Introduction

During the past couple of years I have tried to understand the ifs and buts of the amphibole group. On one hand the amphiboles is a fascinating and intriguing group of minerals, on the other hand it is a complex bunch of ugly black rocks that are impossible to distinguish and impossible to classify.

The problem with the amphibole molecule is that its general formula A B₂ C₅ T₈ O₃₂ W₂ contains 5 positions that all can accommodate different elements. In addition, the amphibole molecule is sufficiently flexible to enable different ion sizes and charges at all the different positions, or to quote Hawthorne and Oberti (2006):

"The chemical composition and variability of the amphiboles may be expressed by the general formula A B₂ C₅ T₈ O₃₂ W₂, where A = , Na, K, Ca, Pb2+; B = Li, Na, Mg, Fe2+, Mn2+, Ca; C = Li, Mg, Fe2+, Mn2+, Zn, Co, Ni, Al, Fe3+, Cr3+, Mn3+, V3+, Ti4+, Zr; T = Si, Al, Ti4+; W = (OH), F, Cl, O."

The result is an ever-changing classification scheme and two unique mineral status’, namely Hypothetical (described and approved mineral, but not yet found in nature) and Named (Found in nature but not yet approved). The most confusing is when a mineral is both Hypothetical and Named ( i.e found in nature and IMA approved, but without a proper description, type material and type locality).

I’ve spent quite some time to try to understand this group in an attempt to explain to myself the relationship between the series and the individual minerals in this group. I soon realized that there were some common nominators between the amphibole minerals in the different root name groups and even between the various subgroups. Note that the terms supergroup, group, subgroups and root name groups in this text to a large extent follows the 2012 amphibole nomenclature, which is built similar to the following hierarchy:



The two differences from the 2012 hierarchy is
1) I have ommitted subgroups in the (OH,Cl,F)dominant amphibole group that does not contain any minerals
2) I have added subgroups to the O dominant amphibole group.

Take Nybøite as an example, this is a rare sodic amphibole in the nybøite series. It has the chemical formula: NaNa₂(Mg₃Al₂)(Si₇Al)O₂₂(OH)₂ This is not very different from Tschermakite in the tschermakite series in the calcic subgroup: Ca₂(Mg₃Al₂)(Si₆Al₂)O₂₂(OH)₂. Also, both these series contains a ferri-ferro mineral, ferri-ferro-nybøite NaNa₂(Fe²⁺₃Fe³⁺₂)(Si₇Al)O₂₂(OH)₂ and ferri-ferro-tscermakite Ca₂(Fe²⁺₃Fe³⁺₂)(Si₆Al₂)O₂₂(OH)₂ respectively. These substitutions and a few others ( such as Na-K and OH-F to mention two) are recurring in many of the series. By looking at all the chemical formulas I found that almost every one of the more than 120 amphiboles could be described by only a handful element substitutions based on the chemical formula of tremolite Ca2Mg5Si8O22(OH)2.

With a few exceptions, all amphibole series can be defined by a combination of the following element substitutions:
The edenite substitution : NaAl<->Si
The tschermakite substitution MgSi <-> AlAl
The winchite substitution ( and it’s Li equivalent) CaMg <-> NaAl
The richterite substitution Ca<-> NaNa
The (Mg,Fe,Mn) substitution Ca2 <-> (Mg,Fe,Mn)2


The individual amphibole minerals can be expressed by the following element <-> element substitutions within a series:

A- Position: Na<->K
C-Position, divalent: Mg<->Fe<->Mn and Mg2<->LiAl
C-position, trivalent Al<->Fe
W-position (OH)<->F<->Cl

These very few and quite simple element substitutions can be used to express the composition of almost all of the amphiboles. Some of the rarest of these minerals will require some additional substitutions, but does not contradict the principles outlined here:


1) The Amphibole supergroup is divided into two groups the (OH,F,Cl) dominant group and the O dominant group based on the Fe²⁺(OH) <-> Fe³⁺O substitution and the Fe²⁺(OH)₂ <-> TiO₂
2) From tremolite, the amphibole root-name groups can be derived by a few standard multi-position element substitutions
3) Within a root name group, the mineral can be identified by element substitutions within the same position.

In the following, these substitutions are applied to the different amphibole sub-groups showing how the different amphibole-root name groups can be expressed by the 5 main substitutions listed above.


Calcium Amphiboles.

The calcium amphiboles are defined to have Ca2 in the B position of the amphibole molecule. The other positions can have a wide variety of elements, but all the different root name groups in the calcium amphibole subgroup (with a few exceptions) can be derived from the tremolite-actinolite series by one or both of the following element substitutions:

1) Tschermakite substitution MgSi-AlAl ( or more correctly (Mg,Fe)Si-(Al,Fe)Al)- but I'll use MgSi-AlAl for simplicity).
The Tschermakite substitution replaces one Mg in the C position and one Si in the T position with one Al in the C position and one Al in the T position.

2) Edenite substitution Si-NaAl
The Edenite substitution replaces one Si in the T position with an Al i the T position and an Na in the A position.

The relationships between the calcium amphibole root name groups are expressed as a function of element substitutions are shown in the illustration below:



Or in table format:



The actual composition of a natural calcium amphibole will contain elements of both substitutions and will be assigned to a root name group based on its dominant substitutions. It appears that the Tschermakite substitution is slightly more common than the Edenite substitution, so that calcium-amphiboles will tend to have a composition near the hornblende-pargasite-tscermakite join. Amphiboles near the edenite or tschermakite end members are rare, and possibly not even stable at normal conditions in the earth's crust. The substitutions are solid solution substitutions in many environments and one of them does not exclude the other one. In reality, the root name grous are defined arbitrary, and, as can be seen in the tables, not always by defining the end members. It can also be seen that:

1) Amphibole-series requiring many substitutions are rarer than those requiring few
2) Amphibole-series requiring uncommon substitutions are rarer than the ones that can be formed by the common ones
3) End members are rarer than intermediate members ( i.e. pargasite is more common than edenite and tschermakite)


In addition, there is an extensive element by element substitution within each root name group, defining the multiple individual minerals in each root name group. The example below shows the dominant substitutions that are forming minerals for the pargasite-root name group:



The long names like fluoro-potassic-pargasite means that this is a pargasite, but with K>Na in the A position and F2>OH2 in the W position. Similar illustrations can be made for all the root name groups in the calcium amphibole subgroup and similar element substitutions (most often the exact same) define the minerals in each amphibole root name group, whether calcium, sodium or something else. Similar schematics defining the individual minerals in each root name group can easily be prepared, but I have not included them here.

The following show two randomly selected real life analyses with data taken from Deer, Howie and Sussman- Rock forming minerals volume 2b- Double chain silicates: It is not necessary to follow these examples to follow the rest of the text.

-------

Example 1: pargasite from Duke Island, Alaska with the following formula:


(Na_0,646,K_0,116,Ca_0,279)_1,041(Ca_1,708(Mg,Fe)_0,292)₂(Mg_3,054,Fe_0,864,Mn_0,005,Al_0,757 Fe_0,455,Ti_0,157)₅(Al_1,951 Si_6,049)₈O₂₂((OH)_0,822 F,Cl)₂
Thus clearly showing what is meant with solid solution in the amphibole molecule. Here is a large number of element substitutions at play, but if we rearrange the presentation of this formula somewhat to reduce the confusing impact of the element by element substitutions:

A position: (Na,K)- 0,762
C position-divalent ions (Mg,Fe,Mn)- 3,923
C-position-trivalent ions: (Al,Fe)- 1,212
T-position Al: 1,951
T position Si: 6,049

Which gives:
Edenite substitution: 0,762(Na,K)from the A position and 0,762 Al from the T position replaces 0,762 Si from the T position
Tschermakite substitution 1,951-0,762= 1,198Al in the T position and 1,212(Al,Fe) in the C position replaces 8-6,049-0,762=1,198 Si in the T position and 5-3,952=1,077 Mg from the C position

In other words, this particular pargasite from Alaska can be described with pretty close to a 0,76 edenite and 1,2 tschermakite substitutions. This pargasite also contains a kaersutitic component that may influence the calculations above.

Example 2: Tshcermakite from Pakistan with the following analysis
(Na_0,368,K_0,063)_0,431(Ca_1,864Na_0,136)2(Mg_2,126Fe_1,489Mn_0,018,Ca_0,035,Al_1,157Fe_0,140,Ti_0,03)(Al_1,686,Si_6,314)₈O₂₂((OH)_1,947,F_0,078)_2,025
yields the following:

Edenite substitution: 0,365(Na,K) from the A position and 0,389 Al from the T position replaces 0,389 Si from the T position

Tschermakite substitution 1,297Al in the T position and 1,297(Al,Fe) in the C position replaces 1,297 Si in the T position and 1,367 Mg from the C position

In addition, this tschermakite contains a small winchite and richterite component that may explain the small deviations in the edenite and tschermakite substitutions. I did not calculate the impact of these components in detail.

-----

All in all it seems fair to be able to express the relationship between the calcium amphiboles in terms of

1: the size of their edenite and tschermakite component
2: singel element substitutions in their formulas.
3: In addition there is room for "special cases" like the kaersutite and cannilloite substitutions.

Sodium-Calcium and sodium amphiboles

The root name groups in the Sodium-Calcium and Sodium subgroup, can be expressed by similar and simple substitutions from tremolite:

The Winchite-substitution: CaMg-NaAl
i.e. a Ca-ion from the B position and a Mg-ion from the C position is substituted with a Na in the B position and an Al in the C position

The Richterite-substitution: Ca-NaNa
A Ca ion in the B position is substituted with a Na in the A position and a Na in the B position.

Many amphiboles show an intermediate composition between the calcium and sodium-calcium so that the distinction between a calcium and a sodic calcium amphibole seems to be rather arbitrary at 0,5Na in the B position and between the sodium and sodium-calcium at 1,5Na in the B position.

All root name groups in the calcium-sodium and sodium subgroups, with the exception of the Leakeite-root name group, can be expressed by these two substitutions + the tschermak substitution. This is illustrated in the figure below:



and also in table format:





The leakite root name group introduces a Li substitution in the sodium subgroup: The Mg-NaLi substitution, where a Mg ion in the C position is replaced by a Na ion in the A position and a Li ion in the C position. This is a substitution that is also present in the Lithium amphiboles.

Lithium amphiboles:

All the Lithium amphiboles are rare minerals with holmquistite as the most common ( listed from some 30 localities at mindat). These amphiboles may still be expressed by a small number of substitutions, valid both for the orthorhombic holmquistite and the monocline clinoholmquistite:

1) The Li equivalent of the winchite substitution: CaMg<->LiAl
i.e Li replaces ca in the B position and one Al replaces one Mg in the C position
2) The Mg-NaLi substitution observed in the leakite series minerals.

see figure



The illustration above also illustrates the close relationship between the Li amphiboles and the sodium/calcium amphiboles. In table format:



Again, it seems that most of the Li amphiboles have an intermediate composition between the sodium and the lithium subgroups, although it should be realized that these amphiboles is very rare.

(Mg,Fe,Mn) amphiboles

The Mg,Fe,Mn amphiboles can by be expressed by a single substitution:

1) Ca2<->(Mg,Fe,Mn)2

Where Mg,Fe,Mn replaces Ca in the B position. This substitution seems to be an "either-or" substitution where the Mg,Fe,Mn amphiboles can hold very little Ca in the B position and vice versa. These amphiboles are split into the monoclinic cummingtonite-gruenerite series and the orthorhombic anthophyllite-gedrite series.

The orthorhombic series allow edenite and tschermak substitutions just like the calcic amphiboles, but it does not seem like the cummingtonite-grunerite series allows any other substitutions than Mg-Fe-Mn substitutions at the B and C positions.

see figure



Published analytical data for Anthophyllite-gedrite show the same variations of both edenite and Tschermak substitutions as the the calcium amphiboles, even sub-silica varieties close to being a gedrite equivalent of the sadanagaite-root name group minerals.


Summary

It appears that the amphibole supergroup minerals can be expressed as variations over tremolite with a handful of substitutions defining mineral root name groups and another handful of substitutions within each root name group that defines the individual minerals.

These substitutions are described in the table below:



It does appear that at amphibole forming PT conditions that a full range of intermedient compositions (except for the Ca<->(Mg,Fe,Mn) substitution) between the different amphibole subgroups and series can be found. It does seem that these intermediate compositions can still be expressed as partial substitutions. Any given amphibole can seemingly be expressed as tremolite modified by fractions of between one and five predefined substitutions defining the amphibole series + an element by element substitution within each position defining the exact mineral.

It appears that the tschermakite, edenite, richterite and winchite substitutions are the most common ones. It seems that there is a correlation between number of substitution and rarity of the mineral.

The formation of a specific amphibole mineral requires that the local geochemical environment and that the Pressure/Temperature conditions supports the formation of this amphibole mineral. Most of the rare amphiboles are formed under slightly deviating condition in small local areas (maybe as small as a mm-scale), so that even at localities hosting very rare amphiboles the bulk rock composition support more common amphiboles, which will be more common even at these localities.

I'd like to emphasis once more that this is how I have come to understand the relationship between the various amphiboles, and that this article is nothing near a scientific view on the matter. I will appreciate comments, corrections and suggestions to this article.

---

In order to leave comments to this article, you must be registered