Log InRegister
Home PageAbout MindatThe Mindat ManualHistory of MindatCopyright StatusWho We AreContact UsAdvertise on Mindat
Donate to MindatCorporate SponsorshipSponsor a PageSponsored PagesMindat AdvertisersAdvertise on Mindat
Learning CenterWhat is a mineral?The most common minerals on earthInformation for EducatorsMindat Articles
Minerals by PropertiesMinerals by ChemistryAdvanced Locality SearchRandom MineralRandom LocalitySearch 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
Mining CompaniesStatisticsThe ElementsUsersBooks & MagazinesMineral MuseumsMineral Shows & EventsThe Mindat DirectoryDevice Settings
Photo SearchPhoto GalleriesNew Photos TodayNew Photos YesterdayMembers' Photo GalleriesPast Photo of the Day Gallery

Glauconite status

Posted by Tom Mortimer  
Tom Mortimer February 22, 2019 06:42PM
The mindat glauconite page states: "IMA status:Approved, 'Grandfathered' (first described prior to 1959)". The given formula requires Mg as an essential element. My recollection (can't prove) was that an earlier mindat formula for glauconite did not require Mg. The webmineral formula for glauconite does not require Mg.
Looking at the IMA data base, it appears glauconite is not an approved species.
I have just EDS analyzed a Hurricane Mtn., Conway, NH "glauconite". No Mg was detected. (Instrument has very good light element detection - down to carbon). The standardless, polished grain EDS analysis suggested an APFU of:
K0.2Fe1.29Al1.10Si4.0O7.8 , normalized for four atoms of Si.
This Hurricane Mtn. glauconite has been accepted as valid by New England collectors for decades. It has appeared in several NH species lists in Rocks & Minerals. A mindat example is shown here: https://www.mindat.org/gallery.php?loc=5942&min=1710
A sample of a similar appearing Government Pit, Albany, NH "glauconite" gave a very similar chemistry.
Comments are solicited.
Tom Mortimer
Frank K. Mazdab February 22, 2019 07:18PM
You can't normalize micas to Si=4 because some of the Al may be tetrahedral as well (and even potentially some of the Fe, as Fe3+). As micas are O10(OH,F,Cl)2, they are most easily normalized to 11 O charge equivalent (=22), which requires a Fe3+/∑Fe assumption. A better but more complex way to normalize them is to normalize ∑(T+M) cations (so K, Ca, and Na are not included) iteratively to a value between 6 and 7 until a desired (=assumed) Fe3+/∑Fe is calculated and ∑T does not exceed 4. This latter method gives the flexibility of allowing for some dehydrogenation in the OH-site, if there's reason to suspect that.

Note that there is a big difference between simple detection and good quantification; hydrous minerals really shouldn't be quantified by standardless EDS, and as H2O goes up, quality and reliability go down.

Glauconite is essentially the K-deficient equivalent of celadonite just as illite is the K-deficient equivalent of muscovite; as such, it's not a mineral, per se, but a compositional space within the K-deficient [celadonite/ferroceladonite/aluminoceladonite/ferroaluminoceladonite] space, and so there can be Mg-, Fe2+-, Al-, and Fe3+-dominant varieties that I don't think are nomenclaturally differentiated.

Edited 3 time(s). Last edit at 02/22/2019 07:41PM by Frank K. Mazdab.
Alfredo Petrov February 22, 2019 07:41PM
I never was able to understand the distinction between glauconite and celadonite. Yes, I'm aware the definitions are a bit different, but was that difference really large enough to qualify 2 distinct species?
Seems to me that for practical purposes outside of laboratory settings, if the mineral occured in a sedimentary rock it was called "glauconite" and if in a volcanic rock it was called "celadonite".
Ralph Bottrill February 23, 2019 08:29AM
The last review of micas left a lot to be desired, as we were left with a lot of mineral names very heavily used in geology without good specific definitions, eg glauconite, brammallite, phengite, illite, sericite, celadonite etc.

Celadonite tends to be mostly used as a group name, with several end members defined, including a Mg-Fe3+ endmember of the same name. Ferroceladonite is the theoretical Fe2+-Fe3+ endmember (unsure how close other analyses get). Glauconite was considered a K-deficient “celadonite” and/or a series name. However celadonites and other micas can be alkali-deficient and I don’t know any formal IMA or other rules as to when you start giving them names like glauconite, illite, brammallite, etc. Consequently we get eg. the same fine grained white mica that can be variously described as muscovite, illite, phengite, sericite, muscovite-celadonite series etc. by different workers. So we really need to get our act together on some of these definitions. I guess one day glauconite will be reclassified into umpteen new end members.

Some photos have been uploaded to Mindat as illite and they look awfully like fairly well crystallised muscovite, but maybe they are K-deficient, I doubt they are well enough analysed to be sure. As Frank said, analysing hydrous phases like micas is a common problem for loss of K, Na etc, giving apparent deficiencies that may not be real, especially if poorly crystallised.

Most geologists and petrologists don’t care much for the mineralogical definitions and, as Alfredo indicates, call a green mica either glauconite or celadonite based on whether it’s in a sedimentary or igneous rock. Thus we would be reluctant to call the Hurricane Mtn material glauconite, but maybe it is.

Tom, it would be a good idea to get an XRD on your sample, it’s unusually K-poor despite the analytical warning above, and it’s possibly a smectite or mixed layer mineral?

PS I just redefined Brammallite as a variety of paragonite rather than illite as that’s the usual usage and the original definition, despite no recognition by th e IMA. I’m sure the last review considered it to be a Na analogue of illite, which itself is still variably defined as a group (of what?), series (to what?) or variety of muscovite, in various references.
Ralph Bottrill February 23, 2019 08:49AM
Also the whole origin of the alkali deficiency is interesting and arguable last I read. Some consider it a simple replacement of alkalis by hydronium ions, some consider it due to interlayers of smectites and other phases, some suggest alkalis are easily leached from the mica crystal surfaces and as the crystallites get smaller there is a corresponding decrease in the alkali contents. Perhaps all three are important?
Tom Mortimer February 23, 2019 06:29PM
My inquiry on this glauconite topic has produced many interesting and informative responses. Thank you.
I have EDS analyzed (qualitative & standardless quantitative) several candidate New Hampshire (my passion) glauconite specimens. Photos of the analyzed specimens and my EDS plots & data may be seen at:
https://mindatnh.org/Glauconite%20sheet.html (specimen photo for this subject inquiry)
https://mindatnh.org/species%20data/Glauconite%20data.html (standardless polished grain analysis)
https://mindatnh.org/Glauconite%20Gallery.html (additional candidate glauconite specimens & analyses)
As a species collector, I wish to apply the most accurate label possible to my specimens and catalog entries. I do like Alfredo’s suggestion: “if the mineral occurred in a sedimentary rock it was called "glauconite" and if in a volcanic rock it was called "celadonite".” (The mindat geologic map identifies Hurricane Mtn., Conway, NH as “Medium-grained mesoperthitic granite containing riebeckite and (or) hastingsite”)
It would seem that the best label for my Hurricane Mtn. specimens would be either celadonite, celadonite-group, or perhaps ferroaluminoceladonite.

While I do understand Frank’s professional mineralogist critique, I believe amateurs such as me can contribute meaningful input to the ever evolving science of mineralogy. Yes, it would be nice to have access to instruments to quantify water and hydroxyl content, and to have access to programs to iteratively compute charge balance. Nevertheless, I believe my preparation of polished grain samples for ZAF corrected, standardless, EDS analysis helps to advance the accuracy of New Hampshire mineralogy.
Tom Mortimer
Frank K. Mazdab February 23, 2019 08:14PM
Hi Tom,

I agree with you that amateurs can contribute meaningful input to the ever evolving science of mineralogy, and I think that's wonderful! I was an amateur once too!

But consider this. You've already gone to the trouble of making a polished grain mount (or you had a polished grain mount made). I also assume you don't have an SEM in your garage (but if you do, that's pretty cool!), and that you ship off or take your samples to some facility to have them run... perhaps to one of the commercial labs that I've seen advertisements for, or maybe to your local university?

What does it cost you to run one of your samples (assuming you really don't have your own SEM)? I've seen $50/sample on some of the ads. Well, here at the University of Arizona, running the electron microprobe costs $32/hour, capped at $360 per day (it used to cost just $17/hour, but due to low usage the university did the head-scratching thing of raising the price!?). Running the SEM here costs $40/hour (it's a different facility; hence the non-sensical price disparity). I don't know where in the country you live (I'm guessing NH?), but I imagine there would be one or more nearby universities that may offer analytical services to the public; we do here. These services aren't typically advertised... you'd normally go in to meet the electron microprobe lab manager, introduce yourself and what'd you like to do, and then arrange a day or two for you to run your polished mounts. We have an older retired gentleman here, a member of the public not affiliated with the university, who comes in weekly and runs analyses and makes element maps on meteorite specimens he's bought, and I think it's great... I've been urging him to put his remarkable accumulated dataset online for the whole world to benefit from.

Of course, if you have just one or two analyses to do, it wouldn't be time efficient and the lab manager would probably suggest you send your sample to a commercial facility. But if you can stockpile a dozen or a couple dozen grains of different minerals you'd like to analyze (they need not all be the same mineral type, but all silicates or all sulfides or all whatever would be helpful), you can put them on a minimum number of epoxy mounts (I can fit ~30 1 mm grains on a single 1" diameter mount) and spend a full day analyzing 50, 60, 70 unknowns by EPMA for the price of what some commercial labs might charge for only 7 samples of inferior EDS analyses. And it would be a great experience to interact with the people at the facility to increase your knowledge and enthusiasm.

If you aren't able to get to a local university yourself, perhaps consider contacting their mineralogy professor and see if there was an interested student who'd volunteer some of his or her time to do the analyses for you. The cost of the instrument time plus lunch for the student (and maybe a cool specimen or two as a perk) might be a reasonable and appreciated deal. I've done those sorts of projects myself periodically, and indeed even still do on occasion.

The irony is that making the polished mount is the biggest hurdle to doing electron microprobe work. If you have a pretty crystal in a rock and you aren't willing to chip off a discreet corner to mount up, then the SEM may be the only option. But once the sample is polished, a world of cool tools opens up to you, and they're really not as inaccessible as they might first appear.

In any case, I do recognize your interest in furthering the science and I admire that. And clearly you've taken the extra step beyond what many collectors are willing to do to even collect EDS data, and that's great too. EDS has its place, and for many minerals it can be an adequate tool. But when I see people here using it for hydrous minerals, minerals with the HREE (even the microprobe doesn't do well with those), minerals with F (even system that "detect" light elements well won't necessarily quantify them well), other element systems that suffer severe peak overlaps, and the like, it's simply not the right tool to use. And ironically it might not even be the least expensive tool to use. I cheer on everyone who wants to learn more about their samples and about the science. But as the old saying goes, sometimes a little knowledge can be a dangerous thing, and contributing subpar data may be worse for advancing the science than contributing no data at all. Your previous posts have indicated to me that you're indeed a thoughtful collector and that you really do want to better the science for your corner of the world... so, I hope you might at least give consideration if these alternative options I've suggested could be practical in your situation (and I similarly mention this for others here who may be reading this as well).

Tom Mortimer February 23, 2019 10:15PM
Responding to your post:
First, I consider myself moderately skilled in preparing epoxy embedded (using EDS recommended epoxy) polished grain samples. Photo of my polished grain block (prior to carbon coating) that contains the subject “glauconite” grain (labeled BC324) is attached. I can (and frequently do) mount two dozen or more grains on a 1.8 by 1.6 cm block. I have personally prepared about 30 of these multi-grain blocks. I utilize the Boston College SEM (Joel JSM 6340F) with an AMETEK Octane Plus EDS detector . The Micromounters of New England (MMNE) sponsors/funds these EDS sessions. By “I utilize” I mean myself and another MMNE member sit at the instrument, image the grain samples, and collect the EDS data. We are aware of the issues of peak over-lap (e.g. Pb, S, Bi, Mo).
This EDS analysis service is provided to members of the MMNE for samples of New England origin. I do the polished grain preparation for member submitted samples. Frequently our analyses turn up the un-expected, indicating the need for follow-up analysis. Another MMNE member has a Raman spectrometer (that has been validated by blind mineral sample testing) to provide another data point. I do use John Attard’s PXRD service when I think the value added is warranted.
Getting microprobe analysis of our grains with ambiguous EDS results is difficult. To my knowledge there are few microprobe instruments in New England (perhaps only one). Researchers with access/control over these instruments are usually heavily committed to their own research projects. I have had polished grain samples in a microprobe queue for over a year. My understanding is that water and hydroxyl quantification is done with a controlled heating oven and an extremely accurate scale to measure weight loss. How many facilities have these?
The AMETEK EDS instrument we use has capability for using standards. I believe I can advance my capability to use this. However, this will be time consuming, and presently the MMNE membership has a desire/preference for more analyses with a 90 to 95% confidence, than fewer analyses with a higher confidence. The results do provide an excellent pre-filter to indicate samples that would benefit from follow-up analysis. More often than not, members have a good idea of possible identifications (from physical characteristics and mineral environment). The EDS usually confirms their ID. We typically do 25 to 35 grain analyses per five our session.
Tom Mortimer

Edited 1 time(s). Last edit at 02/23/2019 10:34PM by Tom Mortimer.
Ralph Bottrill February 23, 2019 10:57PM
“It would seem that the best label for my Hurricane Mtn. specimens would be either celadonite, celadonite-group, or perhaps ferroaluminoceladonite.”
No, none of the above, we don’t even list a celadonite subgroup but probably should. The closest may be ferroceladonite, but I wouldn’t put a name to it without XRD or Raman analysis, K looks too low. But Frank is right, a good probe analysis would help lots.
Frank K. Mazdab February 23, 2019 11:39PM
Thanks for the additional insight, Tom,

I now have a better understanding of the set-up you have available, and it seems like a good arrangement.

Still, given that you're essentially already doing microprobe-style work on an SEM (that is, working on multi-sample polished mounts), you might nonetheless find it worthwhile to drop an email to Ken Domanik at the UofA microprobe facility (domanik[at]lpl.arizona.edu) and inquire if sending a mount out to Tucson (assuming it wasn't practical or cost-effective for you to come out yourself, although you'd be welcome to do that too), and then having him, or potentially me, or potentially a grad student volunteer, run it.

That might provide you with a possibly similarly cost-effective alternative to the BC SEM experience, but with data of a very different nature/quality? It'd then be just a matter to decide which method gives you the data that you want and the most bang for your buck.

It actually surprises me to hear that New England universities, many with their "reputation", would seem to be largely devoid of a basic modern petrologic tool like the electron microprobe (an instrument that has actually been around for more than 50 years), while dog-pile Tucson has two of them (admittedly purchased years ago with generous NASA money for the Lunar and Space Sciences Dept.). In any case, sadly neither instrument gets much use anymore (I'm one of the most prolific users and even I'm only there at best once a month anymore), so there'd be no long wait here. And I know Ken is actively soliciting more work for the lab (which, ironically, may help bring down user costs in the future). So, food for thought...

Tom Mortimer February 24, 2019 04:03PM
Ralph & Frank:
I will send off a sample for PXRD this week. Hopefully, this will bring some resolution to these specimens. I will report results to this thread when they become available.
I will follow up on your suggestion to contact UofA on microprobe analysis of our EDS pre-screened samples that indicate a more refined analysis is needed.
To all:
This exchange has been most helpful and informative. Thank you.
Tom Mortimer March 07, 2019 04:44PM
I have received PXRD results on two New Hampshire glauconite samples. However, before I report these results to this message board thread, I would like to resolve the chemistry question: Is magnesium an essential element for glauconite?
Mindat states the formula for glauconite as:
My understanding of mindat's mineral formula convention is that the (Mg,Fe2+,Fe3+) indicates Mg is an essential element, as it is listed first in this parenthesis group.
Webmineral.com, the RUFF database, and as near as I can determine, the IMA, give the glauconite formula as:
(K,Na)( Fe3+Al,Mg)2(Si,Al)4O10(OH)2 , suggesting that Mg is not essential.
So why is the mindat.org formula different from the sources I cite? What is the source of the mindat formula?
Tom Mortimer
Frank K. Mazdab March 07, 2019 06:42PM
Hi again Tom,

I thought a bit more about this after the original postings, and also went and took a look at Rieder et al., 1999, and come back to my earlier conclusion that glauconite is not a mineral but rather a family of related minerals within a composition space bounded by several end-member minerals, just as biotite is also not a mineral but is instead a family within a composition space. In fact, Rieder et al., 1999 shows the glauconite composition space in Figure 1. Rieder et al., 1999 uses initially "Mg" as a proxy for (Mg+Fe2+), but then further in the text and in Table 3 uses R2+ for Mg and/or Fe2+ and R3+ for Al and/or Fe3+, allowing for a substantial range in Fe2+, Fe3+, Al and Mg composition.

Anything along the Fe2+/(Fe2++Mg) join (as long as the other composition parameters were satisfied to keep it in the glauconite space) technically would still be in the glauconite space, so a Mg-free one is theoretically possible. However, I would be skeptical if such a composition could occur developed from the normal ferromagnesian mineral budget of a precursor rock... where would the Mg in the original rock go if not into the alteration sheet silicates? Some could presumably be removed the system, of course, but all of it? In fact, in many systems alteration ferromagnesian minerals end up more Mg-enriched than their precursors, because during the alteration process some Fe is oxidized and lost to accompanying magnetite and hematite, where Mg is not welcome.


Edited 1 time(s). Last edit at 03/07/2019 06:47PM by Frank K. Mazdab.
Tom Mortimer March 07, 2019 07:28PM
Your response was most informative, as usual.
I have to ask: Do you think the mindat.org formula is acceptable/correct for glauconite?
The mindat.org glauconite page states "Approved, 'Grandfathered' (first described prior to 1959)", indicating to me that glauconite is an approved species, not a family, or group name, as you suggest. Is mindat incorrect here?
Bishop et. al. in their 2008 article in Clay Minerals (March 2008, pgs. 35 - 54) include an analysis (Table 1) of a Hurricane Mtn. glauconite with a formula:
That is a pretty tiny bit of Mg, certainly much smaller than I can detect with the EDS instrument I am using (without averaging for an hour or so).
Ralph Bottrill March 07, 2019 08:35PM
The formula is not very good. For starters as it’s defined as alkali-deficient then that should be indicated in its formula. Also if it’s really a group then it needs to have a group formula of the type ( A,◻)DET4O10Z2, where A=K, Na, etc... But the trouble is it’s variably and vaguely defined as either a variety of celadonite, a series, subgroup or a compositional space. Illite and brammallite have similar problems, none of these should be considered true species and given specific formulae, unless we can actually find minerals of the type that are vacancy-dominated and not mixtures, and define those as new species. Personally I think they are best described as varieties if we don’t have vacant-dominated end members. But I guess epidote eg. is in a similar category. Anyway with a group it’s ok to have 0% of an end member, and a variety of a group like glauconite is the same, though some people would wonder if we should be giving formulae for varieties?

How did the XRD go?
Frank K. Mazdab March 07, 2019 08:40PM
Hi again Tom,

I don't really care for mindat's formula or the webmineral/RRUFF formula, because in neither case are the formulas differentiated from the celadonite family (i.e. for the K-deficient micas, [K+Na] should be between 0.6 and 0.85).

I'd prefer the Rieder et al., 1999 formula (after all, they have the IMA's imprimatur and there hasn't been a more recent update on this that I'm aware of [but there ought to be an update!]). I'd have to sit down for a bit and think about the actual formulas for the eight corners of the glauconite space (and it's only eight corners if Mg+Fe2+ is treated as one component... treat them individually and it becomes a real mess). In any case, you should check out their general formula for a K-deficient dioctahedral mica, given at the start of Table 3. Then one just has to decide if a mineral is a general illite, a general glauconite, or a general brammallite.

By the way, a closer look at that table actually specifically notes that Mg > Fe2+. I don't know if I necessarily agree with this edict (since then an Fe2+-dominant "glauconite" actually wouldn't have a name... kind of a big problem!), but it seems at least "officially", glauconite does indeed need to have Mg.

As for your Bishop et al., 2008 reference, I can't comment on it as I haven't seen it yet. But when I have some additional time I'll take a look. Obviously, I'd like to see what analytical methods they used to arrive at that formula. Also note that even if the analysis is good, just because they call that material glauconite doesn't mean the IMA would agree (case in point... clinoferrogedrite was fairly recently published in a mineralogical journal contrary to the IMA's amphibole nomenclature scheme, but in that example, the authors had the much stronger case!).

Tom Mortimer March 07, 2019 10:01PM
Here is the PXRD of the Hurricane Mtn., Conway, NH "glauconite".
Ralph Bottrill March 09, 2019 11:53AM
Thanks Tom, though the pattern is a bit deficient. With clays you really need to run them from about 4-62 degrees to see the most important peaks. I’m doubtful that you can confidently identify glauconite and sauconite from such an XRD pattern; the pattern could represent quite a number of smectite and mica group minerals, including celadonites. But at least the fact you have a mixture of this sort explains the odd composition. It could be a mixed layer clay.
Tom Mortimer March 09, 2019 09:39PM
Ralph, OK I have requested an extended 2 theta range plot.
Tom Mortimer March 13, 2019 05:29PM
Here is the PXRD (with extended 2-Theta range) for my Hurricane Mtn., Conway, NH.
My view is that this plot strongly favors a ferroceladonite identification, (the pale green reference spikes). The blue reference spikes are for saponite, a smectite group mineral.
My EDS analysis from this specimen contained only K, Fe, Al, Si, and O, consistent with ferroceladonite. Saponite requires a lot of Mg and some Ca, nether of which were found in my EDS analysis.
Comments solicited.

Frank K. Mazdab March 13, 2019 07:07PM
Well, as we've already covered here, glauconite (the ill-defined family) is really just celadonite (the composition space) with a K-deficiency, so their similarity in patterns is unsurprising.

The "evolution" from glauconite ("Cr-bearing", even) in the first run to ferroceladonite in the second run brings home Ralph's point that the analysis of clay minerals is not a trivial undertaking. The proper characterization of a clay is typically beyond the time and patience of even a non-clay-focused mineralogist. That might be why those clay dudes even have their own journal... lol.

That said, the K content from your original chemical analysis is still problematic:

all the celadonites (as "proper" micas) should have K between 0.85 and 1.00 apfu.
glauconite, as it's defined in Rieder et al., 1999, lowers that K range down to 0.6 (maybe 0.5 if you're generous with the dominant cation rule) to 0.85 apfu.
but there's nothing officially defined below that lower end, and certainly not all the way down to 0.2 apfu. That would typically be the realm of chlorites, vermiculites and smectites. Note that the 0.3 apfu Ca in saponite is just an interlayer cation, and so could conceivably be replaced by 0.2 to 0.3 apfu K for a "K-saponite" (or a "K-ferrosaponite") as long as the charge is balanced elsewhere. But I'm not proposing we invent a new mineral, because... see above(!)... "analysis of clay minerals is not a trivial undertaking".

The immediate problem at hand is what to name your sample. "Clay" is a nice name... one of my friend's sons is named Clay and I think he likes the name. And here it might be both the most precise and the least precise choice, but definitely the least satisfying one (assuming the chemical analysis is to be believed... we've already covered reasons why maybe it considered with caution). If in your shoes, I'd probably go with the XRD result but hedge my bet with the chemistry data, opting for a compromise of "Fe-rich glauconite" (glauconite not being an end-member mineral anyway).
W. Richard Gunter March 13, 2019 08:00PM
It would be interesting to see if there was any expandable content with the "Fe-rich glauconite". This is another non-trivial method for the analyses of clay minerals.
Ralph Bottrill March 13, 2019 09:05PM
Ok this latest pattern is more informative. There is a huge peak just below 6 degrees, typical of smectites, though partly truncated. There is a much smaller double peak between 8-9 degrees, indicating some minor mixed layer mica-smectite. So what fits your XRD and EDS analysis best is an aluminous nontronite, with minor intermixed mixed layer clays.
Frank K. Mazdab March 13, 2019 09:38PM
that's great Ralph! so my make-believe "K-ferrosaponite" wasn't such a fanciful guess after all. :-)
Ralph Bottrill March 14, 2019 08:59PM
That’s a pretty good educated guess Frank. But the original analysis shows a high Si/(Fe+Al) ratio indicating a dioctahedral rather than trioctahedral smectite, suggesting the iron is quite oxidised. The formula is probably something like K0.2(Fe1.1,Al0.9)(Si3.8Al0.2)O10
Frank K. Mazdab March 15, 2019 01:31AM
Hi Ralph,

From a purely chemical perspective, the formula can actually be normalized to an all Fe2+ composition, an all Fe3+ composition, or to any Fe3+/∑Fe ratio in between. That's the unfortunate reason sheet silicate normalizations are almost always normalized on an ∑(-) basis (with a guess for Fe3+/∑Fe) rather than on what I consider the preferred cation basis (but where you have to guess □ instead)... there's no single answer without knowing the Fe valence by an independent technique.

For example, the all Fe2+ version would be:

and the all Fe3+ version would be (akin to your formula):

It's true, however, that the all, or nearly all, Fe3+ version is purely dioctahedral, whereas the all Fe2+ version requires a reduction in M-site vacancies and results in something in-between dioctahedral and trioctahedral (admittedly something almost never observed at least in true micas outside of pretty limited ranges on either end). So no doubt you're correct.

I've always found it curious that there wasn't greater solid solution between dioctahedral and trioctahedral micas; chlorites and other sheet silicates with interlayer sheets don't seem to be limited in the same way, at least between the disparate oct-layers.

In any case, all chemistry and apparently the XRD as well point to a smectite group mineral, so the celadonite or even glauconite seem no-gos.
Tom Mortimer March 15, 2019 02:28PM
It appears the mindat glauconite page has undergone a substantial update after my initial posting on this topic on February 22, 2019. The change in the glauconite chemistry from the Feb. 22 and earlier formula: (K,Na)(Mg,Fe+2,Fe+3)(Fe+3,Al)(Si,Al)4O10(OH)2 , (indicating Mg is essential), to the present one: (K,Na)(Fe3+,Al,Mg)2(Si,Al)4O10(OH)2 , leaves readers of this thread questioning my sanity. Also the statement "IMA status: Approved, 'Grandfathered' (first described prior to 1959)" has been removed. It would be most helpful if mindat included a revision date on their species pages. I, (and I would suspect many others), use mindat as a reference for their research, so clear indications of revisions are important!
My .jpg of the extended 2-Theta PXRD is quite fuzzy, so I am adding a PDF of that analysis here. I am also adding a photo of my topic specimen, so chasing my reference link is not needed.
I expect to get a Raman spectrum from a Hurricane Mtn. "glauconite" this week. I am uncertain how much "value added" this will provide, but it will be another data point.

Frank K. Mazdab March 15, 2019 07:09PM
LOL Tom, no one's questioning your sanity. I think users who look through the posts here and observe that a formula you'd mentioned in your original post is now different would just assume that changes were affected through the discussion that evolved from your original post. That's mindat in action... one of the advantages of a dynamic website over a static book.

As for the particular changes here, although I'd thought about adding some text in the main description noting the compositional limits of Rieder et al., 1999, I actually never got around to doing it. Personally, I'd probably still prefer the formula in Table 3 of Rieder et al., 1999 over either the old or new mindat versions, even though the Rieder et al., 1999 one is a bit unwieldy. It seems to delineate the overall range in composition pretty well, and definitely gives the reader a feel that glauconite really a compositional space rather than just some fuzzy end-member.

I hope when the IMA Committee on Changing Mineral Names Just to Annoy Everyone plans their next working group meeting in Tahiti or wherever, that they can specifically re-visit the micas. It's been 20 years since Rieder et al., 1999 and there are still some vaguely unsatisfying parts of it (like boundaries for the names in the "biotite" space). There's a mineral group [the micas], much more so than amphibole 5.0, that can use a thorough 21st century re-working.
Ralph Bottrill March 18, 2019 09:23PM
Frank, yes I was a bit simplistic there, of course there will be both ferrous and ferric iron present and it’s one of our big bug- bears that’s this is so rarely determined. But it seems most people working on micas and clays assume either dioctahedral or trioctahedral and I’m not aware of studies showing whether this is totally valid, but it usually works. I’m actually working on a project at present looking at detailed mica compositions, where we got some hand picked micas analysed for FeO/Fe2O3. But of course when we study these concentrates we mostly find some smectites, chlorite and other mineral inclusions. This of course correlates with nominal alkali-deficiency, making us question the validity of illite and glauconite. It’s a tricky one.

I totally agree that micas and clays are well overdue for a detailed new review. I have made a few changes and comments on some of these mineral pages, though I’m not really an expert on them. I didn’t change the above formulae, and agree with you about using a more generic formula for glauconite, it shouldn’t indicate either Mg of Fe dominance.
Sorry, only registered users may post in this forum.

Click here to login

Mineral and/or Locality  
Mindat.org is an outreach project of the Hudson Institute of Mineralogy, a 501(c)(3) not-for-profit organization. Public Relations by Blytheweigh.
Copyright © mindat.org and the Hudson Institute of Mineralogy 1993-2019, except where stated. Mindat.org relies on the contributions of thousands of members and supporters.
Privacy Policy - Terms & Conditions - Contact Us Current server date and time: March 18, 2019 22:05:27
Go to top of page