American Mineralogist, Volume 96, pages 895–913, 2011
0003-004X/11/0506–895$05.00/DOI: 10.2138/am.2011.3636 895
Nomenclature of the tourmaline-supergroup minerals
Darrell J. Henry,1,
* Milan novák (cHairMan),2
Frank c. HawtHorne,3
anDreas ertl,4
BarBara l. Dutrow,1
Pavel uHer,5
anD FeDerico Pezzotta6
1
Department of Geology and Geophysics, Louisiana Street University, Baton Rouge, Louisiana 70803, U.S.A.
2
Chairman of the Subcommittee on Tourmaline Nomenclature, Department of Geological Sciences, Petrology and Geochemistry,
Masaryk University, Kotlářská 2, CZ-611 37 Brno, Czech Republic
3
Department of Geological Sciences, University of Manitoba, Winnipeg, Manitoba R3T 2N2, Canada
4
Institut für Mineralogie und Kristallographie Geozentrum, Universität Wien, Althanstrasse 14, 1090 Wien, Austria
5
Department of Mineral Deposits, Faculty of Natural Sciences, Department of Mineralogy and Petrology, Comenius University in Bratislava,
Mlynská dolina, 842 15 Bratislava, Slovak Republic
6
Mineralogy Department, Museo di Storia Naturale di Milano, Corso Venezia 55, I-20121 Milan, Italy
aBstract
A nomenclature for tourmaline-supergroup minerals is based on chemical systematics using the
generalized tourmaline structural formula: XY3Z6(T6O18)(BO3)3V3W, where the most common ions
(or vacancy) at each site are X = Na1+
, Ca2+
, K1+
, and vacancy; Y = Fe2+
, Mg2+
, Mn2+
, Al3+
, Li1+
, Fe3+
,
and Cr3+
; Z = Al3+
, Fe3+
, Mg2+
, and Cr3+
; T = Si4+
, Al3+
, and B3+
; B = B3+
; V = OH1–
and O2–
; and W =
OH1–
, F1–
, and O2–
. Most compositional variability occurs at the X, Y, Z, W, and V sites. Tourmaline
species are defined in accordance with the dominant-valency rule such that in a relevant site the
dominant ion of the dominant valence state is used for the basis of nomenclature. Tourmaline can
be divided into several groups and subgroups. The primary groups are based on occupancy of the X
site, which yields alkali, calcic, or X-vacant groups. Because each of these groups involves cations
(or vacancy) with a different charge, coupled substitutions are required to relate the compositions of
the groups. Within each group, there are several subgroups related by heterovalent coupled substitutions.
If there is more than one tourmaline species within a subgroup, they are related by homovalent
substitutions. Additionally, the following considerations are made. (1) In tourmaline-supergroup
minerals dominated by either OH1–
or F1–
at the W site, the OH1–
-dominant species is considered the
reference root composition for that root name: e.g., dravite. (2) For a tourmaline composition that
has most of the chemical characteristics of a root composition, but is dominated by other cations or
anions at one or more sites, the mineral species is designated by the root name plus prefix modifiers,
e.g., fluor-dravite. (3) If there are multiple prefixes, they should be arranged in the order occurring in
the structural formula, e.g., “potassium-fluor-dravite.”
Keywords: Tourmaline, mineral chemistry, nomenclature, substitutions, order-disorder
introDuction
The Subcommittee on Tourmaline Nomenclature (STN) of
the International Mineralogical Association’s Commission on
New Minerals, Nomenclature and Classification (IMA-CNMNC)
has reconsidered the nomenclature of tourmaline-supergroup1
minerals. This was prompted by the general ambiguity in the
assignment of mineral names to specific tourmaline compositions.
There are several reasons for this uncertainty (Hawthorne
and Henry 1999). (1) Formal descriptions of tourmaline minerals
often specify the ideal end-member compositions, but do
not specify the limits for the use of the name. (2) Some of the
formal descriptions of tourmaline minerals specify the general
composition, but do not specify the end-member composition. (3)
Tourmaline is commonly incompletely chemically characterized,
with critical light elements (H, Li, F, and B) and the oxidation
states of transition elements (Fe, Mn) often being undetermined.
(4) Site assignments can be equivocal in the absence of crystalstructure
refinements. (5) Current graphical representations of
tourmaline compositional variations are inadequate to express
the actual substitutional nature of tourmaline. These considerations
motivated Hawthorne and Henry (1999) and the STN
to re-examine and, where necessary, redefine end-members
and potential new end-members and species, which led to the
development of several compositional diagrams that aid in classification
of the tourmaline-supergroup minerals. The proposal
for systematic classification of the tourmaline-supergroup miner-
1
Tourmaline is considered to be a supergroup in terms of nomenclature
procedures because it “consists of two or more mineral
groups, which have essentially the same structure and composed
of chemically similar elements” (Mills et al. 2009).
* E-mail: glhenr@lsu.edu
HENRY ET AL.: TOURMALINE NOMENCLATURE896
als submitted by the STN was accepted by the IMA-CNMNC
(Novák et al. 2009).
The purpose of this paper is to present the essential elements
of tourmaline nomenclature, to define concepts that are central
to tourmaline classification, and to provide practical guidelines
for application of the nomenclature.
crystallograPHic anD crystal-cHeMical
asPects oF tourMaline-suPergrouP Minerals
Tourmaline is a crystallographically acentric borosilicate
mineral with the generalized chemical formula XY3Z6(T6O18)
(BO3)3V3W (Hawthorne and Henry 1999). This general formula
makes no assumptions about site occupancy, besides those sites
known to be occupied exclusively by O2–
. Consequently, as
knowledge of site occupancy in tourmaline progresses, it will be
unnecessary to change the general formula. Only the assignment
of the cations and anions to the letters of the general formula
will be changed. The symmetry of tourmaline is predominantly
rhombohedral in the R3m space group. However, there are some
reports of tourmalines, or sectors within tourmaline crystals,
having orthorhombic, monoclinic, or triclinic symmetry (e.g.,
Akizuki et al. 2001; Shtukenberg et al. 2007; Williams et al.
2010; IMA no. 2009-46).
Table 1 gives the relative abundance of the generalized cations
(R1+
, R2+
, R3+
, R4+
) and anions (S1–
, S2–
) at each of these sites,
and presents the most common cation and anion substituents
for each of the valence states of the ions. Although tourmaline
can accommodate a great variety of cations in minor or trace
amounts, all current tourmaline species are, and most prospective
species will likely be, represented by combinations of the
cations, anions, or vacancies listed in Table 1.
In addition to the ionic size and charge of the cations and
anions, two related factors influence the content and location of
these ions in the tourmaline structure: short-range bond-valence
requirements and order-disorder reactions. (1) In the Y site, the
incident bond-valence requirements at the W site mandate that
only certain short-range configurations are stable (Hawthorne
1996). Table 2 gives the possible stable local cation configurations
at the Y site for generalized Li-free and Li-bearing tourmalines.
For a givenY-site bulk composition, the tourmaline can
have a single local Y-site cation configuration or a mixture of
possible local cation configurations. For example, the chemical
composition of end-member liddicoatite has the stable Y-site
configuration of Li2Al, whereas end-member elbaite, with a Ysite
bulk composition of Al1.5Li1.5, will have equal proportions
of Al2Li and AlLi2 clusters at the Y site (Hawthorne 1996).
These relatively few local cation configurations will constrain
the number and type of potential stable end-members possible
in tourmaline. (2) Order-disorder reactions control the actual
location of ions in the tourmaline structure. When O2–
is located
at the W site, disordering tends to develop at the Y and Z sites
(Hawthorne and Henry 1999). For example, Hawthorne (1996)
showed that in Li-free tourmaline the occurrence of Mg at the
Z site and Al at the Y site is commonly due to disorder reactions
associated with the occurrence of O2–
at the W site, and this can
be expressed as 2Y
Mg2+
+ Z
Al3+
+ W
(OH)1–
↔ 2Y
Al3+
+ Z
Mg2+
+
W
O2–
. For the 3R3+
Y-site configuration in Li-free tourmaline
the disorder relation is 3Y
Mg2+
+ 2Z
Al3+
+ W
(OH)1–
↔ 3Y
Al3+
+
2Z
Mg2+
+ W
O2–
. Effectively, the disordering substitution enhances
the amount of Mg that is located at the Z site and Al at the Y
site with the maximal amount being 2 Mg atoms per formula
unit (apfu) at the Z site (e.g., Bosi and Lucchesi 2007). Similar
arguments have been put forward for the disordering of Fe2+
to
the Z site (Bosi 2008).
In terms of a classification scheme, most of the compositional
variability occurs at the X,Y, Z, W and, to a lesser extent, V sites.
The T site is typically dominated by Si and the B site exclusively
contains B, such that the cationic occupancies at these sites do
not serve as primary parameters for classification, except in some
unusual tourmaline species. The atomic ordering in the structure
will be confidently established only with crystal-structure
refinement data and information from allied techniques such
as Mössbauer and NMR spectroscopy. The influence of W-site
O2–
appears to require disordering reactions to occur such that, in
this paper, the recognized or prospective W-site O2–
-tourmaline
species are presented in their disordered form—their most likely
actual cation distribution. However, when oxy-tourmalines are
considered for classification purposes, they are recast in their
Table 1. Relative site abundances of cations and anions in tourmaline-supergroup minerals
Site Relative abundance of ions with different valence states Common cations and anions at each site in order of relative abundance
X R1+
> R2+
> o (vacancy) R1+
: Na1+
>>K1+
R2+
: Ca2+
Y R2+
> R3+
> R1+
> R4+
R2+
: Fe2+
~ Mg2+
> Mn2+
>>> Zn2+
, Ni2+
, Co2+
, Cu2+
R3+
: Al3+
>> Fe3+
> Cr3+
>> V3+
R1+
: Li1+
R4+
: Ti4+
Z R3+
>> R2+
R3+
: Al3+
>> Fe3+
> Cr3+
> V3+
R2+
: Mg2+
> Fe2+
T R4+
>> R3+
R4+
: Si4+
R3+
: Al3+
> B3+
B R3+
R3+
: B3+
V S1–
>> S2–
S1–
: OH1–
S2:
O2–
W S1–
~ S2–
S1–
: OH1–
~ F1-
S2–
: O2–
Note: The bolded cations and anions represent the most common ions at these sites.
Table 2. Stablelocalshort-rangeY-sitecationconfigurationsforanions
of different charge at the W site
General chemical type W-site anion Y-site stable shortof
tourmaline range configurations
Li-free tourmaline (OH)1–
or F1–
3R2+
or R3+
+ 2R2+
Li-free tourmaline O2–
3R3+
or 2R3+
+ R2+
Li-bearing tourmaline (OH)1–
or F1–
2Al3+
+ Li1+
or Al3+
+ 2Li1+
Li-bearing tourmaline O2–
3Al3+
and Al3+
+ 2Li1+
Note: R is a generalized divalent cation (R2+
) or trivalent cation (R3+
).
HENRY ET AL.: TOURMALINE NOMENCLATURE 897
ordered form. Tourmaline species that contain OH1–
and F1–
at
the W site are also presented in their ordered form.
DeFinitions anD classiFication PrinciPles
There are several terms associated with mineral classification
schemes that should be elucidated. Mineral species (minerals)
are defined on the basis of their unique chemical and crystallographic
properties (Nickel and Grice 1998). In most cases
tourmaline-supergroup minerals are isostructural (space group
R3m), with the exceptions previously noted. Consequently, the
primary criterion for classification is such that most tourmaline
species are defined in terms of chemical composition, with the
dominance of a chemical constituent of the dominant valency
state at a given crystallographic site being the primary criterion
for classification.A“chemical constituent” designates a specific
cation or anion, group of atoms with the same valency state, molecular
group, or vacancies. This dominance criterion is a statement
of the dominant-constituent rule (Hatert and Burke 2008).
Tourmaline solid solutions involve both homovalent substitutions
at a given site and heterovalent coupled substitutions over single
or multiple sites. An extension of the dominant-constituent rule
is the dominant-valency rule that states that in a relevant site,
the dominant ion of the dominant valence state is considered for
nomenclature (Hatert and Burke 2008).Additional complexity is
encountered where heterovalent coupled substitutions occur on
single or multiple sites such that end-members are produced in
which two constituent ions occupy a single site: this is termed the
valency-imposed double-site occupancy (Hatert and Burke 2008).
For example, in the case of the chemical substitution that occurs
from schorl to elbaite, there is a single-site coupled substitution
that involves the incorporation of Li1+
andAl3+
in equal amounts
for Fe2+
to produce an end-member with two cations at the Y site
of elbaite, i.e.,Y
(Al1.5Li1.5). In the case of two-site coupled substitution,
dravite can be transformed to uvite by the substitution of
Ca2+
for Na1+
at the X site, coupled with the substitution of Mg2+
for Al3+
at the Z site, resulting in multiple-cation occupancy of
the Z site as Z
(MgAl5) for the uvite end-member.
An end-member is an algebraic and chemical construct2
that
is irreducible and is conformable with the crystal structure under
consideration (Hawthorne 2002; see also Appendix 1 for more
details). For example, the dravite end-member [NaMg3Al6(Si6O18)
(BO3)3(OH)3OH] is irreducible in that it cannot be expressed as
a combination of other tourmaline end-members. In the case in
which a tourmaline is determined to have dominant cation and
anion occupancies of the crystallographic sites in accordance
with this end-member, it is given the root name dravite and is,
by implication, considered to be the dravite mineral species.
In determining the systematics of tourmaline-supergroup
mineral species, the STN recommends the following general
procedure be considered:
(1) In tourmaline-supergroup minerals that are dominated by
S1–
anions (OH1–
and F1–
) in the W site, it is recommended that
the OH1–
species be the reference root composition for that root
name.As such, this OH1–
species becomes the root name without
a “hydroxyl-” prefix e.g., dravite.
(2) For a tourmaline composition that has most of the chemical
characteristics of a root composition, but is dominated by
other cations or anions at one or more sites, the mineral species is
designated by the root name plus the appropriate prefix modifiers.
For example, a tourmaline with a composition that is generally
consistent with dravite, but that contains F1–
as the most prevalent
S1–
anion where S1–
anions are dominant over S2–
anions at
the W site, the mineral species is termed fluor-dravite. To take
advantage of search capabilities, it is recommended that any
modifiers to tourmaline root names be separated by hyphens.
Hyphens are considered important to clarify the components of
the species name.
(3) If there are multiple prefix modifiers, the modifiers should
be arranged in the order in which it occurs in the structural formula
i.e., X-site modifier,Y-site modifier, Z-site modifier, T-site
modifier, and then W-site modifier. This has the advantage of
ordering the modifiers in a consistent and intuitive manner. For
example, a composition that is generally consistent with dravite,
but with K1+
being the most prevalent R1+
cation for R1+
-dominant
X-site occupancy and F1–
being the most prevalent S1–
anion for
S1–
-dominant W-site occupancy, the hypothetical mineral species
would be termed “potassium-fluor-dravite.”
(4) Consistent with the IMA-CNMNC and the International
Union of Crystallography procedures, any deviation from the
reference rhombohedral space group R3m symmetry is accommodated
in the nomenclature by adding a suffix to the root
name that indicates any atypical symmetry i.e., orthorhombic
(-O), monoclinic (-M), or triclinic (-T) (Bailey 1977). For example,
an elbaite exhibiting triclinic symmetry would be termed
“elbaite-T.”
The cationic and anionic occupancy of the X and W sites
serve as particularly convenient and petrologically meaningful
ways to define the primary tourmaline groups and a subset of
general series of tourmaline species.
PriMary tourMaline grouPs: X-site occuPancy
Tourmaline can be classified into primary groups based on
the dominant occupancy of the X site. Tourmalines have been
described that contain dominant Na1+
, Ca2+
, X
o, and, rarely,
K1+
. However, because of the relatively rare occurrence of
K-rich tourmalines, it is practical to combine the cations with
like charges, Na1+
and K1+
, into an alkali group. This results
in primary groups that are termed the alkali-, calcic-, and Xvacant-tourmaline
groups. This general grouping makes petrologic
sense because X-site occupancy generally reflects the
paragenesis of the rock in which these tourmalines crystallize,
analogous to similar general groupings in the amphibole- and
pyroxene-supergroup minerals. The alkali-, calcic-, X-vacant
ternary system for X-site occupancy can be plotted on the simple
ternary diagram illustrated in Figure 1. Arithmetically, the primary
X-site groups are defined as follows: alkali if (Na1+
+K1+
)
≥ Ca2+
and (Na1+
+K1+
) ≥ X
o; calcic if Ca2+
> (Na1+
+K1+
) and
Ca2+
> X
o; and X-vacant if X
o > (Na1+
+K1+
) and X
o > Ca2+
.
The dominance of Na1+
or K1+
in alkali-group tourmaline is an
example of the dominant-valency rule. In the uncommon case,
2
End-members are important in a thermodynamic sense because
the thermodynamic properties of end-members can be
determined, regardless of whether they exist as stable minerals.
Thermodynamic properties are essential for modeling the behavior
of solid solutions in petrological and geochemical processes.
HENRY ET AL.: TOURMALINE NOMENCLATURE898
in which tourmaline is classified as an alkali-group tourmaline
and K1+
dominates over Na1+
, it is considered a “potassium-”
tourmaline. In this case, the root name should be prefixed by
“potassium-” e.g., “potassium-povondraite.” Arithmetically, to
be considered a “potassium-”tourmaline the following conditions
must be satisfied: (Na1+
+K1+
) ≥ Ca2+
, (Na1+
+K1+
) ≥ X
o,
and K1+
> Na1+
. If there are other cations with relatively large
ion ionic radii that are found in significant amounts at the X site
(e.g., Pb2+
), they should be included with the cations of the same
charge to establish the dominance of the valency at the X site
in accordance with the dominant-valency rule. Consequently,
X-site occupancy as it relates to tourmaline nomenclature can
be extended as needed.
general series oF tourMaline sPecies:
w-site occuPancy
Three distinct anions (OH1–
, F1–
, and O2–
) can occur at the
W site, and the occupancy of the W site forms the basis for
a general series of tourmaline species: hydroxy-, fluor- and
oxy-species (Fig. 2). These are defined as follows: hydroxy if
OH1–
+ F1–
≥ O2–
and OH1–
≥ F1–
; fluor if OH1–
+ F1–
≥ O2–
and
F1–
> OH1–
; and oxy if O2–
> OH1–
+ F1–
. The appearance of the
general series of tourmaline species based on W-site occupancy
differs from the X-site grouping shown in Figure 1 because it
involves two anions with a common 1–
charge (OH1–
and F1–
) and
a single anion with a 2–
charge (O2–
). Consequently, for O2–
to be
the dominant anion of the W site, there must be >50% O2–
(cf.
Nickel 1992; Chopin 2006). Despite the difficulty of measuring
H content in tourmaline, H must be considered in the tourmaline
nomenclature because its content does vary, and it can result in
oxy-tourmaline species. In the case of an oxy-species, a coupled
substitution involving another site is required, and a new root
name is typically warranted rather than placing a modifying
prefix on the root name.
sPeciFic tourMaline sPecies
At the time of the revision of this paper, the IMA-CNMNC
has recognized 18 tourmaline species with an additional species
having a different structure (Appendix 2). Hawthorne and
Henry (1999) and the STN re-examined the compositions of the
holotype material of these species and, in some cases, redefined
the end-member formulas and mineral species in accordance
with guidelines suggested by Hawthorne and Henry (1999),
Hawthorne (2002), and this paper (Appendix 2). In addition to
the IMA-CNMNC-accepted tourmaline species, several varietal
names based on the color of tourmaline in hand sample have been
used in describing tourmaline, but these were not considered as
part of the IMA-CNMNC classification scheme (see Appendix
3). Furthermore, there are a series of names that have been
used for tourmaline that are obsolete or have been discredited
(Appendix 4).
Tourmaline can be broken into several groups and subgroups
that are useful for classification purposes. The primary
tourmaline groups are based on occupancy of the X site, which
serves as a convenient division into the alkali-tourmaline group,
calcic-tourmaline group or the X-vacant-tourmaline group (Fig.
1). Because each of these groups involves a cation or vacancy
with a different charge, coupled substitutions are required to
compositionally shift among the groups (Tables 3–6). Within
Table 3. Significant tourmaline heterovalent coupled substitutions and associated exchange vectors
Generalized coupled substitutions Corresponding exchange vector Resulting actions
(1) X
R1+
+ R2+
↔ X
o + R3+
(X
oR3+
)(R1+
R2+
)–1 relates alkali-vacant groups
(2) X
R1+
+ R3+
↔ X
Ca + R2+
(CaR2+
)(R1+
R3+
)–1 relates alkali-calcic groups
(3) 2Y
R2+
↔ Y
Li1+
+ Y
Al3+
(Li Al) (2R2+
)–1 relates incorporation of Li in all groups
(4) R2+
+ OH1–
↔ R3+
+ O2–
(R3+
O2–
)(R2+
(OH1–
))–1 relates deprotonation in all groups
(5) 0.5Li1+
+ OH1–
↔ 0.5Y
Al + O2–
(Al0.5O)(Li0.5(OH))–1 relates deprotonation in Li species
(6) R2+
+ T
Si4+
↔ R3+
+ T
R3+
(R3+
R3+
)(R2+
Si4+
)–1 relates Tschermak-like tetrahedral-octahedral substitution in all groups
Note: R represents generalized cations such thatX
R1+
= Na1+
, K1+
; R2+
= Mg2+
, Fe2+
, Mn2+
, Co2+
, Ni2+
, Zn2+
;R3+
= Al3+
, Fe3+
, Cr3+
, V3+
, B3+
(T site); and no site designation
reflects possibilities involving multiple sites.
Figure 2. Ternary system for a general series of tourmaline species
based on the anion occupancy of the W site.
Figure 1. Ternary system for the primary tourmaline groups based
on the dominant occupancy of the X site.
HENRY ET AL.: TOURMALINE NOMENCLATURE 899
Table 4. Generalized structural formula types for recognized or prospective tourmaline species listed by X-site alkali-group tourmaline
General formula (X) (Y3) (Z6) T6O18 (BO3)3 V3 W
Alkali-subgroup 1 R1+
R3
2+
R6
3+
R6
4+
O18 (BO3)3 S3
1–
S1–
Dravite* Na Mg3 Al6 Si6O18 (BO3)3 (OH)3 (OH)
Schorl* Na Fe3
2+
Al6 Si6O18 (BO3)3 (OH)3 (OH)
Chromium-dravite* Na Mg3 Cr6 Si6O18 (BO3)3 (OH)3 (OH)
Vanadium-dravite* Na Mg3 V6 Si6O18 (BO3)3 (OH)3 (OH)
Fluor-dravite* Na Mg3 Al6 Si6O18 (BO3)3 (OH)3 F
Fluor-schorl* Na Fe3
2+
Al6 Si6O18 (BO3)3 (OH)3 F
“Potassium-dravite”† K Mg3 Al6 Si6O18 (BO3)3 (OH)3 (OH)
“Tsilaisite”‡ Na Mn3
2+
Al6 Si6O18 (BO3)3 (OH)3 (OH)
Alkali-subgroup 2 R1+
R1
1.
+
5 R1.
3+
5 R6
3+
R6
4+
O18 (BO3)3 S3
1–
S1–
Elbaite* Na Li1
1.
+
5 Al3
1.
+
5 Al6 Si6O18 (BO3)3 (OH)3 (OH)
“Fluor-elbaite”† Na Li1
1.
+
5 Al3
1.
+
5 Al6 Si6O18 (BO3)3 (OH)3 F
Alkali-subgroup 3 R1+
R3
3+
R4
3+
R2
2+
§ R6
4+
O18 (BO3)3 S3
1–
S2–
Povondraite* Na Fe3
3+
Fe3
4
+
Mg2 Si6O18 (BO3)3 (OH)3 O
Chromo-alumino-povondraite* Na Cr3 Al4 Mg2 Si6O18 (BO3)3 (OH)3 O
“Oxy-dravite”†,
|| Na Al3 Al4 Mg2 Si6O18 (BO3)3 (OH)3 O
“Oxy-schorl”†,
|| Na Al3 Al4 Fe2
2
+
Si6O18 (BO3)3 (OH)3 O
“Na-Cr-O root name”† Na Cr3 Cr4 Mg2 Si6O18 (BO3)3 (OH)3 O
“Potassium-povondraite”† K Fe3
3+
Fe3
4
+
Mg2 Si6O18 (BO3)3 (OH)3 O
Alkali-subgroup 4 R1+
R1
1+
R2
3+
R6
3+
R6
4+
O18 (BO3)3 S3
1–
S2–
“Na-Li-O root name ”† Na Li1 Al2 Al6 Si6O18 (BO3)3 (OH)3 O
Alkali-subgroup 5 R1+
R3
3+
R6
3+
R6
3+
O18 (BO3)3 S3
2–
S1–
Fluor-buergerite* Na Fe3
3+
Al6 Si6O18 (BO3)3 (O)3 F
Olenite* Na Al3 Al6 Si6O18 (BO3)3 (O)3 (OH)
“Buergerite”‡ Na Fe3
3+
Al6 Si6O18 (BO3)3 (O)3 (OH)
“Fluor-olenite”‡ Na Al3 Al6 Si6O18 (BO3)3 (O)3 F
Alkali-subgroup 6 R1+
R3
3+
R6
3+
R3
3+
R3
4+
O18 (BO3)3 S3
1–
S1–
“Na-Al-Al-Al root name”‡ Na Al3 Al6 Al3Si3O18 (BO3)3 (OH)3 (OH)
“Na-Al-Al-B root name”‡ Na Al3 Al6 B3Si3O18 (BO3)3 (OH)3 (OH)
“Fluor-Na-Al-Al-Al root name”‡ Na Al3 Al6 Al3Si3O18 (BO3)3 (OH)3 F
“Fluor-Na-Al-Al-B root name”‡ Na Al3 Al6 B3Si3O18 (BO3)3 (OH)3 F
* Tourmaline species currently recognized by the IMA-CNMNC in the original or modified form (Appendix 2).
† Tourmaline species with compositions found in natural settings, but, as of the time of manuscript revision, not currently recognized by the IMA-CNMNC. Examples
of reported compositions consistent with prospective tourmaline species include:“potassium-dravite”(Ota et al. 2008),“fluor-elbaite”(Lussier et al. 2009),
“oxy-dravite”(Žáček et al. 2000),“oxy-schorl”(Novák et al. 2004),“Na-Cr-O root name”(Bosi and Lucchesi 2007) and“potassium-povondraite”(Grice et al. 1993), and
“Na-Li-O root name”(Quensel and Gabrielson 1939).
‡Tourmaline species produced experimentally (e.g., Schreyer et al. 2000; Marler et al. 2002) or found in natural settings in which the tourmalines show a tendency
for development of these compositions, and not recognized by the IMA CNMNC.
§ For the oxy-tourmaline species the formula is given as the disordered form, consistent with the Y- and Z-site occupancy demonstrated for povondraite, and
implied by the short-range disordering effects of O2–
at the W site. However, for the verification of each prospective oxy-tourmaline species the disordering on the
Z site should be demonstrated for the proper determination of the structural formula.
|| Although a new root name is appropriate, because the species names“oxy-dravite”and“oxy-schorl”have been used relatively commonly in the literature (e.g.,
Žáček al. 2000; Novák et al. 2004; Henry et al. 2008), it is recommended that these names be formally accepted as the species names for these end-members.
Table 5. Generalized structural formula types for recognized or prospective tourmaline species listed by X-site calcic-group tourmaline
General formula (X) (Y3) (Z6) T6O18 (BO3)3 V3 W
Calcic-subgroup 1 Ca2+
R3
2+
R2+
R5
3+
R6
4+
O18 (BO3)3 S3
1–
S1–
Fluor-uvite* Ca Mg3 MgAl5 Si6O18 (BO3)3 (OH)3 F
Feruvite* Ca Fe2
3
+
MgAl5 Si6O18 (BO3)3 (OH)3 (OH)
Uvite* Ca Mg3 MgAl5 Si6O18 (BO3)3 (OH)3 (OH)
“Fluor-feruvite”† Ca Fe2
3
+
MgAl5 Si6O18 (BO3)3 (OH)3 F
Calcic-subgroup 2 Ca2+
R2
1+
R1
3+
R6
3+
R6
4+
O18 (BO3)3 S3
1–
S1–
Fluor-liddicoatite* Ca Li1
2
+
Al3+
Al6 Si6O18 (BO3)3 (OH)3 F
“Liddicoatite”† Ca Li1
2
+
Al3+
Al6 Si6O18 (BO3)3 (OH)3 (OH)
Calcic-subgroup 3 Ca2+
R3
2+
R6
3+
R6
4+
O18 (BO3)3 S3
1–
S2–
“Ca-Mg-O root name”‡ Ca Mg3 Al6 Si6O18 (BO3)3 (OH)3 O
“Ca-Fe-O root name”‡ Ca Fe2
3
+
Al6 Si6O18 (BO3)3 (OH)3 O
Calcic-subgroup 4 Ca2+
R1
1.
+
5 R3
1.
+
5 R3
6
+
R4
6
+
O18 (BO3)3 S1
3
–
S2–
“Ca-Li-O root name ”‡ Ca Li1.5 Al1.5 Al6 Si6O18 (BO3)3 (OH)3 O
* Tourmaline species currently recognized by the IMA-CNMNC in the original or modified form (Appendix 2).
† Tourmaline species with compositions found in natural settings, but, as of the time of manuscript revision, not currently recognized by the IMA-CNMNC. Examples
of reported compositions consistent with prospective tourmaline species include:“fluor-feruvite”(Breaks et al. 2008) and“liddicoatite”(Breaks et al. 2008).
‡ Tourmaline species produced experimentally or found in natural settings in which the tourmalines show a tendency for development of these compositions,
and not recognized by the IMA-CNMNC.
HENRY ET AL.: TOURMALINE NOMENCLATURE900
each group there is a fundamental subgroup, subgroup 1, from
which additional subgroups can be generated. Incorporation of
Li1+
via coupled substitution (3) of Table 3 can result in additional
Li-bearing subgroups, which will warrant new root names (e.g.,
elbaite in alkali-subgroup 2 of Table 4). Likewise, the coupled
substitution (4) of Table 3 can result in deprotonated tourmaline
species (e.g., alkali-subgroups 3 and 4 of Table 4), and these
species should, by analogy, have distinct root names. Coupled
substitution (6) of Table 3 could yield species with trivalent
cations in up to 50% of the T site. Although such species have
not been reported from natural settings, experimental syntheses
(e.g., Schreyer et al. 2000; Marler et al. 2002) or compositional
tendencies in natural tourmaline warrant their inclusion as prospective
species. Within a given tourmaline subgroup there can
be several homovalent substitutions that result in extensive
or complete solid solution, but it does not change the fundamental
character of a tourmaline subgroup (e.g., Fe2+
for Mg2+
substitution that relates schorl and dravite in alkali-subgroup 1
tourmaline). Within each group, subgroup 1 generally has the
most species. For instance, alkali-subgroup 1 contains 8 possible
species, mostly variants with dravite or schorl root names. Table
7 presents several additional considerations associated with
various substitutions in the tourmaline-supergroup minerals.
Table 7. Additional considerations associated with homovalent and heterovalent substitutions in tourmaline-supergroup minerals
(a) The most common R1+
homovalent substitution in the X-site is Na1+
↔ K1+
. The tourmaline is considered to be a member of the alkali group if (Na1+
+K1+
) ≥ Ca2+
and (Na1+
+K1+
) ≥ X
o. In the uncommon case in which tourmaline is classified as an alkali-group tourmaline and K1+
> Na1+
, it is considered a“potassium-tourmaline“.
The tourmaline is considered part of the calcic group if Ca2+
> (Na1+
+K1+
) and Ca2+
> X
o. The tourmaline is considered part of the X-site vacancy group if X
o >
(Na1+
+K1+
) and X
o > Ca2+
.
(b) R2+
homovalentsubstitutions involve a number of divalent cations such as Mg2+
, Fe2+
, Mn2+
, Co2+
, Ni2+
, and Zn2+
. For example, the most common substitution is probably
Mg2+
↔ Fe2+
. In this case the complicating factor is that Mg2+
and possibly Fe2+
are generally the primary R2+
cations that can be significantly accommodated on
the Zsite at concentrations up to 2 apfu, most commonly associated with disordering related to incorporation of O2–
on theW site.Within a divalent group of elements
on a given site, the dominant R2+
cations lead to a different modifier/root names. For classification purposes the tourmaline formula should be cast in its ordered form.
(c) R3+
homovalent substitutions occur on both the Y and Z sites and include trivalent cations such as Al3+
, Fe3+
, Cr3+
, and V3+
. For example, a common substitution is
Fe3+
↔ Al3+
. The R3+
cations are most commonly found on the Z site, but the Y site and T site can contain up to 3 apfu R3+
cations. If there is O2–
on the W site, there
is likely to be a disordering resulting in the displacement of the R3+
to the Y site with a concomitant substitution of Mg2+
on the Z site. For classification purposes
the tourmaline formula should be cast in its ordered form.
(d) If there is Li in the tourmaline it is typically introduced via heterovalent substitution (3) Y
2R2+
↔ Y
Li1+
+ Y
Al3+
. Note that for each Li that is introduced there is 2 R2+
displaced. That means that plotting parameters in the ternary elbaite-schorl-dravite subsystem are 2Li-Fe2+
-Mg.
(e) The introduction of O2–
in the W or V site can take place via the heterovalent-deprotonation substitution (4) R2+
+ OH1–
↔ R3+
+ O2–
. Note that for each O2–
there
must be the introduction of an R3+
at the expense of R2+
.
(f) The introduction of R3+
(typically Al or B) into the tetrahedral site can take place via a Tschermak type of heterovalent substitution (5) R2+
+ T
Si4+
↔ R3+
+ T
R3+
.
Table 6. Generalized structural formula types for recognized or prospective tourmaline species listed by X-site vacant-group tourmaline
General formula (X) (Y3) (Z6) T6O18 (BO3)3 V3 W
Vacant-subgroup 1 o R2
2+
R3+
R6
3+
R6
4+
O18 (BO3)3 S3
1–
S1–
Foitite* o Fe2
2
+
Al Al6 Si6O18 (BO3)3 (OH)3 (OH)
Magnesio-foitite* o Mg2Al Al6 Si6O18 (BO3)3 (OH)3 (OH)
Vacant-subgroup 2 o R1
1+
R2
3+
R6
3+
R6
4+
O18 (BO3)3 S3
1–
S1–
Rossmanite* o Li1+
Al3
2
+
Al6 Si6O18 (BO3)3 (OH)3 (OH)
Vacant-subgroup 3 o R1
2+
R2
3+
R6
3+
R6
4+
O18 (BO3)3 S3
1–
S2–
“o-Mg-O root name”‡ o MgAl2 Al6 Si6O18 (BO3)3 (OH)3 O
“o-Fe-O root name”† o Fe2+
Al2 Al6 Si6O18 (BO3)3 (OH)3 O
Vacant-subgroup 4 o R0.
1+
5 R2.
3+
5 R6
3+
R6
4+
O18 (BO3)3 S3
1–
S2–
“o-Li-O root name”† o Li0.5 Al2.5 Al6 Si6O18 (BO3)3 (OH)3 O
* Tourmaline species currently recognized by the IMA-CNMNC in the original or modified form (Appendix 2).
†Tourmaline species with compositions found in natural settings, but, as of the time of manuscript revision, not currently recognized by the IMA-CNMNC. Examples
of reported compositions generally consistent with prospective tourmaline species include:“o-Fe-O root name”(Medaris et al. 2003) and“o-Li-O root name”(Ertl
et al. 2005).
‡ Tourmaline species produced experimentally or found in natural settings in which the tourmalines show a tendency for development of these compositions,
and not recognized by the IMA-CNMNC.
This approach of identifying subgroups based on operation of
general heterovalent coupled substitutions is similar to that used
for the epidote-group nomenclature (Armbruster et al. 2006).
The hypothetical tourmaline species in Tables 4–6 (in quotation
marks) are considered to be species likely to be found naturally
with many published tourmaline analyses being consistent
with these species (see footnotes of Tables 4–6). The hydroxyand
fluor-species are written in the ordered form at the Y and Z
sites. The oxy-species are written in the disordered form, with
the understanding that disorder over the Y and Z sites is likely
(e.g., Hawthorne 1996).Additional end-members that have other
cations dominant at one or more sites are likely, and these can be
added as new tourmaline species when/if they are discovered and
characterized.This proposed scheme is therefore, readily expandable.
Whenever possible or reasonable, new tourmaline species
should be named using currently recognized root names with
appropriate prefix modifiers concatenated to the existing root
name. For example, the IMA-CNMNC-accepted Mg-equivalent
of foitite is magnesio-foitite (Table 6). Tourmaline compositions
generated through heterovalent coupled substitutions of existing
root compositions will generally mandate the introduction of new
root names. All proposed mineral species must be submitted to
the IMA-CNMNC and fulfill the requirements for new mineral
HENRY ET AL.: TOURMALINE NOMENCLATURE 901
species prior to IMA-CNMNC approval as a new mineral species.
For classification, it is recommended that a tourmaline be
named for the dominant species in the dominant subgroup,
i.e., consistent with the dominant constituent of the dominantvalency
state (the last adapted from Hatert and Burke 2008).
For example, if there is a mixture of 40% elbaite, 35% schorl,
and 25% dravite, the resulting subgroups would be 100%
alkali-group tourmaline. However, within this group 60% will be
alkali-subgroup 1 tourmaline species and 40% alkali-subgroup 2
species. Alkali-subgroup 1 is the dominant subgroup and schorl
is the dominant species within this dominant subgroup such that
the tourmaline should be considered a schorl. In this case, these
relations for this ternary subsystem can be graphically illustrated
(Fig. 3). The addition of other cations beyond the three shown
graphically will complicate the application of these subsystem
diagrams, but the grouping of common valency components can
extend the generality of this approach and allow determination
of the proper tourmaline species.
For identification of the wider range of possible natural and
synthetic tourmaline species there are general procedures that
should be followed.After the primary X-site group is established,
the appropriate subgroup should be determined within each of the
primary groups. This can be done graphically or by considering
the numerical thresholds that serve to separate the subgroups.
Identification of the appropriate subgroup 1–4 within each of
the primary X-site group tourmalines can be established with a
series of diagrams that use the X-site occupancy andYZ
R2+
/(YZ
R2+
+ 2Li1+
) ratio as the primary discriminating factor, with the W-site
occupancy as a further discriminator, which will further refine
the species within the subgroup (Fig. 4). An alternative diagram
of YZ
R2+
/(YZ
R2+
+ 2Li1+
) vs. W
O2–
/(W
O2–
+ W
OH1–
+ F1–
) results
in a comparable discrimination diagram (Fig. 5). Significant
variability of T-site and V-site occupancy can result in other
relatively uncommon tourmaline species or prospective species.
For tourmalines in which the V site contains more than 50% O2–
,
tourmaline in the alkali group will fall in alkali subgroup 5 and
species such as fluor-buergerite and olenite can be recognized.
In the alkali-group tourmalines, up to 3 apfu (Al3+
+ B3+
) can
substitute for Si assumingYand Z are fully occupied by trivalent
cations and W and V sites, by monovalent cations. Thus, the
criterion for distinguishing alkali subgroup 6 is Si4+
< 4.5 apfu.
In this case, the dominant tetrahedral trivalent cation (Al3+
or
B3+
) becomes the basis for discriminating the species (Table 4).
Once the subgroup is determined, the appropriate species name
is given as the dominant species within that subgroup (Tables
4–6). Uncertainties arise when the tourmalines are incompletely
analyzed and procedures for dealing with this possibility are
addressed in the discussion below.
Figure 4. Diagrams useful for establishing the appropriate
tourmaline subgroups within the alkali, calcic, and vacant groups. (a)
Determination of subgroups 1–4 for alkali- and calcic-group tourmalines
use parametersYZ
R2+
/(YZ
R2+
+ 2Li1+
) vs. Ca2+
/(Ca2+
+ Na1+
+ K1+
) together
with the dominant valency anion(s) in the W site i.e., (OH1–
+ F1–
) vs.
O2–
. YZ
R2+
represents the total number of divalent cations in the Y and
Z site. (b) Determination of subgroups 1–4 for alkali- and X-vacantgroup
tourmalines use parameters YZ
R2+
/(YZ
R2+
+ 2Li1+
) vs. X
o/(X
o +
Na1+
+ K1+
) together with the dominant valency anion(s) in the W site
i.e., (OH1–
+ F1–
) vs. O2–
.
Figure 3.Ternary dravite-schorl-elbaite subsystem. Note that dravite
and schorl are species within alkali-subgroup 1 and elbaite is a species
within alkali-subgroup 2.
HENRY ET AL.: TOURMALINE NOMENCLATURE902
Additional adjectival modifiers
Although not required, tourmaline species can be further
modified with various adjectives that precede the species names
(Tables 4–6). In the past, the IMA-CNMNC recommended use of
the “Schaller modifiers” in which the valency of the substituent
ion is indicated by the suffix “-oan” (for the lower valency cation)
and “-ian” (for the higher valency cation). However, this scheme
has several shortcomings that have prompted the approval of
alternative chemical-element adjectival modifiers in place of the
Schaller-type modifiers (Bayliss et al. 2005). In the most general
case, it is recommended that adjectival modifier such as “-rich”
or “-bearing” be used together with the specific element(s) and,
where known, an indication of the oxidation state of the cation
and/or the site it occupies, e.g., “Fe2+
-rich,” “K-bearing,” or “T
Alrich.”
The chemical-element adjectival modifiers are not part of
the name of the tourmaline species, and, consequently, authors
are free to use chemical-element modifiers that are chemically
correct and meet the needs of the author in expressing significant
chemical information about the mineral species (Bayliss et al.
2005). However, it is recommended that the authors define the
magnitude of the compositional parameters implied by these
adjectival modifier terms. Multiple modifiers are possible, and
the order of the modifiers should be such that the modifier with
the greatest percentage of site occupancy is next to the mineralspecies
name, the modifier with the second most percentage
preceding that one, etc. For example, a hypothetical tourmaline
solid solution with 60% dravite-schorl, 40% uvite-feruvite and
2/1 ratio of Mg-Fe2+
at the Y site, has a structural formula of
(Na0.6Ca0.4)(Mg2Fe2+
)(Mg0.4Al5.6)(Si6O18)(BO3)3(OH)3(OH) and
the appropriate name could be expressed as Fe2+
-,Ca-rich dravite
i.e., reflecting 33% Fe2+
at the Y site and 40% Ca at the X site.
site allocation oF cations anD anions
Hawthorne (1996) notes that chemical analyses of tourmaline
merely establish which elements are present, but do not determine
where they are located in the structure. Structural refinement and
site assignment based on crystallographic evidence is required for
accurate site allocation. Furthermore, tourmaline site occupancies
can be modeled with appropriate optimization procedures (e.g.,
Wright et al. 2000). In the absence of site assignments directly
established by crystal structure refinements, it is possible to make
some “reasonable” assumptions concerning site assignments of
specific cations and anions (Table 1). With this basic information,
it is recommended that cations and anions in tourmaline be
allocated with the following procedure:
(1) Based on the type of analytical techniques and data generated,
the most appropriate normalization scheme is used and any
significant unanalyzed cations or anions are calculated whenever
possible (see procedures in Appendix 5).
(2) Only B3+
is allocated to the B site. With compelling
chemical, crystallographic, or spectroscopic evidence, excess B3+
(B3+
> 3.0 apfu) may be assumed to be in the T site. Compelling
evidence includes NMR spectra indicating tetrahedral B3+
, structural
refinements with tetrahedral bond length determinations
and well-constrained analytical evidence that demonstrate B3+
> 3.0 apfu (e.g., Tagg et al. 1999; Hughes et al. 2000; Schreyer
et al. 2002; Ertl et al. 2006a; Lussier et al. 2009).
(3) Na1+
, Ca2+
, and K1+
are assigned to the X site with any site
deficiency assumed to represent X-site vacancy (X
o).Additional
large cations, such as Pb2+
, are likely located in the X site and
should be assigned to that site.
(4) Si4+
is assumed to be exclusively located in the T site with
any deficiency made up by Al3+
(MacDonald and Hawthorne
1995). If there is compelling chemical, crystallographic, or
spectroscopic evidence for tetrahedral B3+
, this tetrahedral B3+
should be assigned to the T site prior to the assignment of the
tetrahedral Al3+
(e.g., Lussier et al. 2009).
(5) The relative distribution of anions in the V and W sites
is reasonably well established. F1–
is exclusively contained in
the W site and O2–
tends to be preferentially contained in this
site (Grice and Ercit 1993). Consequently, it is appropriate
that all F1–
be assigned to the W site, and then O2–
. Any excess
O2–
is assigned to the V site. To date, evidence from bond angle
distortion of the ZO6 octahedron and Y-O distances and bondvalence
sums at the V site indicate that most of the tourmaline
species (except buergerite and some olenitic tourmalines) have
∼ 3(OH1–
) at the V site (Ertl et al. 2002; Cempírek et al. 2006;
Bosi and Lucchesi 2007).
(6) The Y and Z site assignments can be more ambiguous.
The least problematic assignment is the exclusive allocation of
Li1+
, Mn2+
, Zn2+
, Ni2+
, Co2+
, Cu2+
, and Ti4+
to theYsite. However,
the smaller cation allocation between the Y and Z sites can be
more uncertain. Based on crystallographic and mineral chemical
information, it is considered that, with the presence of O2–
at the
W site, Mg2+
and possibly Fe2+
may be disordered into the Z site
and trivalent cations (especiallyAl3+
and Fe3+
) may be disordered
into the Y site (Henry and Dutrow 1990, 2001; Hawthorne et
al. 1993; Taylor et al. 1995; Ertl et al. 2003a; Bosi et al. 2004).
Consequently, actual tourmaline structures can exhibit a significant
amount of disordering.
For the purposes of classification of tourmaline species only,
the ordered form of the tourmaline is assumed for all tourmaline
species including the oxy-tourmaline. Consequently, the proceFigure
5.Alternative diagram useful for establishing the appropriate
tourmaline subgroups within the alkali-, calcic- and vacant groups. The
determination of subgroups 1–4 for alkali-, calcic- and vacant-group
tourmalines use parametersYZ
R2+
/(YZ
R2+
+ 2Li1+
) vs. W
O2–
/(W
O2–
+ OH1–
+
F1–
) together with the dominant cation in the X-site and dominant valency
anion(s) in the W site, i.e., (OH1–
+ F1–
) vs. O2–
.
HENRY ET AL.: TOURMALINE NOMENCLATURE 903
dure that is recommended for classification involves allocation
of the small cations among the Z and Y sites as follows: Initially
assign the most abundant R3+
cations to the Z site (not including
any Al3+
and B3+
assigned to the tetrahedral site). Next, the
remainder of the R3+
cations should be assigned in accordance
with their abundance. If there is an excess of R3+
cations on
the Z site, the excess R3+
cations go into the Y site. If there is a
deficiency in the Z site after assigning all of the R3+
to that site
(i.e., <6.0 cations), assign Mg2+
and then Fe2+
to the Z site up to
2 apfu (Bosi and Lucchesi 2007).
tourMaline classiFication ProceDure: a
HierarcHical aPProacH
Tourmaline investigations generally have varying levels of
information available. The tourmaline information may range
from complete analytical and crystal structural data to incomplete
chemical data. However, it is important that a hierarchical
classification procedure be used to accommodate the levels of
information that are accessible.
Level 1—complete analytical and structural data
This level considers those tourmalines in which all elements
are measured, including the oxidation states of transition elements,
and the specific cation and anion site occupancies are
established by crystal-structure refinement. Note that an ordered
structural formula is assumed for classification purposes only,
and proper site occupancies should be included in the ultimate
tourmaline structural formula, whenever possible (e.g., Ertl
et al. 2003a; Bosi and Lucchesi 2004). This level of complete
characterization of tourmaline is the optimal situation, but one
which is currently relatively uncommon.
Level 2—complete analytical data
This level implies direct knowledge of all elements (light
elements and oxidation states of transition elements), but
generally with assumed site assignments. Because an ordered
structural formula is assumed, the site allocation procedure in
the previous section can be used and is appropriate for classification
purposes.
Recommended classification procedure for tourmaline
with Level 1 and Level 2 data. With complete analytical data,
the following procedure for systematically naming tourmaline
species is suggested:
(1) Cast the structural formula in an ordered form consistent
with the site allocation procedures given above.
(2) Determine the dominant X-site cation or vacancy to
establish the primary tourmaline group (Fig. 1).
(3) Establish the dominant anion (OH1–
, F1–
, or O2–
) at the
W site (Fig. 2).
(4)Ascertain the dominant anion (OH1–
or O2–
) at the V site.
The current state of knowledge is that most tourmalines are
dominated by OH1–
at the V site. The exceptions are buergerite
and some olenitic tourmalines (Ertl et al. 2005; Cempírek et
al. 2006; Bosi and Lucchesi 2007).
(5) Determine whether Si4+
< 4.5 apfu with the remainder of
the tetrahedral site being occupied by T
Al3+
and T
B3+
.Although,
natural tourmalines with these characteristics have not been
identified, tourmalines with these characterizations have been
synthesized (Schreyer et al. 2000; Marler et al. 2002).
(6) Establish the dominant Z-site cation (Al3+
, Cr3+
, Fe3+
, or
V3+
). This can be done simply by inspection or calculation, or
can be illustrated graphically. For example, in tourmaline with
low amounts of Fe3+
at the Z site of the ordered formula, the
Al-V-Cr ternary can be used to graphically display the Z-site
dominant cation (Fig. 6). The most common Z-site dominant
cation is Al3+
, and a procedure for graphically classifying 24
possible Z
Al-dominant tourmaline species is given in Figures 7
to 9. There are only a few Cr3+
-, Fe3+
-, V3+
- end-members that
have been described at this time (e.g., chromium-dravite, povondraite,
and vanadium-dravite) so comparable diagrams were
not generated for these chemical systems. The Fe3+
-dominant
end-member povondraite is commonly in solid solution with
the “oxy-dravite” and dravite and the distinction between these
species can be evaluated by examining the dominant YZ
R3+
cation
i.e., Fe3+
orAl3+
(Henry et al. 1999, 2008; Žáček et al. 2000).
(7) Determine the Y-site cation occupancy, recalling that an
ordered form of the structural formula is used for classification
purposes only. The dominant subgroup (subgroup1–4 of
each of the three groups) can be established graphically from
Figures 4 and 5.
(8) Once the dominant subgroup is determined, the dominant
species in that subgroup defines the species name. In
many instances, this can be done simply by comparing the
structural formula with the possible species found within the
appropriate subgroup (Tables 4–6). If tourmaline compositions
fall within the appropriate subsystems, the species can
also be generally established through the use of compositional
diagrams such as Figures 7–9. A more inclusive approach is to
calculate the dominant cation or anion of the dominant valency
on given sites to directly establish subgroup and species. An
Excel spreadsheet program (TourmalineSpecies-Henry1-1) is
available3
for these calculations and species determinations.
Figure 6. Ternary diagram for theAl-V-Cr subsystem of theAl-FeV-Cr
quaternary system used for illustrating the dominant occupancy of
the Z site for tourmaline, assuming minor Fe3+
on the Z site.
3
Deposit item AM-11-036, Excel spreadsheet program. Deposit items are available
two ways: For a paper copy contact the Business Office of the Mineralogical Society of
America(seeinsidefrontcover)forpriceinformation.Online,visittheMSAwebsiteat
http://www.minsocam.org, go to the American Mineralogist Contents, find the table of
contents for the specific volume/issue wanted, and then click on the deposit link there.
HENRY ET AL.: TOURMALINE NOMENCLATURE904
lined inAppendix 5 allow estimation of some of the unmeasured
cations and anions found in tourmaline. However, there can be
considerable uncertainty associated with these procedures. As
noted earlier, an ordered structural formula is assumed and the
cations allocated accordingly.
Recommended classification procedure for tourmaline
with Level 3 data. With the appropriate analytical data, estimation
of unanalyzed elements and assumed site occupancies,
the following procedure for systematically naming tourmaline
species is suggested:
(1) Cast the structural formula in an ordered form consistent
with the site allocation procedures outlined above.
(2) Determine the dominant X-site cation or vacancy to
establish the principal tourmaline subgroup (Fig. 1). This data
are readily accessible from a good-quality electron microprobe
analysis.
(3) Establish the dominant anion (OH1–
, F1–
, or O2–
) at the
W site (Fig. 2). In the absence of direct measurement of H, the
uncertainty associated with estimating H can be large, and the
resultant W
O2–
estimate inaccurate. In contrast, F1–
can be accurately
measured with the electron microprobe if proper care is
taken in the analytical procedure. If H is undetermined (measured
or calculated), it is recommended that the criterion F1–
> 0.5 apfu,
be met for the tourmaline to be considered a fluor-species.
(4) Estimate the dominant anion (OH1–
or O2–
) at the V site.
The current state of knowledge is that most tourmalines are
greatly dominated by OH1–
at the V site. Consequently, the assumption
of V
(OH)1–
= 3 is generally correct. The exceptions
are buergerite and some olenitic tourmalines (Ertl et al. 2002;
Cempírek et al. 2006; Bosi and Lucchesi 2007).
(5) Determine whether Si4+
< 4.5 apfu with the rest of the tetrahedral
site being occupied by T
Al3+
and T
B3+
.Although, natural
tourmalines with these characteristics have not been identified,
tourmalines with these characterizations have been synthesized
(Schreyer et al. 2000; Marler et al. 2002).
(6) Establish the dominant Z-site cation (Al3+
, Cr3+
, Fe3+
, or
Figure 9. Ternary diagrams useful for plotting and classifying
vacant-group tourmaline species with Al3+
dominance at the Z site and
OH1–
dominance at the V site. The two ternary subsystems represent
dominance of OH1–
or O2–
at the W site. F1–
is likely not a dominant anion
at the W site. Determination of the species is made by plotting Y-site
cations on the appropriate ternary. TheY-site occupancy is determined by
an ordered structural formula. Significant deviation from this chemical
subsystem will result in greater uncertainty of species designation.
Figure 7. Ternary diagrams useful for plotting and classifying
alkali-group tourmaline species with Al3+
dominance at the Z site and
OH1–
dominance at the V site. The three ternary subsystems represent
dominance of OH1–
, F1–
, or O2–
at the W site, respectively. Determination
of the species is made by plottingY-site cations on the appropriate ternary.
The Y site occupancy is determined by an ordered structural formula.
Significant deviation from this chemical subsystem will result in greater
uncertainty of species designation.
Figure 8. Ternary diagrams useful for plotting and classifying calcicgroup
tourmaline species with Al3+
dominance at the Z site and OH1–
dominance at theVsite.The three ternary subsystems represent dominance
of OH1–
, F1–
, or O2–
at the W site. Determination of the species is made by
plotting Y-site cations on the appropriate ternary. The Y site occupancy is
determinedbyanorderedstructuralformula.Significantdeviationfromthis
chemicalsubsystemwillresultingreateruncertaintyofspeciesdesignation.
In addition, a table of representative tourmaline analyses with
diagnostic parameters for identification of tourmaline species
is given in Appendix 6.
Level 3—Partial tourmaline compositional data without
direct measurement of B, H, Li, and the oxidation states of
transition elements
This is the most common situation, typical of tourmalines
analyzed only by the electron microprobe. The procedures out-
HENRY ET AL.: TOURMALINE NOMENCLATURE 905
V3+
). This can be done simply by inspection or calculation, or
can be illustrated graphically. For tourmaline with low amounts
of Fe3+
at the Z site of the ordered formula, the Al-V-Cr ternary
can be used to graphically establish this Z-site subgroup (Fig. 6).
The most common Z-site dominant cation isAl3+
, and a graphical
approach for classifying 24 possible Z
Al-dominant tourmaline
species is given in Figures 7 to 9.The Fe3+
-dominant end-member
povondraite is commonly in solid solution with the “oxy-dravite”
and dravite species (Henry et al. 1999, 2008; Žáček et al. 2000).
(7) Determine the Y-site cation occupancy, recalling that an
ordered form of the structural formula is used for classification
purposes only. The dominant subgroup (subgroup 1–4 of each
of the three subgroups) can be roughly established graphically
from Figures 4 and 5.
(8) With the determination of the dominant subgroup, the
dominant species in that subgroup can be identified with the
same procedure given in the previous section.
The user must decide whether the results of the assumptions
used in calculating unmeasured elements are accurate enough to
adequately characterize the tourmaline composition. If the user
considers the calculated unanalyzed elements inadequate, it is
recommended that a more generalized name be used (Table 8,
Fig. 10). However, if F1–
> 0.5 apfu the tourmaline will necessarily
be a “fluor-tourmaline” species and the earlier classification
procedures can be followed.
acknowleDgMents
The authors thank Alfonso Pesquera and Ferdinando Bosi for very detailed,
insightful, and instructive reviews. Ed Grew, American Mineralogist Associate
Editor, provided very important input into the final version of the paper. Scores of
interested individuals helped keep this process vital and dynamic. This work was
supported in part by the Austrian Science Fund (FWF) project no. P20509-N10
to A. Ertl; Grant Agency of Czech Republic, GAČR P210/10/0743 to M. Novák;
and NSF grant EAR-9405747 to D. Henry.
reFerences citeD
Akizuki, M., Kuribayashi, T., Nagase, T., and Kitakaze, A. (2001) Triclinic liddicoatite
and elbaite in growth sectors of tourmaline from Madagascar. American
Mineralogist, 86, 364–369.
Andreozzi, G.B., Bosi, F., and Longo, M. (2008) Linking Mössbauer and structural
parameters in elbaite-schorl-dravite tourmalines. American Mineralogist, 93,
658–666.
Anthony, J.W., Bideaux, R.A., Bladh, K.W., and Nichols, M.C. (1995) Handbook of
Mineralogy. Volume II Silica, Silicates, Part 1 and 2. Mineral Data Publishing,
Tucson, Arizona.
Armbruster, T., Bonazzi, P., Akasaka, M., Bermanec, V., Chopin, C., Gieré, R.,
Heuss-Assbichler, S., Liebscher,A., Menchetti, S., Pan,Y., and Pasero, M. (2006)
Recommended nomenclature of epidote-group minerals. European Journal of
Mineralogy, 18, 551–567.
Aurisicchio, C., Demartin, F., Ottolini, L., and Pezzotta, F. (1999) Homogeneous
liddicoatite from Madagascar: a possible reference material? First EMPA, SIMS
and SREF data. European Journal of Mineralogy, 11, 237–242.
Bailey, S.W. (1977) Report of I.M.A-I.U. Cr. Joint Committee on Nomenclature.
American Mineralogist, 62, 411–415.
Barton, R. Jr. (1969) Refinement of the crystal structure of buergerite and the absolute
orientation of tourmalines. Acta Crystallographica, B25, 1524–1533.
Bastin, G.F. and Heijligers, H.J.M. (2000) Quantitative electron probe microanalysis
of boron. Journal of Solid State Chemistry, 154, 177–187.
Bayliss, P. (2000) Glossary of Obsolete Mineral Names, 235 pp. Mineralogical
Record, Tucson, Arizona.
Bayliss, P., Kaesz, H.D., and Nickel, E.H. (2005) The use of chemical-element adjectival
modifiers in mineral nomenclature. Canadian Mineralogist, 43, 1429–1433.
Bosi, F. (2008) Disordering of Fe2+
over octahedrally coordinated sites of tourmaline.
American Mineralogist, 93, 1647–1653.
Bosi, F. and Lucchesi, S. (2004) Crystal chemistry of the schorl-dravite series. European
Journal of Mineralogy, 16, 335–344.
——— (2007) Crystal chemical relationships in the tourmaline group: Structural
constraints on chemical variability. American Mineralogist, 92, 1054–1063.
Bosi, F., Lucchesi, S., and Reznitskii, L. (2004) Crystal chemistry of the dravitechromdravite
series. European Journal of Mineralogy, 16, 345–352.
Breaks, F.W., Tindle, A.G., and Selway, J.B. (2008) Electron microprobe and bulk
rock and mineral compositions from rare-element pegmatites and peraluminous,
S-type granitic rocks from the Fort Hope pegmatite field, north-central Superior
ProvinceofOntario.OntarioGeologicalSurvey,MiscellaneousReleaseData,235.
Burns, P.C., MacDonald, D.J., and Hawthorne, F.C. (1994) The crystal-chemistry of
manganese-bearing elbaite. Canadian Mineralogist, 32, 31–41.
Cámara, F., Ottolini, L., and Hawthorne, F.C. (2002) Crystal chemistry of three tourmalines
by SREF, EMPA, and SIMS. American Mineralogist, 87, 1437–1442.
Cempírek, J., Novák, M., Ertl, A., Hughes, J.M., Rossman, G.R., and Dyar, M.D.
(2006) Fe-bearing olenite with tetrahedrally coordinated Al from an abyssal
pegmatite at Kutná Hora, Czech Republic: structure, crystal chemistry, optical
spectra and XANES spectra. Canadian Mineralogist, 44, 23–30.
Chopin, C. (2006) Heterovalent substitutions in mineral nomenclature: the dominantion
rule revisited. Abstracts and Programs of the 19th
General Meeting of the
International Mineralogical Association, 300.
Chukhrov,F.V.,Ed.(1981)Minerals:AHandbook.3(3),SilicateswithMultipleChains
of Si-O Tetrahedra. Nauka, Moscow, Russia (in Russian).
Clark, C.M. (2007)Tourmaline: Structural formula calculations. Canadian Mineralogist,
45, 229–238.
Dietrich, R.V. (1985) The Tourmaline Group. Van Nostrand Reinhold Company,
New York, 300 p.
Figure 10. Example of a diagram that can be used to illustrate the
generalized tourmaline species based on Mg/(Mg + Fe) vs. X
o/(X
o +
Na1+
+ K1+
) ratios.
Table 8. Recommended names for incompletely or inadequately
determined tourmalines
General formula (X) (Y3) (Z6) V3 (likely)* W (unknown)†
Alkali-group tourmaline
Elbaitic tourmaline Na Li and Al Al (OH1–
) unknown
Schorlitic tourmaline Na Fe2+
Al (OH1–
) unknown
Dravitic tourmaline Na Mg Al (OH1–
) unknown
Buergeritic tourmaline Na Fe3+
Al (O2–
) unknown
Olenitic tourmaline Na Al Al (OH1–
) unknown
Povondraitic tourmaline Na Fe3+
Fe3+
(OH1–
) unknown
Chromium-dravitic Na Mg Cr (OH1–
) unknown
tourmaline
Vanadium-dravitic Na Mg V (OH1–
) unknown
tourmaline
Calcic-group tourmaline
Liddicoatitic tourmaline Ca Li and Al Al (OH1–
) unknown
Uvitic tourmaline Ca Mg Al (OH1–
) unknown
Feruvitic tourmaline Ca Fe2+
Al (OH1–
) unknown
Vacant-group tourmaline
Rossmanitic tourmaline o Li and Al Al (OH1–
) unknown
Foititic tourmaline o Fe2+
Al (OH1–
) unknown
Magnesio-foititic o Mg Al (OH1–
) unknown
tourmaline
* Anion presumed to occupy the V site.
† Unknown amounts of OH1–
and O2–
. However, F1–
>0.5 apfu will signify a fluorspecies,
and earlier species-designation procedures are warranted.
HENRY ET AL.: TOURMALINE NOMENCLATURE906
Donnay, G. and Barton, R. Jr. (1972) Refinement of the crystal structure of elbaite
and the mechanism of tourmaline solid solution. Tschermaks Mineralogische
und Petrographische Mitteilungen, 18, 273–286.
Donnay, G., Ingamells, C.O., and Mason, B.H. (1966) Buergerite, a new species of
tourmaline. American Mineralogist, 50, 198–199.
Dunn, P.J. (1977) Chromium in dravite. Mineralogical Magazine, 41, 408–410.
Dunn, P.J., Appleman, D., Nelen, J.A., and Norberg, J. (1977a) Uvite, a new (old)
common member of the tourmaline group and its implications to collectors.
Mineralogical Record, 8, 100–108.
Dunn, P.J., Appleman, D.E., and Nelen, J.E. (1977b) Liddicoatite, a new calcium
end-member of tourmaline group. American Mineralogist, 62, 1121–1124.
Dutrow, B.L. and Henry, D.J. (2000) Complexly zoned fibrous tourmaline, Cruzeiro
mine, Minas Gerais, Brazil: A record of evolving magmatic and hydrothermal
fluids. Canadian Mineralogist, 38, 131–143.
Dyar, M.D., Taylor, M.E., Lutz, T.M., Francis, C.A., Guidotti, C.V., and Wise, M.
(1998) Inclusive chemical characterization of tourmaline: Mössbauer study of
Fe valence and site occupancy. American Mineralogist, 83, 848–864.
Dyar, M.D., Lowe, E.W., Guidotti, C.V., and Delaney, J.S. (2002) Fe3+
and Fe2+
partitioning among silicates in metapelites: a synchrotron micro-XANES study.
American Mineralogist, 87, 514–522.
Ertl, A. (2006) Über die Etymologie und die Typlokalitäten des Minerals Schörl
(About the etymology and the type localities of schorl). Mitteilungen der Österreichischen
Mineralogischen Gesellschaft, 152, 7–16.
——— (2007) Über die Typlokalität und die Nomenklatur des Minerals Dravit
(About the type locality and the nomenclature of dravite). Mitteilungen der
Österreichischen Mineralogischen Gesellschaft, 153, 23–29.
——— (2008) About the nomenclature and the type locality of elbaite: A historical
review. Mitteilungen der Österreichischen Mineralogischen Gesellschaft,
154, 35–44.
Ertl, A., Pertlik, F., and Bernhardt, H.-J. (1997) Investigations on olenite with
excess boron from the Koralpe, Styria, Austria. Österreichische Akademie der
Wissenschaften, Mathematisch-Naturwissenschaftliche Klasse, Abteilung I,
Anzeiger, 134, 3–10.
Ertl, A., Hughes, J.M., Pertlik, F., Foit, F.F. Jr., Wright, S.E., Brandstätter, F., and
Marler, B. (2002) Polyhedron distortions in tourmaline. Canadian Mineralogist,
40, 153–162.
Ertl, A., Hughes, J.M., Brandstätter, F., Dyar, M.D., and Prasad, P.S.R. (2003a)
Disordered Mg-bearing olenite from a granitic pegmatite at Goslarn,Austria:A
chemical, structural, and infrared spectroscopic study. Canadian Mineralogist,
41, 1363–1370.
Ertl, A., Hughes, J.M., Prowatke, S., Rossman, G.R., London, D., and Fritz, E.A.
(2003b) Mn-rich tourmaline from Austria: structure, chemistry, optical spectra,
and relations to synthetic solid solutions.American Mineralogist, 88, 1369–1376.
Ertl, A., Rossman, G.R., Hughes, J.M., Prowatke, S., and Ludwig, T. (2005) Mnbearing
“oxy-rossmanite” with tetrahedrally coordinatedAl and B fromAustria:
Structure, chemistry, and infrared and optical spectroscopic study. American
Mineralogist, 90, 481–487.
Ertl, A., Hughes, J.M., Prowatke, S., Ludwig, T., Prasad, P.S.R., Brandstätter,F.,
Körner, W., Schuster, R., Pertlik, F., and Marschall, H. (2006a) Tetrahedrally
coordinated boron in tourmalines from the liddicoatite-elbaite series from
Madagascar: structure, chemistry, and infrared spectroscopc studies. American
Mineralogist, 91, 1847–1856.
Ertl, A., Kolitsch, U., Prowatke, S., Dyar, M.D., and Henry, D.J. (2006b) The Fanalogue
of schorl from Grasstein, Trentino South Tyrol, Italy: crystal structure
and chemistry. European Journal of Mineralogy, 18, 583–588.
Ferrow, E.A. (1994) Mössbauer-effect study of the crystal chemistry of tourmaline.
Hyperfine Interactions, 91, 689–695.
Ferrow, E.A.,Annersten, H., and Gunawardane, R.P. (1988) Mössbauer-effect study
on the mixed-valence state of iron in tourmaline. Mineralogical Magazine, 52,
221–228.
Fleischer, M. (1969) New mineral names. American Mineralogist, 54, 326–330.
Foit, F.F. (1989) Crystal chemistry of alkali-deficient schorl and tourmaline structural
relationships. American Mineralogist, 74, 422–431.
Fortier, S., and Donnay, G. (1975) Schorl refinement showing composition dependence
of the tourmaline structure. Canadian Mineralogist, 13, Part, 2, 173.
Gorelikova, N.V., Perfil’yev, Y.D., and Bubeshkin, A.M. (1978) Mössbauer data on
distributionofFeionsintourmaline.InternationalGeologyReview,20,982–990.
Grew, E.S., Chernosky, J.V.,Werding, G.,Abraham, K., Marquez, N., and Hinthorne,
J.R. (1990) Chemistry of kornerupine and associated minerals, a wet chemical,
ion microprobe, and X-ray study emphasizing Li-contents, Be-contents, Bcontents
and F-contents. Journal of Petrology, 31, 1025–1070.
Grice, J.D. and Ercit, T.S. (1993) Ordering of Fe and Mg in the tourmaline crystal
structure—the correct formula. Neues Jahrbuch für Mineralogie,Abhandlungen,
165, 245–266.
Grice, J.D. and Robinson, G.W. (1989) Feruvite, a new member of the tourmaline
group, and its crystal structure. Canadian Mineralogist, 27, 199–203.
Grice, J.D., Ercit,T.S., and Hawthorne, F.C. (1993) Povondraite, a redefinition of the
tourmaline ferridravite. American Mineralogist, 78, 433–436.
Harris, N.B.W., Gravestock, P., and Inger, S. (1992) Ion-microprobe determinations
of trace-element concentrations in garnets from anatectic assemblages. Chemical
Geology, 100, 41–49.
Hatert, F. and Burke, E.A.J. (2008) The IMA-CNMNC dominant-constituent rule
revisited and extended. Canadian Mineralogist, 46, 717–728.
Hawthorne, F.C. (1996) Structural mechanisms for light-element variations in
tourmaline. Canadian Mineralogist, 34, 123–132.
——— (2002) The use of end-member charge-arrangements in defining new
mineral species and heterovalent substitutions in complex minerals. Canadian
Mineralogist, 40, 699–710.
Hawthorne, F.C. and Grice, J.D. (1990) Crystal structure analysis as a chemical
analytical method—application to light-elements. Canadian Mineralogist, 28,
693–702.
Hawthorne, F.C. and Henry, D.J. (1999) Classification of the minerals of the tourmaline
group. European Journal of Mineralogy, 11, 201–215.
Hawthorne, F.C., MacDonald, D.J., and Burns, P.C. (1993) Reassignment of cation
site occupancies in tourmaline—Al-Mg disorder in the crystal-structure of
dravite. American Mineralogist, 78, 265–270.
Hawthorne, F.C., Cooper, M., Bottazzi, P., Ottolini, L., Ercit, T.S., and Grew, E.S.
(1995) Microanalysis of Minerals for Boron by SREF, SIMS and EMPA—a
comparative study. Canadian Mineralogist, 33, 389–397.
Hawthorne, F.C., Selway, J.B., Kato,A., Matsubara, S., Shimizu, M., Grice, J.D., and
Vajdak, J. (1999) Magnesiofoitite, o (Mg2Al)Al6 (Si6O18) (BO3)3 (OH)4, a new
alkali-deficient tourmaline. Canadian Mineralogist, 37, 1439–1443.
Henry,D.J. and Dutrow, B.L.(1990) Casubstitution in Li-pooraluminous tourmaline.
Canadian Mineralogist, 28, 111–124.
——— (1996) Metamorphic tourmaline and its petrologic applications. In L.M.
Anovitz and E.S. Grew, Eds., Boron: Mineralogy, Petrology and Geochemistry,
33, p. 503–557. Reviews in Mineralogy, Mineralogical Society of America,
Chantilly, Virginia.
——— (2001) Compositional zoning and element partitioning in nickeloan tourmaline
from a metamorphosed karstbauxite from Samos, Greece. American
Mineralogist, 86, 1130–1142.
Henry, D.J., Kirkland, B.L., and Kirkland, D.W. (1999) Sector-zoned tourmaline
from the cap rock of a salt dome. European Journal of Mineralogy, 11, 263–280.
Henry, D.J., Viator, D., and Dutrow, B.L. (2002) Estimation of light element concentrations
in tourmaline: How accurate can it be? Programme with Abstracts
of the 18th
International Mineralogical Association, 209.
Henry,D.J.,Sun,H.,Slack,J.,andDutrow,B.L.(2008)Tourmalineinmeta-evaporites
and highly magnesian rocks: perspectives from Namibian tourmalinites. European
Journal of Mineralogy, 20, 889–904.
Hermon, E., Sankin, D.J., Donnay, G., and Muir, W.B. (1973) The distribution of
Fe2+
and Fe3+
in iron-bearing tourmalines: a Mössbauer study. Tschermaks
Mineralogische und Petrographische Mitteilungen, 19, 124–132.
Hughes, J.M., Ertl, A., Dyar, M.D., Grew, E.S., Shearer, C.K., Yates, M.G., and
Guidotti,C.V.(2000)Tetrahedrallycoordinatedboroninatourmaline:Boron-rich
olenite from Stoffhutte, Koralpe,Austria. Canadian Mineralogist, 38, 861–868.
Hughes, J.M., Ertl, A., Dyar, M.D., Grew, E.S., Wiedenbeck, M., and Brandstätter,
F. (2004) Structural and chemical response to varying B-[4] content in zoned
Fe-bearing olenite from Koralpe,Austria.American Mineralogist, 89, 447–454.
Jarosowich, E. (1966) Chemical analyses of ten stony meteorites. Geochimica et
Cosmochimica Acta, 30, 1261–1265.
Kantipuly, C.J., Longerich, H.P., and Strong, D.F. (1988)Application of inductively
coupled argon plasma mass-spectrometry (ICAP-MS) for the determination of
uranium and thorium in tourmalines. Chemical Geology, 69, 171–176.
King, R.W., Kerrich, R.W., and Daddar, R. (1988) REE distributions in tourmaline—an
INAAtechnique involving pretreatment by B-volatilization.American
Mineralogist, 73, 424–431.
Kunitz, W. (1929) Die Mischungsreihen in der Turmalingruppe und die genetischen
BeziehungenzwischenTurmalinenundGlimmern.ChemiederErde,4,208–251.
Lussier,A.J.,Aguiar, P.M., Michaelis,V.K., Kroeker, S., and Hawthorne, F.C. (2009)
The occurrence of tetrahedrally coordinatedAl and B in tourmaline:An 11
B and
27
Al MAS NMR study. American Mineralogist, 94, 785–792.
MacDonald, D.J. and Hawthorne, F.C. (1995) The crystal chemistry of Si = Al
substitution in tourmaline. Canadian Mineralogist, 33, 849–858.
MacDonald, D.J., Hawthorne, F.C., and Grice, J.D. (1993) Foitite, o [Fe2
2+
(Al,Fe3+
)]
Al6 Si6O18 (BO3)3 (OH)4, a new alkali-deficient tourmaline—description and
crystal structure, American Mineralogist, 78, 1299–1303.
Marler, B., Borowski, M.,Wodara, U., and Schreyer,W. (2002) Synthetic tourmaline
(olenite) with excess boron replacing silicon in the tetrahedral site. II Structural
analysis. European Journal of Mineralogy, 14, 763–772.
Mathesij, J. (1562) Sarepta oder Bergpostill sampt der Jochimßthalischen kurtzen
Chroniken, 233 pp. Gedruckt zu Nürnberg, durch Johann vom Berg und Ulrich
Newber.
McGee, J.J. and Anovitz, L.M. (1996) Electron probe microanalysis of geologic
materials for boron. In L.M. Anovitz and E.S. Grew, Eds., Boron: Mineralogy,
Petrology and Geochemistry, 33, p. 771–788. Reviews in Mineralogy, Mineralogical
Society of America, Chantilly, Virginia.
Medaris, L.G., Fournelle, J.H., and Henry, D.J. (2003) Tourmaline-bearing quartz
veins in the Baraboo Quartzite,Wisconsin: Occurrence and significance of foitite
and “oxy-foitite.” Canadian Mineralogist, 41, 749–758.
Mills, S.J., Hatert, F., Nickel, E.H., and Ferraris, G. (2009) The standardisation
HENRY ET AL.: TOURMALINE NOMENCLATURE 907
of mineral group hierarchies: application to recent nomenclature proposals.
European Journal of Mineralogy, 21, 1073–1080.
Nickel, E.H. (1992) Nomenclature for mineral solid solutions. American Mineralogist,
77, 660–662.
Nickel, E.H. and Grice, J.D. (1998) The IMA commission on new minerals and
mineral names: Procedures and guidelines on mineral nomenclature. Canadian
Mineralogist, 36, 913–926.
Novák, M., and Taylor, M.C. (2000) Foitite: Formation during late stages of evolution
of complex granitic pegmatites at Dobra Voda, Czech Republic, and Pala,
California, USA. Canadian Mineralogist, 38, 1399–1408.
Novák, M., Selway, J.B., Černý, P., Hawthorne, F.C., and Ottolini, L. (1999)
Tourmaline of the elbaite-dravite series from an elbaite-subtype pegmatite at
Blizna, southern Bohemia, Czech Republic. European Journal of Mineralogy,
11, 557–568.
Novák, M., Povondra, P., and Selway, J.B. (2004) Schorl-oxy-schorl to dravite-oxydravite
tourmaline from granitic pegmatites; examples from the Moldanubicum,
Czech Republic. European Journal of Mineralogy, 16, 323–333
Novák, M., Henry, D., Hawthorne F.C., Ertl, A., Uher, P., Dutrow, B., and Pezzotta,
F. (2009) Nomenclature of the tourmaline-group minerals. Report of the
Subcommittee on Tourmaline Nomenclature to the International Mineralogical
Association’s Commission on New Minerals, Nomenclature and Classification.
Nuber, B. and Schmetzer, K. (1981) Strukturverfeinernng von Liddicoatit. Neues
Jahrbuch für Mineralogie, Monatshefte, 5, 215–219.
Ota, T., Kobayashi, K., Kunihiro, T., and Nakamura, E. (2008) Boron cycling by
subducted lithosphere; insights from diamondiferous tourmaline from the
Kokchetav ultrahigh-pressure metamorphic belt. Geochimica et Cosmochimica
Acta, 72, 3531–3541.
Peck, L.C. (1964) Systematic analysis of silicates. Bulletin of the U.S. Geological
Survey 1170.
Pezzotta, F., Hawthorne, F.C., Cooper, M.A., andTeerstra, D.K. (1996) Fibrous foitite
from San Piedro. Canadian Mineralogist, 34, 741–744.
Povondra, P. (1981)The crystal chemistry of tourmalines of the schorl-dravite series.
Acta Universitatis Carolinae, Geologica, 223–264.
Povondra, P. and Čech, F. (1976) A method for the chemical analysis of tourmaline.
Acta Universitatis Carolinae, Geologica, 209–218.
Povondra, P. and Novák, M. (1986) Tourmalines in metamorphosed carbonate
rocks from western Moravia, Czechoslovakia. Neues Jahrbuch für Mineralogie,
Monatshefte, 273–282.
Povondra, P., Lang, M., Pivec, E., and Ulrych, J. (1998) Tourmaline from the Pribyslavice
peraluminous granite, Czech Republic. Journal of the Czech Geological
Society, 43, 3–8.
Prowatke, S., Ertl, A., and Hughes, J.M. (2003) Tetrahedrally coordinated Al in
Mn-rich, Li- and Fe-bearing olenite from Eibenstein an der Thaya, Lower Austria:
A chemical and structural investigation. Neues Jahrbuch für Mineralogie,
Abhandlungen, 385–395.
Quensel, P. and Gabrielson, O. (1939) Minerals of theVarutrask pegmatite. XIV.The
tourmaline group. Geologiska Föreningens Stockholm Förhandlingar, 61, 63–90.
Reznitsky, L.Z., Sklyarov, E.V., Ushchapovskaya, Z.V., Nartova, N.V., Kashaev,
A.A., Karmanov, N.S., Kanakin, S.V., Smolin, A.S., and Nekrosova, E.A.
(2001) Vanadiumdravite, NaMg3V6[Si6O18][BO3]3(OH)4, and new mineral of
the tourmaline group. ZapiskiVsesoyuznogo Mineralogicheskogo Obshchestva,
130, 59–72 (in Russian).
Roda, E., Pesquera,A., and Velasco, F. (1995) Tourmaline in granitic pegmatites and
their country rocks, Fregeneda area, Salamanca, Spain. Canadian Mineralogist,
33, 835–848.
Romé de l’Isle, J.-B.L. (1772) Essai de Cristallographie ou Description des Figures
Géométriques, Propres à Differents Corps du Regne, 427 pp. Didot jeune, Paris.
Rumyantseva, F.V. (1983) Chromdravite, a new mineral. Zapiski Vsesoyuznogo
Mineralogicheskogo Obshchestva, 112, 222–226 (English summary inAmerican
Mineralogist, 69, 210).
Saegusa, N., Price, D.C., and Smith, G. (1979)Analysis of the Mössbauer spectra of
several iron-rich tourmalines (schorls). Journal of Physics (Paris), 40C, 456–459.
Schmetzer, K., Nuber, B., andAbraham, K. (1979) Crystal chemistry of magnesiumrich
tourmalines. Neues Jahrbuch für MineralogieAbhandlungen, 136, 93–112.
Schreyer, W., Wodara, U., Marler, B., van Aken, P.A., Seifert, F., and Robert, J.L.
(2000) Synthetic tourmaline (olenite) with excess boron replacing silicon in the
tetrahedral site: I. Synthesis conditions, chemical and spectroscopic evidence.
European Journal of Mineralogy, 12, 529–541.
Schreyer,W.,Hughes,J.M.,Bernhardt,H.J.,Kalt,A.,Prowatke,S.,andErtl,A.(2002)
Reexamination of olenite from the type locality: detection of boron in tetrahedral
coordination. European Journal of Mineralogy, 14, 935–942.
Selway, J.B., Černy, P., and Hawthorne, F.C. (1998a) Feruvite from lepidolite
pegmatites at Red Cross Lake, Manitoba. Canadian Mineralogist, 36, 433–440.
Selway, J.B., Novák, M., Hawthorne, F.C., Černý, P., Ottolini, L., and Kyser, T.K.
(1998b) Rossmanite, (LiAl2)Al6(Si6O18)(BO3)3(OH)4, a new alkali-deficient tourmaline:
Description and crystal structure.American Mineralogist, 83, 896–900.
Shearer, C.K. and Papike, J.J. (1986) Distribution of boron in the Tip Top Pegmatite,
Black Hills, South-Dakota. Geology, 14, 119–123.
Shtukenberg, A., Rozhdestvenskaya, I., Frank-Kamenetskaya, O., Bronzova, J.,
Euler, H., Kirfel,A., Bannova, I., and Zolotarev,A. (2007) Symmetry and crystal
structure of biaxial elbaite-liddicoatite tourmaline from theTransbaikalia region,
Russia. American Mineralogist, 92, 675–686.
Smith, D.G.W. and Nickel, E.H. (2007)Asystem of codification for unnamed minerals:
Report of the subcommittee for unnamed minerals of the IMA commission
on new minerals, nomenclature and classification. Canadian Mineralogist, 45,
983–990.
Sokolov, P.B., Gorskaya, M.G., Gordienko, V.V., Petrova, M.G., Kretser, Yu.l., and
Frank-Kamentskii, V.A. (1986) Olenite, Na1–xAl3Al6B3Si6O27(O,OH)4—novyy
vysokoglinozemistyy mineral iz gruppy turmalinov. Zapiski Vsesoyuznogo
Mineralogicheskogo Obshchestva, 115, 119–123 (Russian). [Translated title:
“Olenite Na1–x Al3Al6B3Si6O27(O,OH)4—a new high-alumina mineral of the
tourmaline group” inAmerican Mineralogist, 73, 441, “New Mineral Names.”]
Tagg, S.L., Cho, H., Dyar, M.D., and Grew, E.S. (1999)Tetrahedral boron in naturally
occurring tourmaline. American Mineralogist, 84, 1451–1455.
Taylor, M.C., Cooper, M.A., and Hawthorne, F.C. (1995) Local charge-compensation
in hydroxyl-deficient uvite. Canadian Mineralogist, 33, 1215–1221.
Tippe,A. and Hamilton, W.C. (1971) Neutron-diffraction study of ferric tourmaline,
buergerite. American Mineralogist, 56, 101–113.
Tschermak, G. (1884) Lehrbuch der Mineralogie. IX, 589 pp., 700Abb., 2 Farbtafeln,
Hölder, Wien.
Vernadsky, W. (1913) Über die chemische Formel der Turmaline. Zeitschrift für
Kristallographie,KristallogeometrieKristallphysik,Kristallchemie,53,273–288.
Walenta, K. and Dunn, P.J. (1979) Ferridravite, a new mineral of the tourmaline
group from Bolivia. American Mineralogist, 64, 945–948.
Wallerius, J.G. (1747) Mineralogy, and mineral kingdom, 479 pp. Printed by Lorentz
Ludwig Grefing, Stockholm. (In Swedish)
Wang, H. and Hsu, H. (1966)Anew variety of dravite and its significance in mineralogy.
Kexue Tangbao, 17, 91–96 [English summary in American Mineralogist
(1967), 52, 562–563].
Werner,A.G. (1780) Constedt’s Versuch einer Mineralogie, vol. 1, part 1. Übersetzt
und Vermehrt, Leipzig.
Williams, P.A., Hatert, F., Pasero, M., and Mills, S. (2010) New minerals, nomenclature
modifications approved in 2010. Mineralogical Magazine, 74, 375–377.
Wilson,G.C.andLong,J.V.P.(1983)ThedistributionoflithiuminsomeCornishminerals—Ion
microprobe measurements. Mineralogical Magazine, 47, 191–199.
Wright, S.E., Foley, J.A., and Hughes, J.M. (2000) Optimization of site occupancies
in minerals using quadratic programming.American Mineralogist, 85, 524–531.
Žáček, V., Frýda, J., Petrov, A., and Hyršl, J. (2000) Tourmalines of the povondraite—(oxy)dravite
series from the caps rocks of meta-evaporite inAlto Chapare,
Cochamba, Bolivia. Journal of the Czech Geological Society, 45, 3–12.
Manuscript received June 24, 2010
Manuscript accepted January 28, 2011
Manuscript handled by edward Grew
aPPenDiX 1. tourMaline enD-MeMBers: general
cHaracteristics
There are inconsistencies in the earlier definitions of several of
the tourmaline end-member formulas that prompt redefining some
of these end-members (Hawthorne and Henry 1999). In defining
end-member compositions there are several characteristics that
should be considered (Hawthorne 2002).
(1) An end-member composition must be fixed. Consequently,
formulas expressed with variable cations or anions on a given site,
such as (Al3+
, Fe3+
) on the Y site or (F−
, OH−
) at the W site, can
be factored into two or more end-member components of fixed
composition. For example, in the case of the original end-member
definition of foitite, the composition was expressed as o [Fe2
2
+
(Al, Fe3+
)] Al6 (Si6O18) (BO3)3 (OH)4 (MacDonald et al. 1993).
However, this definition is incorrect because the composition is
variable and can be factored into the two fixed compositions: o
[Fe2
2
+
Al] Al6 (Si6O18) (BO3)3 (OH)4 and o [Fe2
2
+
Fe3+
] Al6 (Si6O18)
(BO3)3 (OH)4.
(2) An end-member may have more than a single cation or
anion at a single given site if it is necessary to attain electroneutrality
in the crystal structure. Electroneutrality in crystal structures
mandate that, in some instances, two cations or anions may be
required for charge balance on a single site. One of the best examples
is the tourmaline end-member elbaite: Na [Li1.5 Al1.5] Al6
HENRY ET AL.: TOURMALINE NOMENCLATURE908
(Si6O18) (BO3)3 (OH)3 (OH). The X, Z, T, B, V, and W sites are
completely ordered (occupied by only one type of cation or anion)
such that the aggregate charge is 6−
. This requires that cations on
the Y site have a total charge of 6+
. In the case of elbaite, with
a mix of Li+
and Al3+
, the aggregate charge is only met with the
Y-site composition of [Li1.5 Al1.5].
(3) Anions are critical in defining end-members, and similar
anions can occupy more than one crystallographically distinct site
in a crystal structure. Anions such as OH−
, F−
, and O2–
can define
distinct end-members. Anions sites can have crystallographic
distinctions that mandate preferences of certain anions for specific
anion sites. For example, in the tourmaline structure theVsite can
be occupied by O2–
and OH−
, but theWsite can be occupied by F−
,
O2–
, and/or OH−
. In the case of homovalent anion substitution of
F−
for OH−
, tourmaline will range from (OH)3(OH) to (OH)3(F),
thus defining the OH (hydroxy) and F (fluor) end-members. In
the case of heterovalent anion substitution of O2–
for OH−
or F−
,
coupled substitutions must involve other cations in the crystal
structure to define an “oxy” end-member.
aPPenDiX 2: tourMaline sPecies, enD-MeMBer
ForMulas, MoDiFications, etyMology, tyPe
localities, anD rePresentative cell DiMensions
The following tourmaline species have either been previously
accepted by the IMA-CNMNC or have been modified by the STN
for internal consistency among the tourmaline species. Modifications
to the original tourmaline species descriptions are noted.
Chromium-dravite
Structural formula: Na Mg3 Cr6 (Si6O18) (BO3)3 (OH)3 OH
IMA number: IMA 82-055. Original name: chromdravite.
Modifications:Duetoitsdistinctioninthetourmalinestructure,thestructuralformula
is written to designate two OH-bearing sites (the V and W sites). This contrasts
with the earlier formula that combined these two sites into a single “OH” site
with four possible OH1–
anionic groups. The full term “chromium” is used as a
prefix for consistency among species.
Etymology: Named for its relationship to dravite and chemical composition.
Type locality: Velikaya Guba uranium occurrence, Zaonezhskiy Peninsula, Karelia,
Russia.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 16.11 Å, c = 7.27 Å, V =
1634 Å3
Selected references: Rumyantseva (1983)
Chromo-alumino-povondraite
Structural formula: Na Cr3 (Al4Mg2)(Si6O18) (BO3)3 (OH)3 O
IMA number: IMA 2009-088
Modifications: none
Etymology: Named for its chemical composition and its relationship to povondraite.
Type locality: Chromite deposits of Nausahi, Keonjhar District, Orissa, India.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 16.036 Å, c = 7.319 Å, V
= 1589.9 Å3
Selected references: Williams et al. (2010)
Dravite
Structural formula: Na Mg3 Al6 (Si6O18) (BO3)3 (OH)3 OH
IMA number: First described prior the establishment of IMA in 1958—Grandfathered
status.
Modifications:Duetoitsdistinctioninthetourmalinestructure,thestructuralformula
is written to designate two OH-bearing sites (the V and W sites). This contrasts
with the earlier formula that combined these two sites into a single “OH” site
with four possible OH1–
anionic groups.
Etymology: Named in 1884 by Tschermak for the Drava river area, the location
of a Mg- and Na-rich tourmaline. The Drava river area, which is the district
along the Drava River (in German: Drau; in Latin: Drave) is in Austria and
Slovenia (Ertl 2007).
Type locality: Unterdrauburg, Carinthia, Austria (today Dobrava pri Dravogradu,
Slovenia); Ertl (2007).
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.96 Å, c = 7.21 Å, V =
1590 Å3
Selected references: Tschermak (1884), Kunitz (1929), and Dunn (1977)
Elbaite
Structural formula: Na (Li1.5Al1.5) Al6 (Si6O18) (BO3)3 (OH)3 OH
IMA number: First described prior the establishment of IMA in 1958—Grandfathered
status.
Modifications:Duetoitsdistinctioninthetourmalinestructure,thestructuralformula
is written to designate two OH-bearing sites (the V and W sites). This contrasts
with the earlier formula that combined these two sites into a single “OH” site
with four possible OH1–
anionic groups.
Etymology: Named after the type locality, on the island of Elba, Tuscany, Italy
Type locality: San Piero in Campo, Campo nell’Elba, Elba Island, Livorno Province,
Tuscany, Italy.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.85Å, c = 7.11 Å, V =
1546.9 Å3
Selected references: Vernadsky (1913), Donnay and Barton (1972), Novák et al.
(1999), and Ertl (2008)
Feruvite
Structural formula: Ca Fe3
2+
(Al5Mg) (Si6O18) (BO3)3 (OH)3 OH
IMA number: IMA 87-057
Modifications of the end-member definition: The original formula for feruvite is
Ca Fe3 (Al,Mg)6 (Si6O18) (BO3)3 (OH)4 (Grice and Robinson 1989). This is not
considered to be an appropriate end-member and it does not specify the valence
state of Fe. The Z site was determined to be (Al4.72Fe3
0.
+
34Mg0.82Fe2
0.
+
12), which is
close to an (Al5Mg) end-member.
Etymology: Named for its relationship to uvite and chemical composition.
Crystal system: Trigonal; R3m
Type locality: Repanga Island (Cuvier Island),Waikato, North Island, New Zealand.
Representative cell dimensions (type material): a = 16.01 Å, c = 7.25 Å, V =
1609 Å3
Selected references: Grice and Robinson (1989) and Selway et al. (1998a)
Fluor-buergerite
Structural formula: Na Fe3
3
+
Al6 (Si6O18) (BO3)3 O3 F
IMA number: IMA 65-005. Original name: buergerite
Etymology: Named for Martin Julian Buerger (1903–1986), a professor of mineralogy
at the Massachusetts Institute of Technology and a pioneer of crystal
structure analysis.
Modifications: For consistency among tourmaline species, the STN advocates that
hydroxy species becomes the root name of a species. For those tourmalines in
which the W site exhibits dominance of S1–
relative to S2–
anions and dominance
F1–
relative to OH1–
, the species becomes the fluor-species equivalent.
Type locality: Mexquitic, San Luis Potosí, Mexico.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.87 Å, c = 7.19 Å, V =
1568 Å3
Selected references: Donnay et al. (1966), Barton (1969), and Tippe and Hamilton
(1971)
Fluor-dravite
Structural formula: Na Mg3 Al6 (Si6O18) (BO3)3 (OH)3 F
IMA number: 2009-089
Modifications: For consistency among tourmaline species, the STN advocates that
hydroxy species becomes the root name of a species. For those tourmalines in
which the W site exhibits dominance of S1–
relative to S2–
anions and dominance
F1–
relative to OH1–
, the species becomes the fluor-species equivalent.
Etymology: Named for its relationship to dravite and chemical composition.
Type locality: Crabtree Emerald mine, Mitchell County, North Carolina, U.S.A.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.955 Å, c = 7.153 Å, V
= 1576.9 Å3
Selected references: Williams et al. (2010)
Fluor-liddicoatite
Structural formula: Ca (Li2Al) Al6 (Si6O18) (BO3)3 (OH)3 F
IMA number: IMA 76-041. Original name: liddicoatite
Modifications of the end-member definition: Dunn et al. (1977b) gave the structural
formula for the holotype liddicoatite material as (Ca0.72Na0.27) (Li1.59Al1.13
Fe0.11Ti0.05Mn0.04Mg0.04) Al6.00 Si6O18 (BO3)3 [(OH)2.73O0.27] [F0.87(OH)0.13]. Based
on this composition and the structural analogy with elbaite, they extrapolated the
end-member composition of liddicoatite as Ca (Li1.74Al1.26) Al6 (Si6O18) (BO3)3
[(OH)2.48O0.52] (F,OH), but stipulated that this is not a “pure end-member.” Fur-
HENRY ET AL.: TOURMALINE NOMENCLATURE 909
thermore, they gave the “ideal” composition of liddicoatite as Ca (Li,Al)3 Al6
(Si6O18) (BO3)3 (O,OH)3 (OH,F). Several aspects of this formula warrant modification
of the end-member structural formula to the one given here. (1) Because
F partitions exclusively in the W site, the holotype material is an F-dominant
species. (2) Any oxygen at the W and V sites should be assigned to the W site
in the holotype material structural formula (Hawthorne 1996). (3) The divalent
charge of Ca requires an adjustment of the Li:Al ratio so the Y-site is (Li2Al).
Etymology: Named for RichardT. Liddicoat (1918–2002), gemologist and president
of the Gemological Institute of America.
Type locality: Pegmatitic area southwest of the towns of Antsirabé and Betafo,
Vakinankaratra region, Antananarivo Province, Madagascar.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.87 Å, c = 7.14 Å, V =
1557 Å3
Selected references: Dunn et al. (1977b), Nuber and Schmetzer (1981), and Aurisicchio
et al. (1999)
Fluor-schorl
Structural formula: Na Fe2
3
+
Al6 (Si6O18) (BO3)3 (OH)3 F
IMA number: 2010-067
Modifications: For consistency among tourmaline species, the STN advocates that
hydroxy species becomes the root name of a species. For those tourmalines in
which the W site exhibits dominance of S1–
relative to S2–
anions and dominance
F1–
relative to OH1–
, the species becomes the fluor-species equivalent.
Etymology: Named for its relationship to schorl and chemical composition.
Type locality: Area near Zschorlau, Erzgebirge, Saxony, Germany and area near
Grasstein, Trentino, South Tyrol, Italy.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.996 Å, c = 7.186 Å, V
= 1576.9 Å3
Selected references: Ertl et al. (2006)
Fluor-uvite
Structural formula: Ca Mg3 (Al5Mg) (Si6O18) (BO3)3 (OH)3 F
IMA number: First described prior the establishment of IMA in 1958—Grandfathered
status. Original name: uvite.
Modifications of the end-member definition: Uvite, originally defined by Kunitz
(1929), was considered to have an ideal formula of Ca Mg3 (Al5Mg) (Si6O18)
(BO3)3 (OH)4. This formula was based on the examination of tourmalines from
Uva (Sri Lanka), De Kalb (New York), and Gouverneur (New York) with OH
contents of approximately 4 apfu. However, all other analyses from these localities
contain F in the 0.5–1.0 apfu range. Dunn et al. (1977a) designated a sample
of uvite from Uva, Sri Lanka as a neotype, and this sample has an anion content
of (OH2.90F0.76O0.34). This anion content would mandate that the W site would be
dominated by F and that uvite should be an F-end-member.
Etymology: Named for the type locality area, Uva Province, Sri Lanka.
Type locality: Uva Province, Sri Lanka.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.97 Å, c = 7.21 Å, V =
1592 Å3
Selected references: Kunitz (1929), Schmetzer et al. (1979), and Dunn et al. (1977a)
Foitite
Structural formula: o (Fe2+
2 Al) Al6 (Si6O18) (BO3)3 (OH)3 OH
IMA number: IMA 92-034
Modifications of the end-member definition: The original end-member composition
of foitite was given with a variable cation occupancy at the Y site i.e., Y3
= [Fe2
2
+
(Al,Fe3+
)] (MacDonald et al. 1993). Because Al is dominant relative
to Fe3+
in the holotype material, the Y-site composition of the end-member is
considered Fe2
2
+
Al.
Etymology: Named for Franklin F. Foit Jr. (1942-), mineralogist at Washington
State University, Pullman, in recognition of his work on tourmaline-supergroup
minerals.
Type locality: Found as museum specimens designated only as from “southern
California”, U.S.A., probably from White Queen mine, Pala district, San Diego
County, California, U.S.A. (Anthony et al. 1995; Novák and Taylor 2000).
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.97 Å, c = 7.13 Å, V =
1575 Å3
Selected references: MacDonald et al. (1993) and Pezzotta et al. (1996)
Magnesio-foitite
Structural formula: o (Mg2Al) Al6 (Si6O18) (BO3)3 (OH)3 OH
IMA number: IMA 98-037
Modifications:Duetoitsdistinctioninthetourmalinestructure,thestructuralformula
is written to designate two OH-bearing sites (the V and W sites). This contrasts
with the earlier formula that combined these two sites into a single “OH” site
with four possible OH1–
anionic groups.
Etymology: Named for its relationship to foitite and chemical composition.
Type locality: Kyonosawa, Mitomi-mura, Yamanashi Prefecture, Chubu region,
Honshu Island, Japan.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.88 Å, c = 7.18 Å, V =
1568 Å3
Selected references: Hawthorne et al. (1999)
Olenite
Structural formula: Na Al3 Al6 (Si6O18) (BO3)3 (O3) OH
IMA number: IMA 85-006
Modificationsoftheend-memberdefinition:Sokolovetal.(1986)gavetheformula
for the type olenite as Na1–x Al3 Al6 B3 Si6 O27 (O,OH)4.To create an end-member
formula with the ideal occupancy of one Na at the X site, three Al at the Y site,
and where the T site is completely occupied by Si, the V and W site can only
be occupied by [O3(OH)]. To create an ordered formula, the (OH) is eventually
assigned to the W site, while O3 is assigned to the V site.
Etymology: Named for the type locality area, Olenek River basin, Russia.
Type locality: Olenek River basin, Olenii (Oleny) Range, Voron’i Tundry, Kola
Peninsula, Murmansk Oblast region, Russia.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.80 Å, c = 7.09 Å, V =
1533 Å3
Selected references: Sokolov et al. (1986), Ertl et al. (1997), Hughes et al. (2000
2004), Schreyer et al. (2002), and Cempírek et al. (2006)
Povondraite
Structural formula: Na Fe3
3
+
(Fe4
3+
Mg2) (Si6O18) (BO3)3 (OH)3 O
IMA number: Renamed/redefined by IMA 90-E
Modifications of the end-member definition: Povondraite was redefined from the
original “ferridravite” (Walenta and Dunn 1979) because the initially assumed
site assignments were incorrect and did not correspond to the Fe3+
-equivalent of
dravite (Grice et al. 1993). Hawthorne and Henry (1999) suggested that Mg is
ordered at the Z site and should be part of the end-member formula.
Etymology: Named for Pavel Povondra (1924-), mineralogist and chemist at the
Charles University, Prague, Czech Republic, for his extensive work on the
tourmaline supergroup.
Type locality: Alto Chapare district, Chapare Province, Cochabamba Department,
Bolivia.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 16.19 Å, c = 7.44 Å, V =
1689 Å3
Selected references: Walenta and Dunn (1979), Grice et al. (1993), and Žácek et
al. (2000)
Rossmanite
Structural formula: o (LiAl2) Al6 (Si6O18) (BO3)3 (OH)3 OH
IMA number: IMA 96-018
Modifications:Duetoitsdistinctioninthetourmalinestructure,thestructuralformula
is written to designate two OH-bearing sites (the V and W sites). This contrasts
with the earlier formula that combined these two sites into a single “OH” site
with four possible OH1–
anionic groups.
Etymology: Named after George R. Rossman (1945-), California Institute of
Technology, Pasadena, California, U.S.A., in recognition for his work on the
spectroscopy of the tourmaline-supergroup minerals.
Type locality: Hradisko quarry, Rožná, Morava (Moravia), Czech Republic.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.77 Å, c = 7.09 Å, V =
1527 Å3
Selected references: Selway et al. (1998b)
Schorl
Structural formula: Na Fe3
2+
Al6 (Si6O18) (BO3)3 (OH)3 OH
IMA number: First described prior the establishment of IMA in 1958—Grandfathered
status.
Modifications:Duetoitsdistinctioninthetourmalinestructure,thestructuralformula
is written to designate two OH-bearing sites (the V and W sites). This contrasts
with the earlier formula that combined these two sites into a single “OH” site
with four possible OH1–
anionic groups.
Etymology: Name probably derived from the early German mining term “Schor”
(mud) or after the name of the former village “Schorl“ (today: Zschorlau) in
Saxony,Germany.Thefirstrelativelydetaileddescriptionofschorlwiththe name
“schürl” and its occurrence (various tin mines in the Saxony Ore Mountains)
was written by Johannes Mathesius (1504–1565) in 1562 under the title “Sarepta
oder Bergpostill” (Ertl 2006).
Type locality: Saxony Ore Mountains, Germany.
Crystal system: Trigonal; R3m
HENRY ET AL.: TOURMALINE NOMENCLATURE910
Representative cell dimensions: a = 15.98 Å, c = 7.16 Å, V = 1583 Å3
.
Selected references: First described in 1562 by Mathesius (Mathesij 1562) as reported
in Wallerius (1747), Romé de l’Isle (1772), Werner (1780), Fortier and
Donnay (1975), Foit (1989), and Ertl (2006)
Uvite
Structural formula: Ca Mg3 (Al5Mg) (Si6O18) (BO3)3 (OH)3 OH
IMA number: IMA 2000-030a
Modifications of the end-member definition: Uvite, originally defined by Kunitz
(1929), was considered to have an ideal formula of Ca Mg3 (Al5Mg) (Si6O18)
(BO3)3 (OH)4. This formula was based on the examination of tourmalines from
Uva (Sri Lanka), De Kalb (New York) and Gouverneur (New York) with OH
contents of approximately 4 apfu. However, all other analyses from these
localities contain F in the 0.5–1.0 apfu range. Dunn et al. (1977a) designated a
sample of uvite from Uva, Sri Lanka as a neotype, and this sample has an anion
content of (OH2.90F0.76O0.34). This anion content would mandate that the W site
would be dominated by F and that the type uvite should most properly be termed
fluor-uvite. More recently, uvitic species with OH dominant on the W site were
described by Christine M. Clark, Frank C. Hawthorne and Joel D. Grice from
the Brumado mine, Brazil locality.
Etymology: Named for the type locality area of the currently defined fluor-uvite, Uva
Province, Sri Lanka. Uvite is redefined as the hydroxy equivalent to fluor-uvite.
Type locality: Brumado mine, Bahia, Brazil.
Type material: deposited in the Royal Ontario Museum (Toronto, Canada), specimen
number M55101.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 15.954 Å, c = 7.214 Å, V
= 1590.2 Å3
Selected references: Williams et al. (2010)
Vanadium-dravite
Structural formula: Na Mg3 V6
3+
(Si6O18) (BO3)3 (OH)3 OH
IMA number: IMA 1999-050
Modifications:Duetoitsdistinctioninthetourmalinestructure,thestructuralformula
is written to designate two OH-bearing sites (the V and W sites). This contrasts
with the earlier formula that combined these two sites into a single “OH” site
with four possible OH1–
anionic groups.
Etymology: Named for its relationship to dravite and chemical composition.
Type locality: Slyudyanka complex, Lake Baikal region, Siberia, Russia.
Crystal system: Trigonal; R3m
Representative cell dimensions (type material): a = 16.12 Å, c = 7.39 Å, V =
1663 Å3
Selected references: Reznitsky et al. (2001)
Additional Potential, but Unnamed, Tourmaline Species
At the time of the manuscript revision the following tourmalines
are described either in a proposal to IMAor via the codification
for unnamed minerals (Smith and Nickel 2007) and further
expanded in the IMA-CNMNC website (http://pubsites.uws.edu.
au/ima-cnmnc/).
IMA No. 2009-046
Proposed structural formula: (Na,o)(Fe2+
, Mg)3Al6(BO3)3Si6O18(OH)4
Localities: Cleveland tin mine, Luina, Waratah, Tasmania, Australia (41°28′57′′S,
145°23′7′′E; type locality); Mount Bendoc, Victoria, Australia (37°7′60′′S,
148°54′0′′E); Mount Bischoff, Tasmania, Australia (41°25′S, 145°31′E); Blue
Mountain Saddle (Bald Hornet Claim), North Bend, King County, Washington,
U.S.A. (47°31′N, 121°43′W).
Crystal system: Monoclinic: Cm; structure determined.
Representative cell dimensions (type material): a = 10.408(3), b = 15.991(5), c
= 7.189(2) Å, β 117.44(2)°, V = 1061.88 Å3
UM2000-//-SiO:AlBFLiNa (invalid list)
Proposed structural formula: Na(Al,Li)3 Al6 (BO3)3 Si6 O18 (F,OH)4
Comment: This is the fluorine-dominant analog of elbaite (e.g., Lussier et al.
2009). The suggested name “fluor-elbaite” has not yet been approved by the
IMA-CNMNC.
UM2000-64-SiO:BFeHKMg
Proposed structural formula: (K,Na) (Fe3+
)3 (Fe3+
,Mg)6 (BO3)3 Si6 O18 (OH,O)4
Comment: This is a K-dominant tourmaline (Žáček et al. 2000).
aPPenDiX 3. color-BaseD tourMaline
varietal naMes
Although mineral varietal names are not under the jurisdiction
of the IMA-CNMNC, the STN acknowledges the use of the colorbased
tourmaline varietal names. Appendix Table 1 is a listing
of some of the varietal names that have been used, with possible
tourmaline species associations.
aPPenDiX 4. oBsolete or DiscreDiteD
tourMaline naMes
Throughout the history of tourmaline investigations several
names have been used for tourmaline or minerals that were considered
to be tourmaline. A list of former tourmaline names that
are considered obsolete or have been discredited are presented in
Appendix Table 2.
aPPenDiX 5. analytical consiDerations anD
norMalization ProceDures
The crystal chemistry of tourmaline can be evaluated utilizing
various analytical procedures, each with strengths and weaknesses
that should be carefully considered when assessing the results and
classifying a given tourmaline. Several of the techniques used
specifically for tourmaline analysis are briefly mentioned below.
Special attention is given to normalization procedures used for
the electron microprobe data of tourmaline.
Bulk tourmaline analysis techniques
Several “bulk” sample techniques have been used to analyze
tourmaline. The general disadvantage of these techniques is the
averaging effect of “bulk” samples in which chemical zonation
may be masked, and mineral and fluid inclusions may introduce
contamination. (1) Wet chemical analysis was the standard procedure
for analysis of materials prior to the widespread use of the
electron microprobe (e.g., Peck 1964; Jarosowich 1966; Povondra
and Čech 1976). The chief advantages were that light elements
could be analyzed and oxidation states of transition elements were
determined. (2) Inductively coupled plasma-atomic emission
spectroscopy (ICP-AES) has commonly been used to supplement
electronmicroprobeanalysisbydetermininglightelementconcentrations
as well as trace element levels in tourmaline (e.g., Shearer
and Papike 1986; Kantipuly et al. 1988). Sample amounts as small
as 200–300 mg have been analyzed. (3) Induced neutron activation
analysis (INAA) has been successfully used to determine REE in
tourmaline (e.g., King et al. 1988; Roda et al. 1995).
Tourmaline structural refinement and spectroscopic
analysis techniques
Single crystal refinement and spectroscopic techniques provide
valuable information about the local environments of specific
cations, trace elements and oxidation states of transition elements.
(1) Crystal structure refinement is an electron-counting technique
that, in combination with chemical analysis, helps determine the
site assignments through definition of stereochemicalrelationships
and refined site-scattering values (e.g., Hawthorne and Grice
1990; Hawthorne et al. 1995; Cámara et al. 2002). This technique
exhibits spatial resolution that will work on well-constrained
analytical problems such as those encountered in elbaite (Burns
HENRY ET AL.: TOURMALINE NOMENCLATURE 911
Fe3+
/Fe2+
ratios as well as the locations of Fe2+
and Fe3+
in the
tourmaline structure (e.g., Hermon et al. 1973; Gorelikova et al.
1978; Saegusa et al. 1979; Ferrow et al. 1988; Ferrow 1994; Dyar
et al. 1998;Andreozzi et al. 2008), but this approach is also a bulk
technique. (4) Proton-induced γ-ray emission has also been used
for light-element analysis in tourmaline (e.g., Dyar et al. 1998).
Tourmaline electron microprobe analytical techniques
The electron microprobe is currently the most commonly used
analytical tool for tourmaline analysis. However, there are several
analytical limitations that must be considered when calculating a
structural formula from electron microprobe data. Although it is
possible to analyze most of the major elements in tourmaline, Li
and H as well as the valence states of transition elements, cannot be
directly measured. In addition, a critical constituent of tourmaline
is the light element B. Boron has been analyzed with the electron
microprobe (e.g., Bastin and Heijligers 2000; McGee andAnovitz
1996; Hawthorne et al. 1995), but the analytical accuracy and
precision remain unsatisfactory for confident structural formula
calculations. These shortcomings require that the normalization
procedure for tourmaline be carefully considered to minimize
the limitations of electron microprobe analysis. Because of a
greatly improved understanding of tourmaline’s crystal structure
and chemistry as well as the constraints imposed by geochemical
settings of tourmaline, estimation techniques for unanalyzed light
elements and oxidation states are possible.
Normalization procedures, and light element and ferric Fe
estimation
A variety of tourmaline normalization procedures have been
used. Each procedure has strengths and weaknesses. However, the
most appropriate procedure depends on the completeness of the
tourmaline analysis, the quality of the analysis, the assumptions
for unanalyzed elements and the petrologic environment in which
the tourmalines develop.
Normalization of tourmaline with complete chemical
characterization.
Ultimately,thebestapproachfor determinationofthestructural
formulaoftourmalineistoanalyzetourmalineascompletelyaspossible
and by using as many techniques as possible.With a complete
analysis, normalization can be properly done on a 31 anion basis
(Henry and Dutrow 1996; Clark 2007). The most accurate site assignments
are made by means of crystal structure refinements with
acombinationofsite-scatteringandmeanbond-lengthinformation
andstructuralformulaoptimizationprocedure(e.g.,Hawthorneand
Grice 1990; Hawthorne et al. 1993; Taylor et al. 1995; Hawthorne
1996; Ertl et al. 2003a, 2003b; Bosi and Lucchesi 2004).
Normalization of electron microprobe data and estimation
of light elements and oxidation states of elements.
Most widespread tourmaline normalization procedures make
assumptions concerning the anionic and cationic assignments. In
the absence of accurate analyses of B, O, H, Li and the oxidation
states of transition elements, several procedures can be used to
calculate a structural formula. Each of these procedures, in turn,
can provide the basis for estimation of light elements and oxidation
states of transition elements.
appendix Table 1. Color-based tourmaline varietal names
Varietal name Description
Achroite colorless tourmaline, probably elbaite or rossmanite
Aphrizite dark gray schorl
Brazilian chrysolite yellow-green tourmaline resembling chrysolite
Brazilian emerald green and transparent tourmaline
resembling emerald
Brazilian peridot honey-brown to green tourmaline
resembling peridot
Brazilian ruby transparent red tourmaline resembling ruby
Brazilian sapphire transparent blue tourmaline resembling sapphire
Canary tourmaline bright yellow tourmaline
Cat’s eye tourmaline chatoyant tourmaline in a variety of colors
Ceylon chrysolite yellow-green tourmaline resembling chrysolite
Ceylon peridot honey yellow tourmaline resembling peridot
Chameleonite alexandrite-like tourmaline; color changes
under different lighting conditions
Cromolite green tourmaline, originally considered
Cr-bearing tourmaline
Deuterolite alexandrite-like tourmaline, probably a Cr-bearing
dravite; color changes under different
lighting conditions
Emeralite emerald-green tourmaline
Indicolite blue tourmaline, probably elbaite-schorl series
Iochroite violet tourmaline
Moor’s Head tourmaline light-colored tourmaline with black top
Paraibaíte blue-green (neon blue) Cu-bearing elbaite, originally
described from the Paraíba State of Brazil
Rubellite pink or red tourmaline, probably elbaite
Siberian ruby red tourmaline resembling ruby
Siberite purple rubellite, probably elbaite
Verdelite green tourmaline, probably elbaite-schorl series
Watermelon tourmaline color-zoned tourmaline with pink interiors
and green rims, generally elbaite
Note: Source of terms: Dietrich (1985).
appendix Table 2. Obsolete, discredited, or misidentified tourmaline
species
Name Additional information
Aluminobuergerite Term for suggested end-member Na1–x Al3 Al6
(BO3)3 O3–x (OH)1–x or synthetic NaAl3 Al6 (BO3)3 O3 (OH)
Alumoelbaite Elbaite
Chrome tourmaline Green Cr-bearing tourmaline or chromdravite
Cockle Schorl
Coronite Dravite
Eicotourmaline Tourmaline-like mineral without boron
Ferridravite Povondraite
Ferroelbaite Schorl
Ferroschorlite Schorl
Gouverneurite Dravite
Iochroite Tourmaline
Jochroit Tourmaline
Lithia tourmaline Elbaite
Magnesioschorlite Dravite
Magnodravite Extremely Mg-rich uvite-like tourmaline
Mineral H Ti-rich tourmaline from pegmatites
Oxytourmaline Tourmaline with noteworthy replacement of O for (OH,F)
Pierrepontite Schorl
Schirl Schorl
Schorlite Schorl
Schorlomite Name mistakenly used for schorl; schorlomite
is an accepted garnet group mineral
Taltalite Mixture of tourmaline with Cu-ore or green tourmaline
Titanschorl (schörl) Rutile
Titantourmaline Ti-rich tourmaline
Titanturmaline Ti-rich tourmaline
Tsilaisite Mn-rich tourmaline, mostly in solid solution with elbaite
Zeuxite Acicular tourmaline, Fe3+
-rich tourmaline, green tourmaline
Note: Sources of terms:Wang and Hsu (1966); Fleischer (1969); Chukhrov (1981);
Dietrich (1985); Bayliss (2000).
et al. 1994). (2) Secondary ion microprobe spectroscopy has
provided information on spatial distribution of light elements as
well as trace and minor elements in tourmaline (e.g., Wilson and
Long 1983; Grew et al. 1990; Harris et al. 1992; Hawthorne et
al. 1995). (3) Mössbauer spectroscopy has been used to evaluate
HENRY ET AL.: TOURMALINE NOMENCLATURE912
Normalization procedure 1—fixed number of O atoms.
Assume OH fills the fourV+Wsites after accounting for F and Cl.
In essence, this implies normalization of the cations on the basis
of 31 anions (O + F + OH), 29 oxygen atoms (both normalization
assuming B is calculated by stoichiometry, B = 3 apfu) or 24.5
oxygen atoms (without B calculation). Generally, this approach
provides a good first approximation (Clark 2007). However, if
there is significant O2–
substitution for OH1–
, the oxygen normalization
factor will be underestimated as will be the number
of cations. In fact, significant amounts O2–
substitution for OH−1
have been demonstrated in uvite (Taylor et al. 1995) and in several
other tourmaline compositions (e.g., Povondra 1981; Povondra
and Novák 1986; Dyar et al. 1998). Consequently, this approach
should be used with those tourmalines in which it is reasonably
assured that OH+F+Cl∼4 (e.g., Burns et al. 1994).
Normalizationprocedure2—fixedY+Z+Tcations. Normalize
the sum of the T+Z+Y cations to 15. This is the recommended
normalization approach for tourmaline with low Li contents
and minor B in the tetrahedral site. It assumes that there are no
vacancies in the Y, Z and T sites, a valid assumption based on a
majority of crystal structure refinement data. It also assumes that
the amount of tetrahedral B is not significant. However, some
recent refinement results indicate small numbers of vacancies in
Y and Z sites (<0.22 apfu; Ertl et al. 1997, 2003b; Prowatke et al.
2003).To the extent that all of the significant cations on these sites
are accurately analyzed, the structural formula can be calculated
without having to rely on estimates of the OH content or valence
statesoftransitionelements.Fortunately,thosetourmalinecontaining
even a moderate amount of Mg (>0.02 apfu) and that coexist
with minerals such as biotite, muscovite and staurolite typically
have minor-to-insignificant amounts of Li due to the preferential
partitioning of Li into these coexisting minerals (Henry and
Dutrow 1996). This normalization procedure will be appropriate
for almost all metamorphic and most igneous tourmalines.
Normalizationprocedure3—fixedSi.Normalizeonthebasis
of Si = 6.This procedure is useful for normalization of tourmalines
with significant amounts of unanalyzed elements, particularly Li.
It assumes that there is no significant tetrahedralAl or B. For many
Li-rich tourmalines this general assumption is probably reasonable
(e.g., Dutrow and Henry 2000). However, many low-Li tourmalines
are known to contain significant amounts of tetrahedral Al
and B, so that this procedure should be used with care in these
instances (e.g., Hawthorne et al. 1993; Lussier et al. 2009).
Estimation of B. Based on the total B found in a series of
high-quality wet-chemical analyses of tourmaline from various
lithologies, Henry et al. (2002) indicate that the assumption that
B fully occupies the triangular B site and can be calculated using
stoichiometric constraints (i.e., B = 3) is likely to be valid for
most petrologic occurrences of tourmaline (see also Clark 2007).
Furthermore, crystal structure refinements and bond valence
calculations indicate that there are essentially 3 B apfu in those
tourmalines that have been determined to date (Hawthorne 1996).
However, there are a few uncommon instances involving Al-rich
tourmaline (e.g., olenite) in which some B may also be tetrahedral
(Ertl et al. 1997, 2006a;Tagg et al. 1999; Hughes et al. 2000).This
feature indicates that the assumption of 3 B may not be valid in all
cases, but is likely to be reasonable for most tourmalines.
Estimation of Li. Li can be approximated by assuming that
Li fills any cation deficiency in the Y site. However, this requires
that the formula be calculated either on a fixed cation basis (e.g.,
Si = 6, see Dutrow and Henry 2000) or that Li be iteratively calculated
using a fixed number of oxygens and assuming OH+F =
4 (e.g., Burns et al. 1994; Clark 2007). (1) The Si = 6 approach
appears to work well for some Li-rich tourmaline (Li >0.7 apfu),
but tends to underestimate the Li contents (Henry et al. 2002).The
underestimate of Li is likely due to the existence of tetrahedralAl
substituting for Si will result in calculated Li being too low. Henry
et al. (2002) indicate that this procedure should be within 0.1–0.3
apfu of the correct Li values. (2) The Li by iteration and OH+F =
4.0 approach tends to result in high estimates of Li (Henry et al.
2002). For tourmalines with high values of Li (>0.7 apfu) this generally
results in an overestimate of 0.1–0.3 Li apfu. This is a more
significant problem in low-Li tourmalines, and generally appears
to indicate that these tourmalines contain significant amounts of
Li.The Li overestimates are a consequence of substantial amounts
of substitution of O for OH+F in the W and V sites in many
tourmalines or can result because of the occurrence of significant
vacancies at the Y site in Al-rich tourmalines. Nonetheless, this
procedure is very useful for tourmaline developed in Li-enriched
environments (e.g., highly fractionated granitic pegmatites).
Estimation of H. H contents can be determined by charge
balance if oxidation states can be measured or inferred, and Li
contents can be estimated or inferred. Putative oxidation state
evidence can be derived from tourmaline optical properties or by
mineral assemblages, particularly graphite-bearing assemblages
(Dyar et al. 1998). Minimal Li contents can be generally inferred
by presence of significant Mg in tourmaline, which is a function
of coexisting mineral assemblages that tend to efficiently partition
Li (Henry and Dutrow 1996). Charge-balance estimates of H are
subject to significant amounts of uncertainty, and likely result in
minimum values of H. Despite the larger amounts of uncertainty,
useful information on H contents can be derived using electron
microprobe data (e.g., Henry et al. 2002).
Estimation of Fe oxidation states. Fe3+
contents can be determinedby
chargebalanceifH contents canbemeasuredorinferred,
and Li contents can be estimated or inferred. Despite the larger
amounts of uncertainty, useful information on Fe3+
contents can be
derived using electron microprobe data (e.g., Henry et al. 1999).
However, all these procedures can have relatively large errors.
The best way would be to employ techniques such as Mössbauer
spectroscopy to determine the Fe2+
/Fe3+
ratio (e.g., Dyar et al.
1998). In some instances, the Fe2+
/Fe3+
ratio can be inferred for
tourmalines found in certain mineral assemblages. For example,
Dyar et al. (2002) found that tourmaline from graphite-bearing
metapelites contained 20–34% of the total Fe as Fe3+
.
aPPenDiX 6. rePresentative tourMaline
analyses witH Diagnostic ParaMeters For
classiFication
To illustrate the manner in which various parameters are determined
for use in the tourmaline nomenclature scheme, several
tourmalineanalysesfromtheliteraturearegiveninAppendixTable
3. These represent a range of tourmaline species and demonstrate
the manner in which an ordered tourmaline is written and presents
some of the important chemical parameters used for classification.
HENRY ET AL.: TOURMALINE NOMENCLATURE 913
appendix Table 3. Examples of complete tourmaline analyses with ordered structural formula and important parameters used for classification
Data source Grice and Ercit (1993) Povondra et al. (1998)
Sample no. 144478 49356 43293 43167 43873 55224 PAFG - Pr1
Oxides
B2O3 9.17 10.36 10.14 10.83 10.33 11.56 10.18
SiO2 30.74 33.58 32.92 34.04 35.27 36.50 34.88
Al2O3 1.40 30.62 30.70 27.33 26.69 40.07 32.31
TiO2 1.63 0.54 0.39 0.52 0.67
V2O3 0.04
Fe2O3 43.89 0.86 12.37 7.63 5.46 1.20
FeO 2.69 12.65 5.98 4.91 2.20 0.22 12.66
MnO 0.06 0.11 3.07 0.21
MgO 6.45 2.69 0.16 7.34 10.77 1.60
Li2O 0.03 0.01 0.00 1.61 0.06
CaO 0.20 0.99 1.67 0.20 0.41
Na2O 2.12 2.84 2.49 2.35 2.15 2.15 2.24
K2O 1.04 0.06 0.07 0.05 0.23
F 0.34 1.32 0.79 1.24 0.61
H2O 2.56 2.96 1.15 3.18 2.92 3.13 2.57
Subtotal 100.10 98.68 98.15 99.00 98.82 99.75 99.83
O=F 0.14 0.56 0.33 0.52 0.26
Total 100.10 98.54 97.59 99.00 98.49 99.22 99.57
31 O atoms normalization
Total B 3.059 3.089 3.099 3.173 3.001 3.181 3.007
B site: B 3.000 3.000 3.000 3.000 3.000 3.000 3.000
T site: Si 5.941 5.801 5.829 5.778 5.935 5.818 5.969
B 0.059 0.089 0.099 0.173 0.001 0.181 0.007
Al 0.000 0.110 0.072 0.049 0.064 0.001 0.025
T-site Total 6.000 6.000 6.000 6.000 6.000 6.000 6.000
Al (Total) 0.319 6.234 6.407 5.467 5.294 7.528 6.516
Z site: Al 0.319 6.000 6.000 5.418 5.230 6.000 6.000
V3+
0.006
Fe3+
5.675 0.582 0.691
Mg2+
0.079
Z-site Total 6.000 6.000 6.000 6.000 6.000 6.000 6.000
Y site: Al 0.125 0.335 1.526 0.491
Ti 0.212 0.072 0.050 0.066 0.086
Fe3+
0.707 0.112 1.648 0.392 0.155
Fe2+
0.435 1.828 0.886 0.697 0.310 0.029 1.812
Mn2+
0.009 0.016 0.414 0.030
Mg 1.858 0.693 0.042 1.857 2.623 0.408
Li 0.022 0.003 0.002 1.030 0.041
Y-site Total 3.000 3.000 3.000 3.000 3.001 3.000 3.024
Ca 0.038 0.180 0.301 0.034 0.075
Na 0.794 0.951 0.855 0.773 0.702 0.664 0.743
K 0.256 0.013 0.016 0.011 0.050
X-site vacancy 0.036 0.091 0.047 0.301 0.131
X-site Total 1.051 1.000 1.000 1.000 1.013 1.000 1.000
V+W site: OH 3.300 3.411 1.358 3.600 3.278 3.328 2.933
V site: OH 3.000 3.000 1.358 3.000 3.000 3.000 2.933
V site: O 1.642 0.067
W site OH 0.300 0.411 0.600 0.278 0.328
W site: F 0.186 0.739 0.420 0.625 0.330
W site O 0.700 0.403 0.261 0.400 0.302 0.047 0.670
V,W-site Total 4.000 4.000 4.000 4.000 4.000 4.000 4.000
X-site primary group: Alkali Alkali Alkali Alkali Alkali Alkali Alkali
W-site species series: Oxy- Hydroxy- Fluor- Hydroxy- Fluor- Fluor- OxyV-site
dominant anion: OH OH O OH OH OH OH
Y-site dominant divalent cation: Mg2+
Fe2+
Fe2+
Mg2+
Mg2+
Mn2+
Fe2+
Y-site dominant trivalent cation: Fe3+
Al3+
Fe3+
Fe3+
Al3+
Al3+
Z-site dominant trivalent cation: Fe3+
Al3+
Al3+
Al3+
Al3+
Al3+
Al3+
Plotting parameters
R2+
/(R2+
+ 2Li) 1.000 0.983 1.000 0.997 0.999 0.177 0.965
Ca/(Ca+Na+K) 0.042 0.189 0.297 0.049 0.087
Vac/(Na+K+Vac) 0.036 0.095 0.057 0.312 0.142
W site O/(O+OH+F) 0.700 0.403 0.261 0.400 0.302 0.047 0.670
Alkali group
Subgroup subgroup 3 subgroup 1 subgroup 5 subgroup 1 subgroup 1 subgroup 2 subgroup 3
Tourmaline species Povondraite Schorl Fluor-buergerite Dravite Fluor-dravite “Fluor-elbaite” “Oxy-schorl”