The standardisation of mineral group hierarchies: application to recent nomenclature proposals

There are many ways which are in current use to help with the classification of minerals, such as: Dana’s New Mineralogy (Gaines et al., 1997), the Strunz classification (Strunz & Nickel, 2001), A Systematic Classification of Minerals (Ferraiolo, 2003) and the various volumes of Deer, Howie and Zussman (Rock-forming Minerals series), which use combinations of mineral structure and chemical composition to classify minerals. There is also Fleischer’s Glossary of Minerals (Mandarino, 1999; Back & Mandarino, 2008) which lists ‘groups’ of minerals in the back section of the glossary. All are useful aids, however, there has been no systematic approach to mineral group naming or a definite hierarchical system put in place. There is also no system for the proposal and approval of mineral groups and group names.

In the past, some mineral groups have been referred to by different names and some mineral species have been proposed as members of more than one group. This proposal aims to standardise group nomenclature by introducing a hierarchy in which to classify mineral species (applied to recent nomenclature proposals) and to introduce a new procedure for the approval of new mineral groups. The following proposal has been approved by the IMA Commission on New Minerals and Mineral Names prior to publication (Voting Proposal 09–A).

From time to time, the issue of how the names of groups have been applied and its consistency has been discussed by both the CNMMN/CNMNC and the Commission on Classification of Minerals (CCM)¹. In 2004, a proposal was prepared by Drs Pushcharovsky, Pasero, Nickel and Ferraris which set out some definitions and criteria for establishing a standard set of group names. This document was circulated for comment at that time; however, there were many competing views. A revised version of this document was commented on by the CNMNC in 2008 and the definition of a mineral group in this scheme has been incorporated from that document.

Mineral species can be grouped in a number of different ways, on the basis of chemistry, crystal structure, occurrence, association, genetic history, or resource, for example, depending on the purpose to be served by the classification. However, if the classification is to adequately meet the needs of the CNMNC, it is proposed that the grouping be based on chemical composition and crystal structure, as these are the two essential components in the characterisation of a mineral species. Consequently, the simplified definition of a mineral group is:

A mineral group consists of two or more minerals with the same or essentially the same structure, and composed of chemically similar elements.

The expression “the same structure” means isotypic structures, i.e., structures belonging to one structural type. Crystal structures regarded as being ‘essentially the same’ can be encompassed by the term ‘homeotypic’. As defined by the IUCr, “two structures are considered as homeotypic if all essential features of topology are preserved between them” (Lima-de-Faria et al., 1990). In particular, homeotypic structures do not necessarily have the same space group. Therefore crystallographic variants such as superstructures, substructures and differences in the ordering of atoms that may give rise to multiple cells and/or different space groups, are considered to be homeotypic (e.g., as in the recently defined labuntsovite (Chukanov et al., 2002) and eudialyte (Johnsen et al., 2003) groups). Some polymorphs, such as triclinic and monoclinic feldspars, can be regarded as homeotypic and can therefore be included in a group; others, such as the carbon polymorphs diamond and graphite, are topologically too dissimilar (i.e., they are not homeotypic) and should not belong to the same group.

Homologous series (e.g., lillianite and pavonite series), polysomatic series (e.g., biopyriboles, heterophyllosilicates) and other structural categories that comprise modular structures (Ferraris et al., 2008) go beyond the strict definition of a homeotype, and therefore are not to be regarded as groups. However, some mineral species in these categories may belong to groups if they meet the necessary criteria.

Polytypic variations within mineral species, as defined in Guinier et al.(1984), are not regarded as comprising groups.

“Chemically similar elements” is taken to mean elements that have similar crystal-chemical behaviour. Thus, isoconfigurational minerals composed of elements with dissimilar crystal-chemical behaviour, such as galena, periclase and halite, are not to be regarded as belonging to the same group. Unoccupied structural sites are to be treated in the same way as chemical elements for the purpose of group placement.

The hierarchical scheme draws on the strengths of the various publications mentioned above. We have subdivided the scheme into six levels, however, because some of these levels are described differently in various texts, we provide the following definitions:

1. Mineral class.

2. Mineral subclass.

3. Mineral family.

4. Mineral supergroup.

5. Mineral group(s).

6. Mineral subgroup or mineral series².

1. At the highest level, mineral species can be classified primarily on the main anion (O²⁻, S²⁻ etc.), anionic complex (OH⁻, SO₄²⁻, CO₂²⁻, PO₄³⁻, B_xO_y^Z−, Si_xO_y^Z− etc.) or lack of an anion (native elements) to form classes. The most common mineral classes are: native elements, sulphides, sulphosalts, halides, oxides, hydroxides, arsenites (including antimonites, bismuthites, sulphites, selenites and tellurites), carbonates, nitrates, borates, sulphates, chromates, molybdates, tungstates, phosphates, arsenates, vanadates, silicates and organic compounds.

2. Mineral subclasses apply to the borate and silicate classes, where the configuration and bonding of tetrahedra are used to group structurally similar minerals. The subclasses are: neso-, soro-, cyclo-, ino-, phyllo-and tectosilicates(borates). Traditionally the borates are divided into monoborates, diborates, triborates, tetraborates etc. (e.g.Strunz & Nickel, 2001), however, enough structural data is known to base classification of borates on the polymerisation of the borate anion.

3. Mineral families apply to groups and/or supergroups having similar structural and/or chemical features that make them unique. A mineral family can also consist of two or more supergroups. An example of a mineral family established on the basis of structural criteria is the zeolite family, where all members are characterised by their framework structures containing cavities, but individual minerals themselves may also belong to different groups (and supergroups). The feldspathoid family also belongs to this type of ‘structural’ family. Other families are defined on the basis of chemical features, as for example the pyrite–marcasite family (which would consist of the pyrite and marcasite supergroups).

4. A mineral supergroup consists of two or more groups which have essentially the same structure and composed of chemically similar elements. Generally, a supergroup will contain members from the same mineral class (e.g. the epidote supergroup, Table 1c⁰), however in rare cases a supergroup may also contain groups belonging to different classes, as for example in the alunite supergroup (Table 1a⁰). A supergroup may also contain isolated mineral species which do not belong to any mineral group, as for example vanadinite, which is the only vanadate in the apatite supergroup.

5. A mineral group consists of two or more minerals with the same or essentially the same structure and composed of chemically similar elements (see above).

6. A mineral subgroup or mineral series should be used for minerals of a homologous series (e.g., the lillianite and pavonite series and other sulphosalt series, Moëlo et al., 2008) or polysomatic series (e.g., biopyriboles and heterophyllosilicates, Ferraris et al., 2008), where they do not meet the strict definition of a mineral group.

It is desirable that the group name be that of the first mineral to have been adequately characterised. This will generally require full structural characterisation. However, in some cases it may be preferable to name a group by a particular chemical or structural attribute (e.g., sodic-calcic amphiboles, Leake et al., 2003) rather than by a specific species name. The historical name should be used as the group name wherever possible.

The supergroup name should also be taken from the first mineral to have been adequately characterised (i.e. the first group name) or a historically significant name which no longer defines a single mineral species, such as tourmaline.

In a few cases, a group or a supergroup name can be selected contrary to the precedence rule because the name of this group (supergroup) is very firmly established in the literature. For example, it would be confusing to refer to the alunite supergroup as the “plumbogummite supergroup”, even though plumbogummite was described in 1819 and has precedence over alunite which was described in 1824.

A mineral group (or supergroup) can be introduced in the following different ways:

1. If an author (or group of authors) is submitting a new mineral proposal to the CNMNC Chairman, whereby the new mineral would either become the second mineral of a new group, or a group has not been validated previously (and has more than two members), the author(s) can submit at the same time as the new mineral proposal, a proposal for the creation of a new group.

2. During a nomenclature report by a CNMNC subcommittee handled by the CNMNC Secretary.

3. In a proposal by an author (or group of authors) to the CNMNC Vice-Chairman (responsible for changes to existing nomenclature) to create a new mineral group (or supergroup) based on data collected by those author(s).

The term ‘subgroup’ has been used in a number of different nomenclature proposals (e.g. the epidote nomenclature report, Armbruster et al., 2006), which is the equivalent of ‘level 5 mineral group’ in the hierarchical system described above. The term ‘subgroup’ is often applied where ‘group’ has been used as the equivalent of ‘level 4 mineral supergroup’. Thus, there is a simple transformation between this style of mineral group naming and the one described here.

The terms ‘family’, ‘supergroup’ and ‘group’ have been used interchangeably in several different schemes; in particular when referring to the alunite (jarosite) supergroup, which has been called the alunitejarosite family, alunite and jarosite supergroups, alunite–jarosite supergroup or the alunite supergroup. There have also been many different allocations of minerals into various groups within the ‘supergroup’ (Mills, 2007). In this case, the terms family and supergroup are equivalent to ‘level 3 mineral supergroup’ in the hierarchical scheme described above. The new scheme provides a way of simplifying large mineral groups so that confusing and inconsistent terms cannot be used to indicate the same thing.

As a test that the new hierarchical scheme can classify all minerals, we have applied it to recent nomenclature proposals (Table 1a⁰ to 1j^000000000). Here, we have created new groups and supergroups where they were previously ‘subgroups’ and ‘groups’, such as the new högbomite, labuntsovite, epidote and sapphirine supergroups.

In the case of the högbomite supergroup, the ‘subgroups’ that were established on the basis of Mg²⁺, Fe²⁺ or Zn²⁺ by Armbruster (2002) have been combined so that they fall within the högbomite, nigerite or taaffeite groups. Likewise, the alunite supergroup has been simplified (from Scott, 1987; Birch et al., 1992; Bayliss et al., 2009), so that all members fall within the alunite (SO₄-dominant), dussertite (AsO₄-dominant), plumbogummite (PO₄-dominant), or beudantite (mixed SO₄-PO₄ and SO₄-AsO₄) groups.

In the sapphirine supergroup, surinamite is an unassigned member because there is no other member to allow a group to be established. Paralabuntsovite is also an unassigned member of the labuntsovite supergroup, for the same reason.

The new hierarchical scheme has been successfully applied to recent nomenclature reports in order to standardise the way in which minerals are organised into groups. A new procedure has been put in place in order to facilitate the future proposal and naming of new mineral groups within the IMA–CNMNC framework.

The next step is to apply the hierarchical scheme to all known minerals (via a CNMNC subcommittee) and to compile a list of mineral groups, supergroups, etc., which should be published in an appropriate mineralogical journal and on the CNMNC website.