline shows the boundary of Chapada Diamantina National Park. Modified from map on “geografia-geral-e-do-brasil” website.

Modern alluvial deposits were mined extensively for placer diamonds in the Chapada Diamantina beginning in the 18th Century, but large-scale mining ended soon after kimberlitic diamonds were discovered in Africa in 1869. Prior to that discovery, all diamonds used in the gem-stone trade, world-wide, came from stream gravels such as those in Brazil. With the eventual expansion of kimberlite mining elsewhere, only small-scale mining of placer diamonds continued in Brazil. A few diamonds were also mined from outcroppings of ancient conglomerates in Minas Gerais. During the 20th Century, geologists and engineers from the De Beers company studied the prospects for economic diamond-mining operations in Brazil. They left the country without finding any encouragement, farming out their exploration rights to smaller, mainly Brazilian companies. More recently, at least two kimberlites have been mined for diamonds in Bahia and one in Minas Gerais, but large-scale diamond mining from kimberlites in Brazil is in its infancy, and the degree of economic success relative to international competition has yet to be determined (e.g., see Svisero et al., 2017).

In 1985 the Brazilians set aside a portion of the Chapada Diamantina as a national park (red outline in Figure 3) in order to protect the unique terrain and associated ecosystems from further environmental damage resulting from continued, relatively unprofitable, placer mining activities. They also hoped to preserve historical buildings and other aspects of the earlier diamond and gold mining industries, encouraging the local population to transition from mining to working in ecotourism and associated support systems. Although mechanized mining has been prohibited in the region since 1996 (Pedreira, 2001), it is still legal to pan for diamonds and pick up loose rocks most anywhere except in the parks and various other reserves.

My specimen may have been attributed (or mis-attributed) to the Rio Formoso because that would mean it was found far from protected areas. At its closest, the Rio Formoso drainage is 230 km WSW of the Chapada Diamantina/Espinhaço region, so people would have no basis for suspecting the specimens had been collected illegally. Another possible reason for the Rio Formoso attribution is (supposedly) that some professional collectors purposely misidentify certain localities in order to keep competitors away from potentially productive sites. Finally, it may also be that a minor stream elsewhere in Bahia has been referred to as Rio Formosa or Rio Formosa do Norte by local people, apparently incorporating the incorrect syntax mentioned earlier--not unusual in local dialects of Portuguese in Brazil.

Where Has My Diamond Been, and How Did It Get Into Conglomerate?

The Espinhaço Supergroup is the only known source of diamonds in conglomerate in Bahia. The supergroup reaches more than 2,000 meters in thickness in the Chapada Diamantina, and consists primarily of siliciclastic strata deposited in non-marine to shallow marine settings in a continental rift basin (e.g., Pedreira and Waele, 2008). Diamonds have been found in two of the several conglomeratic units in the supergroup, the Tombador Formation and the younger Morro do Chapéu Formation (Battilani et al., 2007). Igneous intrusions that cut those strata, and lava flows or pyroclastic deposits interlayered with them, have been dated at between 1.5 and 1.1 billion years old.

In fact, all terrestrial diamonds had originally crystallized in the Earth’s upper mantle or lower crust. Only at those depths would conditions of temperature and pressure be high enough for crystalline carbon to take the form of diamond rather than graphite. The diamond crystals were later carried to the surface in kimberlitic magmas. Those ultrabasic magmas contained CO2 gas derived from carbonate mineral precursors. Generation of large quantities of the gas resulted in almost explosive upward migration and eruption of the magma at the surface. On the way up, the magmas broke off chunks of previously-lithified upper mantle and lower crustal rocks containing diamond crystals that had crystallized prior to the eruptions by perhaps hundreds of millions, or even billions, of years (Shirey and Shigley, 2013). Another ultrabasic rock type from deep in the Earth, called lamproite, has also sourced diamonds in Brazil, but the diamonds are uncommon, and none of gem quality have been found so far. Dark-colored, polycrystalline diamond known as carbonado has been found and mined in Bahia, but it has little value and is used primarily as an industrial abrasive. Its origin is uncertain. Carbonado is not the topic of this discussion, and is not mentioned further.

After their eruption and solidification, the ancient kimberlite bodies in Brazil were weathered and eroded, releasing the contained diamonds and other rock components into stream gravels that ultimately became Espinhaço conglomerates. These kimberlites are considered to be the primary sources of the diamonds, even though the diamonds did not originally crystallize in them. This seemingly odd situation (the labeling) probably has to do with the fact that we lack access to the primordial rocks in which diamonds crystallized because they are more than 100 km in depth. If any of those rocks were to work their way to the surface by plate tectonic processes, the diamonds within them would become resorbed well before reaching a level where they could be collected. The host rocks themselves would undergo considerable retrograde metamorphic changes, so they would be different rocks by the time we could collect them. Our only direct information about these super-deep rocks comes from xenoliths in the kimberlites. Conglomerates of the Espinhaço Supergroup are considered to be secondary sources of the diamonds, and the modern gravelly stream alluvium where most placer diamonds have been mined are a tertiary source. Younger kimberlites in both Bahia and Minas Gerais have produced diamonds, but those kimberlites did not source Espinhaço diamonds because they were emplaced after deposition of Espinhaço strata. The surficial and hypabyssal portions of kimberlites that did source Espinhaço diamonds must have been largely removed by erosion, accounting in large part for our inability to locate any of them. The region is also covered by deep lateritic soils, further hindering our search.

The Diamond

The diamond in my conglomerate specimen is a single crystal having maximum dimension of 6.6 mm, not including a small portion that is covered by matrix (Figure 2). It probably weighs about 1.5 carats. The crystal is a rhombic dodecahedron whose major faces are slightly curved. Those faces are also marked by subdued, somewhat-elongated hillocks having multiple parallel steps, plus possible impact marks (see examples of such features in Kaminsky et al., 2001; especially Figs. 4 and 14). Smaller faces modifying the dodecahedron include elongated trapezohedral faces that bevel edges between dodecahedral faces, and small triangular octahedral faces. Corners and edges are sharp, indicating a short distance of transport between the ancient kimberlite outcrops and where they were deposited in Espinhaço stream gravels. My diamond seems to have a subtle brownish or pinkish-brown color. Reddish brown color of matrix beneath and on the sides of the diamond, however, might be responsible for that tint. High clarity allows major irregularities in the matrix beneath the diamond to show through, but multiple reflections and refractions obscure the details. Svisero et al. (2017) reported that diamonds from the Chapada Diamantina and Espinhaço Range of Bahia are mostly high-clarity dodecahedrons having curved faces, and that brown is the second-most common color, after colorless.

The Conglomerate

The conglomerate is generally reddish brown owing to abundant presence of iron oxides in the matrix. Sedimentary particles (clasts) in my specimen span the size range from clay to gravel. The largest clasts are about 18 mm in maximum dimension. To be conglomerate, a rock must have at least 30% of its volume consisting of clasts larger than 2 mm, as this specimen does. These are called framework grains, and grains smaller than 2 mm make up the matrix. The framework grains in this specimen are not in contact with each other, seemingly “floating” in matrix (Figure 4-A). This is therefore a matrix-supported conglomerate, as opposed to grain supported. Matrix-supported conglomerates are often called diamictite, but that more technical term may not be widely known and does not appear in much of the literature on Bahia and Minas Gerais diamonds. Therefore I use the more common, and more general term “conglomerate” for this discussion. In a broader sense, many mineral collectors also use the indefinite term “matrix” for a whole rock of any kind which includes an important mineral of interest (e.g., “diamond in matrix”), adding another possibility for confusion. Unfortunately, both usages of “matrix” have become entrenched, so we will have to live with them.

Figure 4. Two views of the freshly broken surface of the conglomerate. A) Close-up of the unweathered conglomerate, showing both framework and matrix material. B) Same view with labels. Grains labeled “Q” are dominantly quartz, but other subordinate minerals such as feldspars may be present in them; “Li-maf” shows a mafic lithic grain; “IL?” is for possible ilmenite. Arrows show partial linings of orange hematitic material around larger framework grains. The pockmark is a hole left by a large well-rounded framework grain, probably quartz, falling out of the rock.

The mineral composition of both framework and matrix in conglomerates varies from one location to another and from one stratum to another. Original mineralogy as possibly modified by metamorphism and weathering, depending on specific history, are prime determinants of mineral composition of the sedimentary debris in streams. Processes of transportation and deposition of this material, and then lithification and diagenesis during the transition to sedimentary rock, also played a role in determining the composition and texture of the resulting conglomerate. Framework grains in my specimen are dominantly quartz. Lithic grains are also present, as well as an opaque black mineral that may be ilmenite. Most of the larger grains are well-rounded, but progressively smaller grains are subrounded to subangular, with grains near the lower limit for framework (2 mm) being fully angular. A few show rounded crystal forms and/or striations. Note that many (but not all) diamonds in conglomerate retain some or all of their crystal faces because of their considerable durability, even if other framework grains are largely well rounded. My specimen is finer-grained than most specimens of diamond-bearing conglomerate from Brazil pictured on the Internet, and lacks the very large clasts (25 mm and larger) seen in many of those specimens. Therefore, my specimen may not be from the precise locality as the majority of illustrated specimens, but it is similar enough to other material attributed to the Rio Formoso drainage to accept it as from that area.

Most of the larger framework grains are at least partially coated by a very thin layer of hard, light brown to dark yellowish-orange (GSA Color Chart), clay-sized hematitic material (Figure 4-B). These coatings appear to have precipitated in situ around the grains, as they are present well into fresh matrix, demonstrating their formation naturally and independently of modern weathering. Portions of the partial hematitic coatings are still present on weathered surfaces of the clasts, showing that the coating does not slake when wetted, and therefore does not include clay minerals. Pockmarks in freshly-broken surfaces of the rock are where large clasts were originally present but “popped out” of their pockets. Hematitic linings are still present in some of those pockmarks.

I recognized two end-members of matrix, fresh and weathered (there are also intermediate stages):
(1) Fresh, unweathered matrix is tough, compact and dusky red to dark reddish brown (colors from Geological Society of America’s Rock Color Chart, 1991), ranging to nearly black in small patches probably due to local presence of manganese oxides. Most of the visible particles in this matrix are quartz silt and sand, grading upward into framework size (larger than 2 mm). The smaller gains among these are very angular. Actually, the matrix is dominated by clay-sized particles. Here is another possible source of confusion. The term “clay” can be used for the finest grain size in sediments and sedimentary rocks (smaller than 1/256 mm in most classifications), as well as for phyllosilicate minerals and minerals which impart plasticity to clay and which harden upon drying or firing (Clay Mineral Society, 2017). “Clay” (singular) usually refers to a sedimentary size class, whereas “clay mineral” or simply “clays” (plural) refer to the phyllosilicate minerals. How you use the term should be clarified where necessary.

(2) Weathered matrix is friable and pale yellowish-orange to dark yellowish-orange (GSA Color Chart), the lighter color being probably a result of goethite replacing much of the hematite. Grains in the silt and sand size-ranges are readily visible on weathered surfaces of the specimen, but clay-sized particles are difficult to discern individually using ordinary microscopes. Fluffy-looking, dark yellowish-orange clumps of very fine material are present here and there in the specimen, and those are probably rich in clay minerals. Exposed surfaces are coated by micaceous flakes oriented parallel to specimen surfaces, imparting a conspicuous sheen to those surfaces (Figure 5). This sheen is the hallmark of sericite, a variety of muscovite. The mica may be what has been referred to as illite in much of the literature on Espinhaço strata. Illite is also a variety of muscovite, and Süssenberger et al. (2014) noted that Espinhaço illite is indistinguishable from muscovite by x-ray data (I have not done XRD on any material in the specimen). Rocks in this part of Bahia were indeed weakly metamorphosed, reaching greenschist grade at most. Because no distinct minerals of metamorphic paragenesis are present, degree of illite crystallinity (the “Kübler index”) must be used to verify intensity of metamorphism (Süssenberger et al., 2014). Micaceous crystals oriented parallel to specimen surfaces is not true foliation, which would reflect the regional stress field(s) during orogenic episodes. Apparently, tectonic stresses brecciated the rock, forming small-scale fracture surfaces along which hydrothermal solutions flowed, introducing the potassium required for precipitation of muscovite. Potassium-argon dates reported for Espinhaço illite by Süssenberger et al. (2014) indicate three major times of crystallization ranging from 650 to 440 million years, consistent with production during various phases of the Brasiliano Orogeny.

Figure 5. Slightly weathered surface parallel to bedding in the conglomerate that is coated by micaceous material having a silky sheen caused by parallel alignment of mica flakes. This sheen shows the micaceous material to be a variety of muscovite known as sericite. Some of the larger detrital grains seen here are in the framework size grade (just over 2 mm), and many show crystal faces or are otherwise fully angular. Field of view is 21 mm.

On weathered surfaces of my specimen, framework grains stand in relief above the level of the matrix (Figure 6), a result of micaceous material falling away during modern weathering. Along contacts between matrix and the framework grains, mica flakes and finer particles seem to lap onto the framework grains, where the layer gradually thins to nothing across a curved surface resembling the outer edge of a wetting (concave) meniscus (red arrow, Figure 7). Note that this comparison refers only to that geometry, as true menisci are actually traits of liquids related to surface tension. These meniscus-resembling contacts must be natural, as manufacturing such contacts between matrix material and all of the exposed framework grains in a sample would be clearly beyond the capabilities of even the most skilled, meticulous, and patient craftspeople.

Figure 6. Heavily-weathered surface of the diamond-bearing conglomerate, on which framework grains stand out in relief due to disintegration of matrix caused by weathering. This surface also has a sheen from micaceous material (muscovite, variety sericite) that must have formed prior to weathering, or it wouldn't be present below the original surface of the rock. Field of view is 25 mm.

Clay-sized matrix in this specimen can be easily brushed away using a child’s circular water-color paint brush. Upon wetting, the water quickly becomes cloudy and yellowish, but the matrix does not completely slake, probably because the silt and sand provide “body” for it. A teasing needle readily “feels” these coarser particles in the matrix. Interestingly, as soon as very fine particles are brushed away from the surface of a framework grain, more come to replace them, jumping from the adjacent matrix to freshly-cleaned surfaces of the framework grain (including the diamond--see green arrows in Figure 7). This effect is probably due to electrostatic forces, and results in framework grains in high-magnification photographs always appearing “dirty.”

Figure 7. Close-up showing framework grains embedded in orange matrix. Field of view is 5.5 mm. The diamond crystal is labeled in blue. Two green arrows at lower left show areas having several bright specks. These are clay-size particles on the surface of the diamond crystal that jumped from matrix to diamond surface immediately after I had used a brush to remove similar specks of “dirt” for photography. This behavior is probably due to electrostatic charges. The plum-colored arrow points to a grain that exhibits aspects of its crystal habit--tabular with striations. The mineral is probably ilmenite, which has been reported to be abundant in many sedimentary deposits of Bahia. The matrix includes micaceous material that, where it thins out, “laps” onto framework grains, including the diamond. This results in curved, thinning surfaces resembling wetting (concave) menisci in liquids (general area beyond point of red arrow). Note the short, shallow groove in the surface of the diamond, just beyond point of red arrow, that is filled with the same kind of matrix as elsewhere in the specimen. Meniscus-like contacts of matrix with framework grains are visible in the larger grooves between framework grains (note large groove from red arrow to upper right corner of photo). This aspect of contacts between matrix and framework grains would be difficult, if not impossible, to create artificially, supporting the conclusion that the diamond is present naturally in the conglomerate rather than having been added to the specimen by a human.

Summary: What to Look For to Determine If the Diamond Really Belongs in That Conglomerate

Each specimen of diamond in conglomerate is unique–the diamond is unique, and the matrix is unique. Actually, this is true for any mineral specimen, although certain generalizations can be made for specific minerals from specific areas. For example, experienced collectors can envision the general appearance of a specimen of sphalerite or galena from a Mississippi-Valley-type lead-zinc deposit, or of aquamarine from a pegmatite in Afghanistan. Far fewer people, however, can envision a diamond in conglomerate from Brazil. In fact, matrix diamond specimens are so rare that our science’s meager “sampling program,” consisting primarily of specimens in our individual collections, probably does not adequately convey the range of lithologic features of Brazilian conglomerates that contain diamonds, or gemological qualities of the diamonds in them. This relative shortage of information may contribute to differences of opinion about whether either the diamond or the conglomerate is sufficiently representative for a certain claimed locality to be correct. Although an indicated locality may be wrong, it may still be that the specimen formed naturally. Of course, we really only have to satisfy ourselves that any specimen of diamond in conglomerate is indeed a natural product. That is important if we are considering purchasing the specimen, selling it, displaying it in a museum, or appraising it. If you collected it yourself, you know the truth, but you may need to point out geological features visible in your specimen to convince others.

Alluvial diamonds often exhibit qualities such as habit, size, color, and/or clarity that are typical for particular localities or regions. Most deposits contain diamonds exhibiting a range of those qualities, however, and most diamond collections from a specific locality will have a number of them that display at least one atypical physical property. This results from the explosive nature of kimberlite eruptions, which caused the rising magma to entrain diamond xenocrysts from different depths and different parent rock bodies. Diamonds having varied crystal forms, impurities, and resorption histories therefore become mixed together in a single kimberlite. Moreover, some alluvial deposits, modern and ancient (e.g., Espinhaço conglomerates), may include diamonds sourced by two or more kimberlites. Thus, the physical traits of a particular diamond do not unambiguously indicate the specific locality or conglomerate bed it came from.

The conglomerate holding my diamond is typical of those attributed to the Rio Formoso that are illustrated on-line. It is not quite identical, but it may be from another site in the river’s drainage basin. Or, it may not be from the Rio Formoso at all, as I already suggested. Other specimens, however, might be so contrary to geological expectation for the region, as to rock type and diamond habit, that they would clearly be suspect. Probably no knowledgeable person would attempt to manufacture a fake diamond in conglomerate using a diamond from Africa or a matrix unlike any previously reported from the state of Bahia.

A legitimate diamond in conglomerate would likely extend into the specimen--the diamond should be in conglomerate, not on it. The diamond should not be alone, on top of a matrix that contains all of the other framework grains, as if the diamond is somehow special. It is part of the rock. It is not a mineral that grew the way calcite crystals grow from the side of a vein or like zeolites grow in a basaltic amygdule. Likewise, the diamond should not be squeezed in among several other framework grains that all seem to be similarly separate from the matrix rather than being part of the whole rock.

Every grain of similar size in any conglomerate, regardless of mineral composition, should be similarly situated, because every tiny parcel of the rock, including perhaps many hundreds of grains, was deposited at the same time and under the same conditions as other framework grains. All of the larger grains are surrounded, at least partially, by finer-grained matrix. Even after having been cleaned by a preparator, matrix should still be visible around the diamond. Diamond-to-matrix contacts and “matrix-to-other-mineral” contacts should all be similar. If the diamond appears to be sharply demarcated from matrix, other framework grains should also be sharply demarcated–not just the diamond or a small group of large grains in contact with the diamond.

Textural details that would seemingly defy production by mortal humans, such as the meniscus-form borders between matrix and framework grains, provide evidence that the diamond was not positioned by human hands. Be suspicious, however, if clays bordering the diamond are different from clays scattered throughout the remainder of the specimen. All framework grains should be embedded in similar material as the diamond. In absence of XRD or other analytical data, simple inspection should be adequate to decided between “similar” and “different.” Cracks and pits in the diamond surface, if any, should be filled with the same kind matrix as found elsewhere. Thus, the groove in my diamond, as a natural product, is packed with the same kind of clayey material that forms the greater part of the matrix, and has the same kind of meniscus-like borders with the diamond surface.

Test the point of attachment of diamond to matrix using a fluorescent light and/or hydrocarbon solvent. The attachment should be through matrix that is the same as matrix elsewhere in the specimen, and none of that should fluoresce, nor should it dissolve in the solvent. Try water, too, in case the suspicious diamond is attached or surrounded by clay minerals of different composition than the matrix (red flag!), or water-based glue (e.g., Elmer’sTM).

Finally, there should be no spilled-and-dried glue on the specimen, even if it is not in direct contact with the diamond. Always be wary of dried glue anywhere on your diamond in conglomerate specimen! This last point seems to go without saying, and I haven’t mentioned it so far. But I once saw a specimen of diamond in conglomerate from Brazil offered in an on-line auction in which splotches of what appeared to be dried glue covered a part of the matrix within a centimeter or two of the diamond. Don’t bid on such a specimen!

References Cited

A. Published references available without charge on the Internet (except the Rock Color Chart):

Battilani, G. B., Gomes, N. S., and Guerra, W. J., 2007, The occurrence of microdiamonds in Mesoproterozoic Chapada Diamantina intrusive rocks – Bahia / Brazil: Anais da Academia Brasileira de Ciências, v. 79, no. 2, p. 321-332; available on-line at

Geological Society of America, 1991, The Geological Society of America Rock Color Chart with Genuine Munsell Color Chips: Rock-Color Chart Committee, Geological Society of America, 9 p.

Kaminsky, F. V., Zakharchenko, O. D., Khachatryan, G. K., and Shiryaev, A. A., 2001, Diamonds from the Coromandel area, Minas Gerais, Brazil: Revista Brasileira de Geociéncias, v. 31, no. 4, p. 583-596; available on-line at

Pedreira, A. J., 2001, Serra do Sincorá, BA, in Schobbenhaus, C., Campos, D. A., Queiroz, E. T., Winge, M., and Berbert-Born, M., eds., Sítios Geológicos e Paleontológicos do Brasil, article 085; available online in English at

Pedreira, A. J., and Waele, B. de, 2008, Contemporaneous evolution of the Palaeoproterozoic–
Mesoproterozoic sedimentary basins of the São Francisco–Congo Craton: Geological Society of London, Special Publications, v. 294, p. 33-48; available on-line at

Shirey, S. B., and Shigley, J. E., 2013, Recent advances in understanding the geology of diamonds: Gems & Gemology, v. 49, no. 4; published on-line at

Süssenberger, A., Brito Neves, B. B. de, and Wemmer, K., 2014, Dating low-grade metamorphism and deformation of the Espinhaço Supergroup in the Chapada Diamantina (Bahia, NE Brazil): a K/Ar fine-fraction study: Brazilian Journal of Geology, v. 44, no. 2, p. 207-220; available on-line at

Svisero, D. P., Shigley, J. E., and Weldon, R., 2017, Brazilian diamonds: A historical and recent perspective (peer-reviewed article): Gems & Gemology, v. 53, no. 1; published on-line at

B. Websites:

Clay Mineral Society, 2017,

Geografia Geral E Do Brasil:

Hunterian Museum Geology Collections, photos of diamond in conglomerate attributed to Rio Formoso:

Mindat Messageboard discussion: Diamonds in Conglomerate from Brazil?,55,141278,211944.

Article has been viewed at least 991 times.


Thanks for a very interesting and thorough work.
Why are there no garnets in the conglomerate? Apart from the ilmenite there isn't a lot of high density material here as one might expect from a pot hole.

Rob Woodside
26th Sep 2017 7:00pm

Good question! Unfortunately, I don’t have a good answer. The most authoritative discussion I could locate on short notice is in a master’s thesis by Cláudio Meira de Andrade, completed in 1999 at the Instituto de Geociências, Sao Paulo, Brazil, titled “Aspectos mineralógicos, geológicos e econômicos de diamantes e carbonados da Chapada Diamantina, Bahia.” Do you read Portuguese? Here are some pertinent passages from the English-language abstract:

“Diamond and carbonado have been prospected from quaternary sediments made up of sands and gravels as a result of the erosion of the Tombador Formation rocks. . . . The heavy minerals associated with diamond and carbonado in the region is comprised of magnetite, ilmenite, hematite, tourmaline, rutile, zircon, hornblende, epidote, kyanite, andalusite, almandine garnet, corindon, crysoberil, staurolite and gold. . . . Kimberlite indicators such as pyrope garnet, Mg-ilmenite, Cr-spinel, Cr-diopside and zircon are absent in the area. . . . The origin of diamond is still a controversial subject due to the lack of systematic surveys in the area. The absence of kimberlite indicators such as pyrope garnet, Mg-ilmenite, Cr-spinel, Cr-diopside and zircon as well suggests that the primary source rocks may have been eroded or are covered now by sediments of the platform.

I am puzzled by the last sentence, as diamonds are present in the Tombador, in spite of the erosion and/or covering by younger platform sediments. So why not also some kimberlite indicators? Perhaps they are simply as few and far between as the diamonds, and we already know that they must be (or were) in the area by virtue of finding the diamonds, so why spend a lot of time looking for them?

I have not run across any papers on sedimentary petrology of the Tombador Formation or heavy minerals in stream alluvium in Bahia to help us. I suspect there aren’t any, but maybe someone out there knows of some. If so, please let us know!

Norman King
28th Sep 2017 1:34am
Thanks Norman. If the diamonds came from African Pipes while Africa and South America were still attached and before the rise of the Andes, I doubt there would be any Kimberlite indicators. If they did come from Brazilian pipes the indicators should be there, unless the diamonds eroded out of the pipes long before the conglomerate formed. There's a lot more garnet in a pipe than diamonds so the latter is unlikely.

I too have a Rio Formoso specimen. The diamond in matrix has survived poking with a red hot pin, sitting for days in water, acetone and alcohol respectively. The only thing I haven't done is a thin section see that it is real conglomerate and not concrete. My specimen is a twin of yours and oddly has a fleck of gold but very few of the minerals listed in the thesis.

The diamonds are supposed to occur in Potholes and the photo of Rock's specimen appears to be one. However our pieces are a limonite cemented conglomerate. They must be from potholes as finding a diamond in uneroded conglomerate would be hopeless without alluvial pothole concentration. There might be some unmapped eroding conglomerate along the Rio Formoso supplying diamonds to the potholes. On the other hand there might be enough time in the last 100 million years for the diamonds washing around in Brazil to have ended up in Rio Formoso's catchment area. Your suggestion that they come from a disguised protected area is simpler and quite credible.

Rob Woodside
3rd Oct 2017 10:05pm

In order to leave comments to this article, you must be registered
Mineral and/or Locality is an outreach project of the Hudson Institute of Mineralogy, a 501(c)(3) not-for-profit organization.
Copyright © and the Hudson Institute of Mineralogy 1993-2017, except where stated. relies on the contributions of thousands of members and supporters.
Privacy Policy - Terms & Conditions - Contact Us Current server date and time: October 20, 2017 13:46:19
Go to top of page