General"Tantalites" from Afghanistan
We have in the database 75 photos of manganotantalites fom Afghanistan, no of them were analyzed.
Today I received 4 crystals labeled as "Tantalite Afghanistan".
Of course I analized them quantitatively. Three of them turned out manganocolumbites, one - high-manganoan ilmenite.
On this photo ilmenite crystal is in hand.
Two columbite crystals (central at the first photo) are low tantaloan ~9-13 mas.% Ta2O5 with Mn>Fe (Mn:Fe ~ 3:2), one columbite (the leftmost) is pure manganoan iron-free with elevated contents of Ta - ~20-25 mas.% Ta2O5.
And this isn't the first similar results. I had analyzed 3 samples of "Tantalites from Afghanistan" and before. Only one of them was confirmed - bright orange-red crystal similar to this one https://www.mindat.org/photo-109822.html in colour. Dark-coloured cherry-red crystals also turned out manganocolumbites.
So 5 of 6 analyzed by me personally "tantalites" from Afghanistan turned out manganocolumbites with Ta < Nb even in mas.%, not only in at.%.
This isn't only Afghan problem - total analyzing of my "tantalites" from Ural, Kazakhstan, Brazil, Australia and Madagascar (look for example https://www.mindat.org/photo-114523.html) gave quite similar results. Tantalites with Ta>Nb in at.% are much more rare, than hopes most of collectors and thinks most of dealers. I suppose, that most part of unanalized tantalites in collections are manganocolumbites in reality.
Hello Pavel,
Maybe the unanalyzed should be changed to Columbite-(Mn) - Tantalite-(Mn) series, pending analysis. Do you think the orange-red color is characteristic of Tantalite-(Mn)? 
The red color depends on Mn:Fe, as in the wolframite series. I don't think Ta:Nb has any effect on the color.
However, one can get a good indication of Ta:Nb by measuring the density! (Fe:Mn has very little effect on the density.) May be not hue of red, but its brightness. Dark coloured samples with cherry-red inner reflections usually turned out mangancolumbites, light coloured (even brown) samples are mangantantalites.
Problem is in the fact, that Ta is significantly heavyer than Nb, so comparatively low contents of Nb2O5 in mas.% give predomination of Nb over Ta in at.%. As result analyses even with Ta2O5 > Nb2O5 in mas.% able to belong to columbites. But in old literature such samples able to be wrongly mentioned as tantalites. Alfredo Petrov Wrote:
-------------------------------------------------------
> The red color depends on Mn:Fe, as in the
> wolframite series. I don't think Ta:Nb has any
> effect on the color.
>
> However, one can get a good indication of Ta:Nb by
> measuring the density! (Fe:Mn has very little
> effect on the density.)
In columbites-tantalites color more depends on Fe2+/Fe3+ ratio than on Mn2+/Fe2+ one as in wolframites is.
Columbites-tantalites containing elements with valencies other than 2+ and 5+ (Ti4+, Sn4+, Sc3+, W6+) are usually dark-coloured (black), because charge compensation reached by changing of Fe oxidation state. Columbites usually are enriched by Ti4+ in comparison with tantalites. This is the reason why columbites usually are darker than tantalites.
In other words, columbites-tantalites with monovalent iron are more light-coloured than ones with heterovalent iron.
And which help able to do measuring of dencity of such https://www.mindat.org/photo-171799.html samples? I will have to get my tantalite-Mn from Afghanistan analyzed. Will let you know what I find. Please give its photo here - I'll try to make prognosis. ;-) I will post a picture shortly. Here is the tantalite-mn from Afghanistan that I plan to get analyzed.
You must also consider the crystallography. E.g. your tantalite or columbite may in fact be tetragonal tapiolite-Fe/Mn. Hello Travis,
Cleavage will tell them apart as well. Tapiolite has none. Or at least no cleavage is reported for tapiolite. It's hard to tell from "tapiolite" photos as well, with only one being analyzed by XRD..
https://www.mindat.org/photo-682171.html
It looks like irregular fracture to me, and no cleavage... Reiner Mielke Wrote:
-------------------------------------------------------
> Here is the tantalite-mn from Afghanistan that I
> plan to get analyzed.
>
>
And without analyze I tell you that this will be almost Fe-free Ta-dominant member with 95% of probability. :-) Hello Pavel,
I hope you are correct. I hope to have some EDS results in a couple weeks. It is interesting to note that the Tanco Tantalite-Mn https://www.mindat.org/photo-427243.html is completely opaque and sub-metallic black with no trace of red internal reflections even on thin splinters. It has an average of 13.33 % MnO, 2.45% FeO, 58.61% Ta2O5 and 21.87% Nb2O5. There is significant TiO2 1.46% and SnO2 0.62%. The Mn:Fe ratio is 3.33:0.61 and the Ta:Nb ratio is 4.70:2.92. This manganotantalite https://www.mindat.org/photo-764085.html has composition (Mn0.92Fe0.08)1.00(Ta1.18Nb0.82)2.00O6.00 with Ta2O5 content 58.60 % and Nb2O5 24.50 %, 15.51 % MnO and 1.39 % of FeO, no Sn and Ti. The mineral is black, submetallic and opaque.
Compare it with manganocolumbite https://www.mindat.org/photo-836628.html (composition given) from the same locality, which is visually indistinguishable on a photo.
But we talking here about analyzed samples. Most of Afghan tantalites in the database weren't analysed. And even their SG wasn't measured. While almost all analysed samples from there are turned out manganocolumbites. 
The following article may be interesting in this context.
Martin & Wülser 2014, Niobium and tantalum in minerals: Siderophile, chalcophile or lithophile,
and polyvalent. Journal of Geochemical Exploration 147, 16–25.
Abstract:
Most geoscientists are convinced thatminerals of both niobium and tantalum contain those elements in the 5+
state only. In fact, there are several well-documented minerals containing Nb0, Nb2+, Nb4+ and Nb5+, and
tantalum follows a similar pattern. Both elements can exhibit a siderophile, a chalcophile and a lithophile character,
as well illustrated empirically, and as documented via XANES and XPS spectra. Of the XPS spectra that
we report here on ferroniobium of metallurgical grade, pyrochlore, samarskite, niocalite, synthetic edgarite
(FeNb3S6) and NbTe2, only the samarskite and niocalite contain only one valence state of Nb (5+). Vladimir
Vernadsky understated the situation when he placed Nb and Ta in the “dispersed” category. Columbite-group
minerals aremost unusual in exhibiting variable mutual disorder involving (Nb,Ta) and (Fe,Mn). Oddly, samples
must be heated to become ordered. We raise the possibility that at the stage of primary crystallization, the
relevant magmas contained mixed valences of Nb, Ta, Fe and Mn, which made a disordered distribution along
the octahedrally coordinated cations energetically more acceptable. Owing to the relative insolubility of members
of this solid-solution series in the ambient aqueous fluid, they have remained metastably frozen in a partly
or completely disordered state.
cheers "Owing to the relative insolubility of members
of this solid-solution series in the ambient aqueous fluid, they have remained metastably frozen in a partly
or completely disordered state."
I don't understand this statement considering that they say it "must be heated to become ordered." So what does solubility have to do with it? Can anyone explain this to me? Thank you. It seems to me, authors of the article has problems with logics and confusing warm with moist.
But from most common considerations disordered members of columbite-tantalite series containing same elements in polyvalent states should to be more dark-colored and less transparent than ordered and monovalent ones. Hi Reiner,
I am also stumbled about their the last phrase. But in it isn't any criminal. They say, only that due to insolubility of tantaloniobates, their disordered crystals survives in geological environment staying disordered (what is truth for many other metastable minerals).
Solubility/insolubility, ordering/disordering, heating/cooling aren't connected here, but style of authors is heavy and muddled enough. Reiner,
I wanted to refer to question 6 on this link
http://www.minsocam.org/msa/Special/Pig/PIG_CQ/PIG_CQ.html
but it was removed, I have inquired why.
well, if I understand it rightly, the precense of hetero-valent Fe and Nb species in the melt-fluid and their relative solubility in it, causes them to crystallize (into the columbite structure) with different degree of order. Obviously the disorder is higher at higher temp. and this disorder is thus frozen metastably into the columbite structure. But to gain order in these disordered columbites researchers heat them. Some authors make a big thing out of this "to get order at a higher temp" but as I see it it is just an energetic push over the metastable treshold.
cheers Pavel Kartashov Wrote:
-------------------------------------------------------
> But from most common considerations disordered
> members of columbite-tantalite series containing
> same elements in polyvalent states should to be
> more dark-colored and less transparent than
> ordered and monovalent ones.
This is what I was aiming at, and at least partly explain dark, seemingly stoichiometric FeNb2O6 - thus (Fe2+,Fe3+,Nb4+)(Nb5+,Fe3+,Nb4+)2O6
cheers The way this is presented suggests that ordered columbite and tantalite crystals do not exist in nature. Is that correct? However if they do and if solubility is a factor then would not ordered crystals crystallize first since disordered crystals should have a slightly greater solubility than the ordered ones? The order of crystallization would then suggest that disordered ones form last and therefore at a lower temperature but that is not consistent with what is generally known about disordered crystals. However if one assumes that ordered crystals are more soluble than disordered then it all makes sense but it raises the question of why ordered crystals would be more soluble than disordered ones?
Thank you Johan and Pavel for your help. Reiner, you have to read up on old papers by Cerny, Ercit, Wise etc. to find out about exact frequency of ordered vs. disordered, it varies.
The "solubility discussion" goes back to solubilities of the different elements, e.g. Nb vs. Ta, in the melt-fluid, and was originally presented as an explanation to observed fractionation trends. Nb (and Fe) had less solubility, hence crystallized before Ta (and Mn), thus ferrocolumbite compositions are observed earlier than manganotantalite in pegmatites.
Assuming that everything else "works", i.e. that di-valent Mn and Fe allways goes into the A-site and penta-valent Nb and Ta into the B-site, they would be completely ordered (by crystal radius size), I assume that should be the case at standard conditions, when Nb is present in its most oxidized state.
The observation of disorder is empirical and that it could be caused by the presence of Fe3+ and Nb4+ was first suggested by Martin.
I send the full article
cheers 
This dark red sample from Paprok has not been analyzed but the SG of the whole sample including albite is 6.65 so density of red stuff must be somewhat over 7. Is this enough to confirm Tantalite-Mn?
Hello Pavel,
You were correct it is very low Fe Tantalite-Mn, see attached
Yesterday I received and analyzed the seventh "tantalite" from Afghanistan.
I took portion of powder from surface of the crystal, and analyzed about 12 microparticles, only one of them shows predomination of Ta over Nb (spectrum #5). The most of them contains ~36 mas.% Ta2O5 vs 48 mas.% Nb2O5 what is equal to 2.2:1 ratio of Nb:Ta in atomic %.
Empirical formulae calculated from these spectra are:
sp#1 Mn1.00(Nb1.38Ta0.62)2.00O6.00
sp#6 (Mn0.97Fe0.03)1.00(Nb1.17Ta0.83)2.00O6.00
sp#5 Mn1.00(Ta1.03Nb0.97)2.00O6.00
All particles were very low-ferroan (not more than 0.45 mas.% of FeO), what making the mineral obviously red, but don't making it tantalite. Note that these are analyzes of outermost zone of crystal, so its internal zones must to be more poor in Ta.
So score columbite:tantalite become 6:1. :-(
