Mindat Mineralogy Messageboard

Marco E. Ciriotti

Forthcoming new mineral: filatovite
January 25, 2004 08:23PM

<HTML>L.P. VERGASOVA, S.V. KRIVOVICHEV, S.N. BRITVIN, P.C. BURNS, V.V. ANANIEV
Filatovite, K[Al,Zn)2(As,Si)2O8], a new mineral species from the Tolbachik volcano, Kamchatka peninsula, Russia.

European Journal of Mineralogy, 16 (Forthocoming)

IMA 2002-052</HTML>

Reply Quote

Peter Haas

Re: Forthcoming new mineral: filatovite
January 26, 2004 02:53AM

<HTML>What means (As,Si)2O8 ?

The maximum oxidation number of arsenic is +5, so the theoretical sum formula of the ortho-oxoacid would be H5AsO5, and a corresponding di-orthoarsenate would have the formula As2O9 8-. Actually, H3AsO4 is considered as ortho-arsenic acid, because its hydrate does not exist in nature. As far as I know (i.e. without consulting the last edition of HoWi), the same holds for salts with AsO5 5- anions. The dimer of H3AsO4 is H4As2O7, so one could think of As2O8 x- only as a peroxodiarsenate (x = 4). I severely doubt that this anion should exist. Of course, the same problem consists with Si2O8 x- ...

[AsO4, SiO4]2 would make a lot more sense, but there is still a problem with the charge balance.</HTML>

Reply Quote

Marco E. Ciriotti

Re: Forthcoming new mineral: filatovite
January 26, 2004 08:23AM

<HTML>Filatovite is a member of feldspar group.
It is monoclinic:
C2/c 13.416 13.370 8.772 90.00 100.67 90.00

Filatovite formula K[Al,Zn)2(As,Si)2O8] to compare with orthoclase K(AlSi3)O8, rubicline (Rb,K)(AlSi3)O8, buddingtonite (NH4)(Si3Al)O8, celsian Ba(Al2Si2)O8, anorthoclase (Na,K)AlSi3O8, hyalophane (K,Ba)[Al(Si,Al)3]O8, microcline K(AlSi3)O8, and sanidine (K,Na)(AlSi3)O8 formulae.</HTML>

Reply Quote

Jim Ferraiolo

Re: Forthcoming new mineral: filatovite
January 26, 2004 11:14PM

<HTML>Marco,

There is a second paper for filatovite listed:

S.K. FILATOV, S.V. KRIVOVICHEV, P.C. BURNS, L.P. VERGASOVA.- Crystal structure of filatovite, K[Al,Zn)2(As,Si)2O8], the first arsenate of the feldspar group.

What the end-member composition should be is anyone guess until the paper comes out, but ordered Al Zn would give
K[AlZn]As2O8 [assuming +5 for As, of course]

Jim</HTML>

Reply Quote

Peter Haas

Re: Forthcoming new mineral: filatovite
January 27, 2004 02:18AM

<HTML>Sorry, but that's still incorrect. As2O8 x- is not an existing anion, even if the formula is considered as a 'net' formula (i.e. representing the relative atomic composition of a polymeric anion). However, (AsO4)2 would be correct. This should not be summed up to As2O8, since this is simply not the same (as, for instance, 2 H2O is not equal to H4O2).

Regards
Peter</HTML>

Reply Quote

Peter Haas

Re: Forthcoming new mineral: filatovite
January 27, 2004 02:22AM

<HTML>Concerning the charge balance: if Al3+ is dominant, AsO4 (3-) cannot be dominant at the same time. It makes more sense to write the formula as follows: K(Al,Zn)2[SiO4,AsO4]2</HTML>

Reply Quote

andy christy

Re: Forthcoming new mineral: filatovite
January 27, 2004 02:34AM

<HTML>As in the feldspars, the anion is an infinite tetrahedral 3-D framework with 4:8 cation:oxygen ratio. You should not really be thinking in terms of the monomeric [AsO4]3- etc species that occur in solution or in normal arsenate minerals. All the Zn, Al, Si and As are part of the tetrahedral framework here. K is the only weakly-bound interstitial cation. Potenbtial charge-balanced versions of the filatovite formula can be derived from K-feldspar K[AlSi3O8]by couple substitutions:

Zn2+ + As5+ for Al3+ + Si4+

and also

Zn2+ + 2As5+ for 3Si4+

or

2Zn2+ + As5+ for 3Al3+

giving

K[ZnSi2AsO8] , K[ZnAlAs2O8], K[Zn0.67Si3As0.33O8] respectively.

It looks from the authors' rendition of the formula as though low-valence (Zn,Al) and high-valence (Si,As) are fractionated, implying (Zn+Al)=(Si+As) and that the dominant endmember is in fact K[(ZnAl)(SiAs)O8].</HTML>

Reply Quote

andy christy

Re: Forthcoming new mineral: filatovite
January 27, 2004 02:39AM

<HTML>Correction: dominant endmember K[(ZnAl)As2O8].

Essential mixed occupation of only one site, a la Frank Hawthorne in Canadian Mineralogist.

the Si is presumably subsidiary to As, incorporated by some substiotution towards K=spar and K[ZnSi2AsO8].</HTML>

Reply Quote

Jim Ferraiolo

Re: Forthcoming new mineral: filatovite
January 27, 2004 02:28PM

<HTML>Thanks, Andy.

I believe that's what I suggested.

Jim</HTML>

Reply Quote

Peter Haas

Re: Forthcoming new mineral: filatovite
January 27, 2004 08:57PM

<HTML>From a mathematical point of view, there is a charge balance. Fine !
Unfortunately, it's impossible to substitute the majority of Si atoms in the three-dimensional network by As, because this makes the whole structure pretty unstable. Actually, the substitution of only one Si atom by an As atom will already break up the network structure. I will explain that below.

Start with this one: take a piece of paper and draw the As2O3 structure. (As2O3)x forms sheets with the As atoms in a psi-tetrahedral configuration and oxygen bridges between them (i.e. lone pairs remain on the As atoms pointing below and above the plane of view alternatingly). To obtain a three-dimensional network, you'll only have to involve the lone pairs in bonds, corresponding to an oxidation which leaves the As atoms with an oxidation number of +5 and, consequently, a positive charge (so-called arsonium ions; due to the s2p3 configuration of the group V elements, an oxidation number of +4 becomes impossible). The repellent forces between neighbouring atoms will make the structure unstable.
However, it is not only the charge; arsonium ions are extremely strong Pearson acids, that is, they readily react with even a weak nucleophile such as chloride or fluoride - not to talk about oxide or hydroxide ions which are much stronger nucleophiles. In the reaction, one bond is cleaved by a nucleophilic attack and the network breaks up.
Stable arsonium ions are not known in inorganic chemistry (there are a couple of inorganic chemistry textbooks adressing such particular structures - see Cotton/Wilkinson or Huheey, for example). The only way to stabilize these ions is to introduce substituents that are not cleaved by a nucleophilic attack, such as alkyl residues. Bulky substituents such as tert.butyl provide a maximum stabilization due to sterical hindrance in the attack of appropriate reactands ('kinetic stabilization'). For instance, tetraalkylarsonium hydroxides can prepared from non-protic, low polarity solvents. Some of these compounds have very interesting IR spectra.</HTML>

Reply Quote

Jim Ferraiolo

Re: Forthcoming new mineral: filatovite
January 28, 2004 01:11PM

<HTML>Looks like we'll have to wait for the description, analysis and structure.

Jim</HTML>

Reply Quote

John Dagenais

Re: Forthcoming new mineral: filatovite
January 28, 2004 01:52PM

<HTML>I second that emotion.</HTML>

Reply Quote

andy christy

Re: Forthcoming new mineral: filatovite
January 28, 2004 02:49PM

<HTML>Reply to Peter Haas, 01-27-04, 20:57:

Peter wrote: "From a mathematical point of view, there is a charge balance. Fine !
Unfortunately, it's impossible to substitute the majority of Si atoms in the three-dimensional network by As, because this makes the whole structure pretty unstable. "

???
There is *no* problem in replacing tetrahedrally coordinated Si4+ by tetrahedrally coordinated As5+. That's how titanite, CaTi[SiO4]O, and durangite, NaAl[AsO4]F, get to be isostructural!

Unfortunately, Peter appears to have become confused by some aspects of arsenic chemistry (admittedly complicated):

1. As3+ has nothing to do with filatovite, which is an arsenate containing As5+.

2. As5+ is quite stable in tetrahedral coordination by oxygen. The sigma bonding is sp3, and pi contributions bring the bond orders (=bond valences) up to 5/4 = 1.25. This is exactly as it is in any of the normal arsenate minerals. These are not particularly reactive as acids, cf. mimetite!

3. Arsonium compounds are those in which
(i)As5+ forms four bonds of bond valence 1 to more electronegative constituents, and retains a formal remnant charge of +1 on the arsenic, or
(ii) As3- (formal charge) forms four bonds to more electropositive ligands.

The former case is analogous to tetrachlorophosphonium, [PCl4]+.
The latter case is analogous to normal phosphonium, [PH4]+, or ammonium.

From the bond-valence point of view, the crucial differences from, say, arsenate(V) are:

1. There is negligible pi-bonding from As to its neighbours, which keeps these bonds fairly weak (they are very strong in arsenate).

2. In case (i) arsonium, the arsenic has some unsatisfied valence and must form additional very long and weak, primarily electrostatic bonds to more distant electronegative neighbours. These are the sort of bonds that keep the easily sublimable solid "PCl5" (= [PCl4]+[PCl6]-) together. In arsenate, the bonding capabilities of the As are fully satisfied by four very short, strong bonds to oxygen.

It is easy to see why arsonium compounds should be very reactive. Filatovite and other arsenates(V), however, are not arsonium compounds!

The closest to a bonding problem in a polysilicate/arsenate like filatovite is overbonding of the bridging oxygens, since the ideal bond valence to As is greater than 1. This problem can be alleviated by ordering cations such that oxygens bonded to As on one side preferentially bond to a lower-valence Al or Zn on the other side.

Incidentally:

1. bridging oxygens (2- and 3- coordinate) occur between As in the structure of As2O5 (not a mineral- too deliquescent) which has As in both tetrahedral and octahedral coordination.

2. polymerised As-O-Si chains are already known in the mineral tiragalloite, Mn4[AsSi3O12(OH)], which contains a linear tetramer with As at one end.</HTML>

Reply Quote

Jim Ferraiolo

Re: Forthcoming new mineral: filatovite
January 28, 2004 05:45PM

<HTML>Thanks, Andy.</HTML>

Reply Quote

Peter Haas

Re: Forthcoming new mineral: filatovite
January 28, 2004 06:00PM

<HTML>Andy,

It seems you did read my last post rather sloppily, so I will try it again:

1. You wrote: "As3+ has nothing to do with filatovite, which is an arsenate containing As+5." - The reference to the (As2O3)x structure was just a way to explain how a feldspar-like three-dimensional network structure exclusively containing As instead of Si could be build up. The oxidation number of As in (As2O3)x is +3. Then the sheets are linked to As atoms in neighbouring layers (formation of additional As-O-As bonds), which corresponds to an oxidation to As+5 (as I already had explained), so where's the problem ?

2. You wrote: "There is *no* problem in replacing tetrahedrally coordinated Si4+ by tetrahedrally coordinated As5+. That's how titanite, CaTi[SiO4]O, and durangite, NaAl[AsO4]F, get to be isostructural" - Yes, but (a) they are not isoelectronic (see below) and (b) neither SiO4 4-, nor AsO4 3- is a polymer, so this example has no relation to our problem.

3. You wrote: "As5+ is quite stable in tetrahedral coordination by oxygen. The sigma bonding is sp3, and pi contributions bring the bond orders (=bond valences) up to 5/4 = 1.25." - Yes, but it is involved in five bonds in this case, four sigma bonds and one p-p pi bond. This means that only three of the oxygen atoms can be engaged in bridging to a neighbouring As or Si or whatever. As you was saying before, you believe the anion was an infinite three-dimensional network, and Si was subsidiary to As in it - but with only three bonds of As +5 (in a stable form), this is practically impossible. As +5 compounds (such as their P +5 analogues) with As=O bonds may form an additional coordinative bond to Lewis acids. When we stay with our perception of a three-dimensional network, this would imply that the As atoms change from a tetrahedral configuration (sp3 hybrids) to a trigonal-bipyramidal configuration (sp2d2 hybrids) which is rather common in both As +5 and P +5 compounds. The coordinative bonds and the As=O double bonds will be arranged on a line with the remaining three ligands in planes perpendicular to it ( - As=O - As=O - As=O, etc, with "-" as the coordinative bond (more correctly, it should be pointed, but I don't have an appropriate character at my disposal) and the remaining ligands omitted). For instance, phosphoric acid esters, when dissolved in non-polar solvents, associate in that manner (as evident from the shift of the P=O stretching vibration in the IR spectrum). However, this geometry is not compatible to the feldspar network.

4. You wrote: "Arsonium compounds are those in which ..." - No, by definition, arsonium compounds are simply compounds which contain an As atom with an oxidation number of +5 and a positive charge on it (IUPAC nomenclature). Besides, hydrogen is less electropositive than arsenic (H: EN = 2.1, As: EN = 2.0). That's one reason why H4As+ does not exist.

5. Referring to arsonium compounds, you wrote: "There is negligible pi-bonding from As to its neighbours, which keeps these bonds fairly weak." - That's nonsense. The sigma-bonds are always much stronger than the additional pi-bonds, so a double bond is stronger than a single bond, but far not twice as much. You may check this by looking at the wavenumbers of the respective stretching vibrations in an IR spectrum.

6. Referring to As+5 compounds, you wrote: "These are not particularly reactive as acids." - I assume you are referring to my previous post, where I stated that arsonium compounds are strong Pearson acids. - Well, I could also say they are strong Lewis acids. The Lewis and Pearson acid-base concepts do not relate to protonation/deprotonation reactions. They have nothing to do with the Bronsted concept, nor with the classical acid/base concept. The association of [PCl4]+ and [PCl6]-, for instance, is a fine example of a Lewis acid/base reaction. These concepts are described in every inorganic chemistry textbook, so I will omit an explanation here.

7. You wrote: "It is easy to see why arsonium compounds should be very reactive. Filatovite and other arsenates(V), however, are not arsonium compounds." - First, arsenates (V) are no arsonium compounds. Second, I don't believe that filatovite is an arsonium compound. However, if it had a feldspar-like three-dimensional network structure with As atoms replacing most of the Si atoms, there would have to be As +5 atoms with four ligands and a remaining positive charge on them. All other, more likely solutions are not compatible with that structure and that's what I am in doubt about.

8. You wrote: "Polymerised As-O-Si chains are already known in the mineral tiragalloite, Mn4[AsSi3O12(OH)], which contains a linear tetramer with As at one end." - That's not what I would call polymerised As-O-Si chains. There are simply polymerised Si-O-Si chains with an As containing end group. Again, this example has no relation to our problem. It would, when the As atoms were included within the Si-O-Si chains. I think, it is evident why they aren't ...

9. You wrote: "In arsenate, the bonding capabilities of the As are fully satisfied by four very short, strong bonds to oxygen."
(a) The bonding capabilities of As are not satisified with four bonds, but with five (four sigma-bonds, one pi-bond).
(b) Actually, these bonds are not very short. They are shorter than a single bond, but not as short as a double bond.
(c) The bond strength depends on the degree of overlap of the atomic orbitals that form the bond. A more effective overlap is often obtained by mixing different atomic orbitals (hybridization), which does not always lead to a reduced bond length.
(d) The reason for the extraordinary stability of the arsenate (V) anion is delocalization (a resonance phenomenon). The valence-bond theory is not able to explain this phenomenon satisfactorily. For this, you have to refer to the MO theory.

Regards
Peter</HTML>

Reply Quote

andy christy

Re: Forthcoming new mineral: filatovite
January 30, 2004 03:36PM

<HTML>Reply to Peter Haas, 01-28-04, 18:00:

Peter: "It seems you did read my last post rather sloppily,"

No. We are just approaching this from quite different points of view, and occasioanlly use the same words in different ways.

1. Re. polymerisation of As2O3 sheets through oxygen bridges. This does not produce a very stable As2O5 framework! Consider the stoichiometry As2O5. If all As5+ is tetrahedral, then there are 2*4 = 8 bonds from As per formula unit (a "bond" is a link from As to O, irrespective of actual bond order). Therefore, the mean coordination of oxygen is only 8/5=1.6, ie 2 of the oxygens per formula unit are 1-coordinate, double-bonded to As, and only three are bridging. Simplest arrangement is for each As to form one As=O link. The remaining three As-O bonds are bridging, so the As=O unit can be regarded as three-coordinate. Three-connected frameworks are known (silicate framework in neptunite, polysilicide in ThSi2) but they are very open.

2. Exact isoelectronicity is rarely important outside the semiconductor industry. Ca2+ is isoelectronic with K+, but far more often forms compounds isostructural with Na+, which is more similar in size. This is an example of a "diagonal relationship" in the periodic table, just like Si and As!

3. Our differing uses of the word "bond" becomes apparent. By "bond", I mean a link, usually between two atoms. This is not necessarily an integral-order bond, as assumed by most molecular chemists. To me, one of the short distances between Na and Cl in NaCl is a "bond", albeit of order (= "bond valence")1/6. In the case of arsenate, yes, I know that the arsenic atom contributes 5 electrons to bonding orbitals. However, there are four bonds to oxygen, of order 5/4 = 1.25.

There is no requirement that a given As-O link has order=1. The bond sum requirements of the oxygen can be satisfied by having a strong bond to As (order 1.25) on one side, and a lower-order bond to something else on the other side. All valences can be matched, so long as there are cations with valence < 4 contributing to the framework as well as As.

There is no compulsory requirement for As form three unit-order bonds and an As=O link, or for As to adopt 5-fold coordination. Such coordination may be common in synthetic As-bearing organics, but it is unknown in arsenates, precisely because As and oxygen readily form a pi-bond, which can be delocalised over all 4 As-O links in the tetrahedron.

4. I accept that the As3- variety of arsonium does not actually occur. Note that this is simply a formal-charge description that would apply if the arsenic ligands were more electropositive than As, which has not been observed to be the case to date. Argument-by-IUPAC-authority is not really applicable to this hypothetical situation. It is as likely as not that if such compounds were to be synthesised, they would expand the definition to include them, so as to maintain uniformity between As, P and N.

5. No. A single sigma bond is indeed weaker than a single sigma bond + a partial pi bond. Hence As-O with bond order 1.25 will be stronger than As-O with bond order 1. Furthermore, the trade-off between forming two single bonds to two different ligands on the one hand, or forming a double bond to one ligand on the other, is affected by factors such as steric repulsion between ligands. A smaller number of tighter-bonded ligands may be favoured. This is certainly the case with As5+ and oxygen, as shown by the fact that arsenic acid is O=As(OH)3, not As(OH)5.

6. I am perfectly aware of the distinction between Lewis and Bronsted acids and bases. 5-coordinate P and As compounds show both Lewis acid and base behaviour (ie, are Lewis "amphoterics"), since 5 single bonds are unstable relative to 4 bonds of order 1.25 and lower inter-ligand repulsion due to low CN on the one hand, or 6 bonds of order 0.83 and reduced repulsion due to increased bond length on the other. In any case, this is not relevant to frameworks that are not 5-connected.

7. The crucial factor here is that the filatovite framework has As5+ replacing not "most of" feldspar (Al,Si) but up to a MAXIMUM OF 50% of the tetrahedral positions. Most of the others are Al3+ or Zn2+. Hence the charge imbalance problem does not arise.

8. Tiragalloite is at least an instance of arsenate-silicate linkage. Acid-base reasons for the terminal location of the arsenate/vanadate were discussed in the descriptive paper(s) for this mineral and medaite, if I recall correctly.

9.(a) Terminological difference again. In my terminology, yes, there are two-centre bonds from As to 4 oxygens, so there are four As-O bonds. The bond order happens to be 1.25, not 1.

(b)(c) this seems to be heckling about the subjective meanings of "short" and "strong".

(d) The original "bond strength" concept devised and mis-represented as an "ionic" model by Linus Pauling, as evolved over the last 30 years or so by mineralogically oriented crystal chemists like Bruce Hyde, Mike O'Keeffe, David Brown and Frank Hawthorne, *is* actually a simple, empirical expression of Molecular Orbital theory for polar-covalent compounds. Hence the occurrence of non-integral order bonds, which is an expression of the delocalisation of electron-pair bonds. It runs into trouble with metals, with necessarily multi-centre systems like boranes, and with systems where strong pair-donation and back-bonding make the distinction between bonding and antibonding orbitals ambiguous (eg carbonyls). It is noteworthy that metals are rare as minerals, and the other two classes of compound mentioned are unknown in nature. The bond valence - bond sum approach describes bonding patterns and bond length-bond valence correlations well for most minerals, since they are polar-covalent materials.

Hence...why did you think I was restricted to VB theory?</HTML>

Reply Quote

Thomas Witzke

Re: Forthcoming new mineral: filatovite
February 01, 2004 12:25PM

<HTML>There is another example of a replacement of 2 Si by Al + As in a tetrahedral framework:
Quartz, SiO2 and Alarsite, AlAsO4.
Alarsite has the quartz structure, but with AsO4 and AlO4 tetrahedra instead of SiO4 tetrahedra.
Both minerals are trigonal, space group P3_121 / P3_221, but c is doubled in Alarsite compared with Quartz.
Analogue phosphates of quartz structure type are Berlinite, AlPO4, and Rodolicoite, FePO4.
Regards
Thomas</HTML>

Reply Quote

Rob Woodside

Re: Forthcoming new mineral: filatovite
February 01, 2004 06:48PM

<HTML>Phosphate Zeolites and Arsenate Feldspars !!! Where is mineralogy going???</HTML>

Reply Quote

Jim Ferraiolo

Re: Forthcoming new mineral: filatovite
February 02, 2004 12:55PM

<HTML>Hi Rob,

Where it's always been - chemistry and crystal structures!

Regards,
Jim</HTML>

Reply Quote

andy christy

Re: Forthcoming new mineral: filatovite
February 03, 2004 02:11AM

<HTML>Re. Rob's comments about phosphate zeolites...

There are a huge number of aluminophosphate and even more exotic zeolite analogues produced synthetically now. Some of them may eventually turn up as minerals.

Zeolite structures are already known for BERYLLOphosphates.

Tiptopite is isostructural with cancrinite, and pahasapaite is a natural example of the structure of (synthetic) zeolite-rho. The structure of weinebeneite also has an IZA zeolite structure symbol.

Conversely, hurlbutite is a clathrasil analogue (ie framework without continuous channels: a "clathrabep"?) isostructural with danburite and paracelsian, if I recall correctly.

Hence, this is closer to the berlinite-alarsite-rodolicoite series, which are aluminophosphate, aluminoarsenate and ferriarsenate analogues of the clathrasil quartz.

Yet more examples where Si4+ can be replaced by (at least) one higher-valence and (at least) one lower-valence cation while preserving the structural topology!</HTML>

Reply Quote

Fade toolbar when not in focus		Fix toolbar to bottom of page
Hide Social Media Links
Slideshow frame delay	seconds

Mineral	and/or Locality
Search Google

Mineral:
and/or Locality: