Content uploaded by Pedro Aguiar
Author content
All content in this area was uploaded by Pedro Aguiar on Jan 20, 2015
Content may be subject to copyright.
Mushro om elbaite from the Kat Chay mine,
Momeik, near Mogok, Myanmar: I. Crystal chemistry by
SREF, EMPA, MAS NMR and M˛ssbauer spec troscopy
A. J. LUSSIER
1
, P. M. AGUIAR
2
, V. K. MICHAELIS
2
, S. KROEKER
2
, S. HERWIG
1
, Y. ABDU
1
AND F. C. HAWTHORNE
1,
*
1
Department of Geological Sciences, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2
2
Department of Chemistry, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2
[Received 15 April 2008; Accepted 26 August 2008]
ABS TR AC T
Tourmaline from the Kat Chay mine, Momeik, near Mogok, Shan state, Myanmar, shows a variety of
habits that resemble mushrooms, and it is commonly referred to as ‘mushroom tourmaline’. The
structure of nine single crystals of elbaite, ranging in colour from pink to white to black and purple,
extracted from two samples of mushroom tourmaline from Mogok, have been refined (SREF) to R
indices of ~2.5% using graphite-monochromated Mo-KaX-radiation.
11
B and
27
Al Magic Angle
Spinning Nuclear Magnetic Resonance spectroscopy shows the presence of
[4]
B and the absence of
[4]
Al in samples with transition-metal content low enough to prevent paramagnetic quenching of the
signal. Site populations were assigned from refined site-scattering values and unit formulae derived
from electron-microprobe analyses of the crystals used for X-ray data collection.
57
Fe Mo¨ssbauer
spectroscopy shows that both Fe
2+
and Fe
3+
are present, and the site populations derived by structure
refinement show that there is no Fe at the Zsite; hence all Fe
2+
and Fe
3+
occurs at the Ysite. The
57
Fe
Mo¨ssbauer spectra also show peaks due to intervalence charge-transfer involving Fe
2+
and Fe
3+
at
adjacent Ysites. Calculation of the probability of the total amount of Fe occurring as Fe
2+
!Fe
3+
pairs
for a random short-range distribution is in close accord with the observed amount of Fe involved in
Fe
2+
!Fe
3+
, indicating that there is no short-range order involving Fe
2+
and Fe
3+
in these tourmalines.
KEY WORDS :elbaite, tourmaline, crystal-structure, electron-microprobe analysis, Mo¨ ssbauer spectroscopy,
magic-angle-spinning nuclear magnetic resonance, Mogok, Myanmar.
Introduction
THE area around Mogok is the main gem-
producing region of Mandalay Division,
Myanmar. Mogok is in north central Myanmar
and is well known for being the source of some of
the world’s finest rubies (Keller, 1983).
Tourmalines from the granitic pegmatites in the
region of Momeik, north-east of Mogok and
north-west of Sakangyi, show unusual mushroom-
like habits (Fig. 1). The granitic pegmatites occur
within a central belt of evolved tin-tungsten
granites and associated topaz-bearing pegmatites
that stretches north!south. The pegmatites occur
as veins and dykes cutting granitoid, migmatite,
gneiss and schist, and range from 2 to 5 m wide
and 30 to 150 m long (Zaw, 1998; Hia et al.,
2005; Themelis, 2007). They contain quartz,
orthoclase, albite, microcline microperthite and
muscovite, with accessory minerals that include
biotite, tourmaline, beryl, garnet, topaz, lepidolite,
magnetite, wolframite, cassiterite and rare colum-
bite (Themelis, 2007). The pegmatites are
commonly zoned; where tourmaline is present, it
is usually confined to the outer zone.
The chemical variability of tourmaline makes it
useful as a petrogenetic indicator for the rocks in
* E-mail: frank_hawthorne@umanitoba.ca
DOI: 10.1180/minmag.2008.072.3.747
Mineralogical Magazine, June 2008, Vol. 72(3), pp. 747–761
#2008 The Mineralogical Society
which it occurs, particularly as a record of
progressive geochemical evolution of meta-
morphic systems (Henry and Dutrow, 1992,
1996; Henry and Guidotti, 1985; Povondra and
Novak, 1986) and crystallization sequences in
pegmatitic systems (Aurisicchio et al., 1999; Dyar
et al., 1998; Nova´k and Povondra, 1995; Nova´k et
al., 1999; Selway et al., 1998, 1999, 2000a,b,
2002; Neiva et al.,2007).Therehasbeen
considerable work done on the characterization
(e.g. Hawthorne et al., 1993; Burns et al., 1994;
Grice and Ercit, 1993; Grice et al., 1993; Taylor
et al., 1995; Dyar et al., 1998; Francis et al., 1999;
Bloodaxe et al., 1999; Kalt et al., 2001; Ca´mara et
al., 2002; Schreyer et al., 2002; Ertl and Hughes,
2002; Ertl et al., 2003a,b, 2004, 2005; Hughes et
al., 2000, 2004; Marschall et al., 2004; Bosi and
Lucchesi, 2004; Bosi et al., 2004, 2005) and
understanding (Hawthorne, 1996, 2002) of site
occupancy in tourmaline. However, the complete
derivation of site populations (and accurate bulk
compositions) is not trivial, and our knowledge of
the crystal chemistry of all the tourmaline
minerals is still far from complete. Here, we
examine the variation in crystal structure,
chemical composition and ordering of cations
over the tetrahedrally and octahedrally coordi-
nated sites in two samples of mushroom tourma-
line from the Mogok region, Shan state,
Myanmar, and characterize the unusual habits of
tourmaline from the Kat Chay mine, Momeik,
Mandalay Division, Myanmar.
Sample appearance
This work focuses on two samples of tourmaline,
labelled SHM and SHP, from this area, both of
which resemble mushrooms. Sample SHM
(Fig. 1a) consists of a base of black dense
tourmaline that grades into a greyish-white
aggregate of acicular crystals. This greyish-
white region is bounded by a thin band of black
that is mantled by a thick cap of pink tourmaline.
On the outside, sample SHP (Fig. 1b) appears as a
greyish purple aggregate of diverging highly
elongated crystals; unlike the first sample, only
slight colour variations are present.
Experimental
To adequately characterize the crystal chemistry
of these tourmalines, they were examined by
single-crystal structure refinement, electron
microprobe analysis (EMPA),
11
Band
27
Al
Magic angle spinning nuclear magnetic resonance
(MAS NMR) and
57
Fe Mo¨ssbauer spectroscopy.
The colours listed in Table 1 may be correlated
with the samples in Fig. 1. Care was taken to
ensure that all techniques examined either the
same crystals, or analogous material where the
method required many milligrams of sample.
X-ra y da ta collec tion
Crystals selected for the study were ground to
approximate spheres (three were not ground due
to their acicular nature) and mounted on thin
tapered glass fibres. The unit-cell dimensions of
the crystals were determined using a Bruker P4
automated four-circle single-crystal diffract-
ometer equipped with a serial detector and a
Mo-KaX-ray source. The software used for data
reduction and least-squares refinement is part of
the SHELXTL PC package.
For six crystals, twenty-five reflections between
10 and 30º2ywere centred using a random search
method. Data were collected from 4 to 60º2ywith
FIG. 1. Tourmalines from Mogok: (a) pink-white-black
Mogok mushroom tourmaline (SHM); (b) purple-black
‘mushroom’ tourmaline (SHP).
74 8
A. J. LUSSIER ET AL.
index ranges from 0, 0, !11 to 23, 23, 11. Scan
speeds were set according to the size and shape of
each crystal; larger spheres were set for a variable
scan-speed depending on the intensity of the
reflection, whereas smaller fibrous crystals were
scanned at a slow fixed scan-speed; the scan range
was set to 1.2º. An intense check reflection was
used to monitor intensity variation during data
collection; no significant changes were observed.
For each data set, c-scan absorption corrections
were applied, together with the usual geometrical
corrections, and the data were reduced to structure
factors. For three crystals, more than a hemisphere
of data was collected for each on a Bruker P4
automated four-circle single-crystal diffractometer
equipped with an Apex CCD detector and a Mo-
KaX-ray source. Reflections were measured out to
60º2ywith a frame width of 0.2º and a frame time
of 20!40 s. Unit-cell dimensions were determined
on all reflections with |I| > 10sI. Absorption
corrections were done using the program SADABS
(Sheldrick, 1998). The data were then corrected for
Lorentz, polarization and background effects,
averaged and reduced to structure factors.
Various data-collection and refinement information
are given in Table 1.
Structure ref|nement
All calculations were done with the SHELXTL PC
Plus software package; each refinement was done
in the space group R3m. In the preliminary stages
of refinement, the Z-site scattering was not
observed to diverge from full Al occupancy, and
hence the Zsite was fixed at full occupancy of Al
in the final stages of refinement. Refinement of all
variable positional parameters, anisotropic-displa-
cement parameters, site occupancies for the X,Y
and Tsites, and an empirical isotropic-extinction
correction converged to final R-indices (Table 1)
ranging from 1.7 to 3.9%. Structures were tested
for absolute orientation and transformed as
appropriate. Refined positional parameters and
equivalent isotropic-displacement parameters are
given in Table 2, selected interatomic distances in
Table 3, and refined site-scattering values in
Table 4.
Electron-microprobe analysis
The crystals used for X-ray data collection were
mounted in epoxy on 2.5 cm diameter Perspex
1
discs, ground, polished, carbon-coated and
analysed with a Cameca SX-100 electron-micro-
TABLE 1. Data collection and refinement information for Mogok tourmaline samples.
SHM1 SHM2 SHM3 SHM3a SHM3e SHM5 SHP1 SHP2 SHP3
a(A
˚) 15.799(1) 15.774(1) 15.818(1) 15.9005(5) 15.8034(4) 15.7972(4) 15.8063(16) 15.8402(16) 15.8375(18)
c(A
˚) 7.094(1) 7.079(1) 7.094(1) 7.1241(2) 7.0880(2) 7.0883(2) 7.0923(7) 7.1015(13) 7.0996(10)
V(A
˚
3
) 1533.49 1525.41 1537.18 1559.84 1533.3(1) 1531.9(1) 1534.54 1543.13 1542.19
Space group R3mR3mR3mR3mR3mR3mR3mR3mR3m
Colour pink colourless black black black very dark grey light purple-pink purple black
Crystal size 0.2 mm
sphere
0.05 mm
equant
0.2 mm
sphere
0.1 mm
block
0.12 mm
sphere
0.15 mm
block
0.2 mm
equant
0.2 mm
equant
0.09 mm
sphere
Z33333 3 3 33
Rad/Mon Mo-Ka/Gr Mo-Ka/Gr Mo-Ka/Gr Mo-Ka/Gr Mo-Ka/Gr Mo-Ka/Gr Mo-Ka/Gr Mo-Ka/Gr Mo-Ka/Gr
No. reflections >10sI!!!5991 7352 8048 !!!
# unique reflections 1107 1097 1109 1109 1094 1096 1110 1109 1109
R
1
(%) 2.20 3.92 1.94 2.55 1.71 1.94 1.85 3.15 3.25
wR
2
(%) 5.24 9.86 5.29 6.25 4.48 5.16 4.89 7.92 6.85
GooF* 1.058 1.089 1.178 1.148 1.151 1.176 1.154 1.059 0.955
* GooF = goodness of fit
CRYSTAL CHEMISTRY OF MUSHROOM ELBAITE
749
TABLE 2. Positional and equivalent anisotropic displacement parameters for Mogok tourmaline samples.
SHM1 SHM2 SHM3 SHM3a SHM3e SHM5 SHP1 SHP2 SHP3
Xx 000000000
y000000000
z0.8404 0.8404 0.8404 0.8408 0.8404 0.8404 0.8404 0.8404 0.8404
U
eq
0.0299(10) 0.0339(17) 0.0316(9) 0.0354(11) 0.0252(7) 0.0255(9) 0.0330(8) 0.0347(12) 0.0352(18)
Yx 0.06066(4) 0.06063(7) 0.06006(3) 0.05959(3) 0.06019(3) 0.06032(3) 0.06088(3) 0.06056(4) 0.06058(7)
y0.93934(4) 0.93937(7) 0.93994(3) 0.94041(3) 0.93981(3) 0.93968(3) 0.93912(3) 0.93944(4) 0.93942(7)
z0.4310(5) 0.4258(9) 0.4271(4) 0.4327(6) 0.4258(4) 0.4260(4) 0.4347(4) 0.4355(6) 0.4350(9)
U
eq
0.0108(4) 0.0104(5) 0.0112(3) 0.0116(3) 0.0114(2) 0.0120(3) 0.0107(3) 0.0114(4) 0.0124(5)
Zx 0.26011(4) 0.26036(7) 0.26032(4) 0.26042(5) 0.26031(3) 0.26022(4) 0.26018(3) 0.26048(5) 0.26046(8)
y0.29679(4) 0.29683(7) 0.29698(4) 0.29730(5) 0.29692(3) 0.29688(4) 0.29690(3) 0.29709(5) 0.29710(8)
z0.4599(5) 0.4560(8) 0.4562(4) 0.4581(5) 0.4556(3) 0.4557(4) 0.4613(4) 0.4600(6) 0.4581(9)
U
eq
0.00775(13) 0.0081(2) 0.00797(13) 0.00766(16) 0.00897(10) 0.00942(12) 0.00768(11) 0.00767(17) 0.0082(2)
Tx 0.18955(4) 0.18949(6) 0.18967(3) 0.19000(4) 0.18956(3) 0.18958(3) 0.18972(3) 0.18984(5) 0.18978(7)
y0.19163(4) 0.19132(6) 0.19158(3) 0.19189(4) 0.19146(3) 0.19143(3) 0.19174(3) 0.19175(4) 0.19168(7)
z0.0687(5) 0.0642(8) 0.0651(4) 0.0687(5) 0.0642(3) 0.0643(4) 0.0706(4) 0.0697(6) 0.0683(9)
U
eq
0.00569(17) 0.0063(3) 0.00570(16) 0.0050(2) 0.00680(13) 0.00720(16) 0.00582(14) 0.0056(2) 0.0059(3)
Bx0.89090(12) 0.89103(19) 0.89072(11) 0.89009(15) 0.89068(8) 0.89071(10) 0.89085(9) 0.89072(14) 0.8908(2)
y0.10910(12) 0.10897(19) 0.10928(11) 0.10991(15) 0.10932(8) 0.10929(10) 0.10915(9) 0.10928(14) 0.1092(2)
z0.6149(7) 0.6114(10) 0.6121(5) 0.6162(7) 0.6108(4) 0.6110(5) 0.6164(5) 0.6159(7) 0.6165(12)
U
eq
0.0089(5) 0.0085(8) 0.0087(5) 0.0103(7) 0.0097(3) 0.0101(4) 0.0088(4) 0.0091(6) 0.0104(11)
O(1)
o
x000000000
y000000000
z0.2897(7) 0.2866(12) 0.2857(7) 0.2822(9) 0.2855(5) 0.2858(6) 0.2901(6) 0.2887(9) 0.2867(13)
U
eq
0.0251(9) 0.0212(3) 0.0255(8) 0.0316(13) 0.0236(6) 0.0238(7) 0.0271(7) 0.0296(12) 0.035(2)
O(1)
d
x0.0055(6) 0.0059(7) 0.0060(4) 0.0075(4) 0.0054(3) 0.0052(5) 0.0074(2) 0.0077(4) 0.0089(6)
y!0.0055(6) !0.0059(7) !0.0060(4) !0.0075(4) !0.0054(3) !0.0052(5) !0.0074(2) !0.0077(4) !0.0089(6)
z0.2897(7) 0.2869(12) 0.2859(7) 0.2824(9) 0.2857(5) 0.2860(6) 0.2902(6) 0.2890(9) 0.2870(13)
U
eq
0.0174(18) 0.012(2) 0.0158(15) 0.016(2) 0.0155(10) 0.0166(14) 0.0119(11) 0.0130(17) 0.013(3)
O(2)
0
x0.93952(8) 0.93998(13) 0.93942(7) 0.93885(10) 0.93966(5) 0.93962(7) 0.93951(6) 0.93938(10) 0.93941(14)
y0.06048(8) 0.06002(13) 0.06058(7) 0.06115(10) 0.06034(5) 0.06038(7) 0.06049(6) 0.06062(10) 0.06060(14)
z0.5785(6) 0.574(1) 0.5772(5) 0.5866(7) 0.5754(4) 0.5752(5) 0.5816(5) 0.5833(7) 0.5834(11)
U
eq
0.0156(5) 0.0157(7) 0.0165(4) 0.0198(9) 0.0172(3) 0.0174(4) 0.0164(4) 0.0189(6) 0.0187(9)
750
A. J. LUSSIER ET AL.
O(2)
d
x0.9464(4) 0.9337(6) 0.9465(3) 0.9307(4) 0.9329(3) 0.9327(3) 0.9320(3) 0.9313(4) 0.9305(6)
y0.0674(4) 0.0538(6) 0.0677(3) 0.0530(4) 0.0536(3) 0.0534(3) 0.0530(2) 0.0525(4) 0.0518(6)
z0.5785(6) 0.574(1) 0.5772(5) 0.5867(7) 0.5754(4) 0.5753(5) 0.5817(5) 0.5834(7) 0.5835(11)
U
eq
0.0113(6) 0.0123(9) 0.0119(5) 0.013(1) 0.0131(4) 0.0130(5) 0.0112(5) 0.0129(7) 0.0115(11)
O(3) x0.13169(9) 0.13093(14) 0.13201(8) 0.13384(10) 0.13152(6) 0.13145(8) 0.13226(7) 0.13279(11) 0.13317(16)
y0.86831(9) 0.86907(14) 0.86799(8) 0.86616(10) 0.86848(6) 0.86855(8) 0.86774(7) 0.86722(11) 0.86683(16)
z0.5609(6) 0.5570(9) 0.5571(5) 0.5592(6) 0.5564(4) 0.5565(5) 0.5630(4) 0.5613(7) 0.559(1)
U
eq
0.0138(5) 0.0135(7) 0.0136(4) 0.0115(5) 0.0141(3) 0.0142(4) 0.0135(3) 0.0139(5) 0.0153(9)
O(4) x0.90617(8) 0.90603(13) 0.90621(7) 0.90646(9) 0.90600(6) 0.90603(7) 0.90618(6) 0.90632(10) 0.90633(14)
y0.09383(8) 0.09397(13) 0.09379(7) 0.09354(9) 0.09400(6) 0.09397(7) 0.09382(6) 0.09368(10) 0.09367(14)
z!0.0046(6) !0.0084(9) !0.0077(5) !0.0027(6) !0.0087(4) !0.0088(5) !0.0020(4) !0.0020(6) !0.002(1)
U
eq
0.0114(4) 0.0123(6) 0.0112(4) 0.0118(5) 0.0128(3) 0.0138(3) 0.0111(3) 0.0107(5) 0.0116(8)
O(5) x0.09349(8) 0.09345(14) 0.09346(8) 0.09353(10) 0.09340(6) 0.09344(7) 0.09354(6) 0.09352(10) 0.09351(14)
y0.90651(8) 0.90655(14) 0.90654(8) 0.90647(10) 0.90660(6) 0.90656(7) 0.90646(6) 0.90648(10) 0.90649(14)
z!0.0261(6) !0.0305(9) !0.0289(5) !0.0240(6) !0.0303(4) !0.0304(5) !0.0239(4) !0.0242(6) !0.0251(10)
U
eq
0.0114(4) 0.0126(6) 0.0120(4) 0.0112(5) 0.0133(3) 0.0135(3) 0.0115(3) 0.0118(5) 0.0115(7)
O(6) x0.18430(10) 0.18405(16) 0.18509(9) 0.18696(12) 0.18463(7) 0.18450(9) 0.18492(8) 0.18557(13) 0.18567(18)
y0.19473(10) 0.19456(16) 0.19522(9) 0.19666(12) 0.19481(7) 0.19475(8) 0.19518(8) 0.19584(12) 0.19610(19)
z0.2944(5) 0.2896(9) 0.2902(5) 0.2937(6) 0.2895(4) 0.2897(4) 0.2958(4) 0.2947(6) 0.294(1)
U
eq
0.0095(3) 0.0097(4) 0.0098(3) 0.0102(3) 0.01050(19) 0.0108(2) 0.0095(2) 0.0098(3) 0.0110(5)
O(7) x0.28570(9) 0.28578(15) 0.28575(9) 0.28590(11) 0.28580(7) 0.28580(8) 0.28576(7) 0.28592(12) 0.28588(16)
y0.28612(10) 0.28599(16) 0.28605(9) 0.28571(12) 0.28603(7) 0.28598(8) 0.28605(8) 0.28574(12) 0.28560(18)
z!0.0080(5) !0.0117(9) !0.0120(5) !0.0107(6) !0.0127(4) !0.0123(4) !0.0068(4) !0.0088(6) !0.0096(9)
U
eq
0.0086(3) 0.0092(4) 0.0088(2) 0.0085(3) 0.00983(18) 0.0101(2) 0.00852(19) 0.0089(3) 0.0094(5)
O(8) x0.27006(11) 0.27004(17) 0.27027(10) 0.27094(13) 0.27017(7) 0.27018(9) 0.27016(8) 0.27052(13) 0.27070(19)
y0.20969(11) 0.20941(17) 0.20970(10) 0.21014(13) 0.20959(7) 0.20966(9) 0.20960(8) 0.20984(13) 0.20995(19)
z0.6305(5) 0.6270(9) 0.6267(5) 0.6286(6) 0.6262(4) 0.6264(4) 0.6319(4) 0.6302(6) 0.6292(9)
U
eq
0.0092(3) 0.0094(4) 0.0096(3) 0.0105(4) 0.01048(19) 0.0109(2) 0.0092(2) 0.0099(3) 0.0104(6)
H(3) x0.1302(14) 0.129(2) 0.1267(14) 0.135(3) 0.1294(10) 0.1310(13) 0.1318(11) 0.1298(18) 0.120(2)
y0.8697(14) 0.871(2) 0.8733(14) 0.865(3) 0.8706(10) 0.8690(13) 0.8682(11) 0.8702(18) 0.880(2)
z0.693(3) 0.694(3) 0.689(3) 0.694(3) 0.687(2) 0.689(3) 0.695(3) 0.695(3) 0.684(4)
*U
eq
0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015
* fixed during refinement;
o
= O(1) and O(2) ordered
d
= O(1) and O(2) disordered
CRYSTAL CHEMISTRY OF MUSHROOM ELBAITE
751
TABLE 3. Selected interatomic distances (A
˚) for Mogok tourmaline samples.
SHM1 SHM2 SHM3 SHM3a SHM3e SHM5 SHP1 SHP2 SHP3
T!O(4) 1.6163(9) 1.608(1) 1.6153(8) 1.623(1) 1.6130(6) 1.6122(8) 1.6157(7) 1.617(1) 1.612(2)
T!O(5) 1.628(1) 1.626(2) 1.6290(9) 1.636(1) 1.6282(7) 1.6288(9) 1.6293(8) 1.632(1) 1.630(2)
T!O(6) 1.605(2) 1.600(3) 1.601(2) 1.607(2) 1.601(1) 1.602(1) 1.601(1) 1.602(2) 1.608(3)
T!O(7) 1.601(1) 1.600(2) 1.603(1) 1.611(2) 1.603(1) 1.602(1) 1.602(1) 1.606(2) 1.604(3)
<T!O> 1.613 1.609 1.612 1.619 1.611 1.611 1.612 1.614 1.614
B!O(2) 1.359(4) 1.366(6) 1.361(4) 1.365(5) 1.368(3) 1.367(3) 1.359(3) 1.361(5) 1.359(9)
B!O(8) 62 1.381(2) 1.377(3) 1.380(2) 1.383(3) 1.377(2) 1.378(2) 1.380(2) 1.383(3) 1.385(4)
<B!O> 1.374 1.373 1.374 1.377 1.374 1.374 1.373 1.376 1.376
X!O(2) 63 2.493(4) 2.503(6) 2.503(3) 2.476(4) 2.503(2) 2.505(3) 2.477(3) 2.475(4) 2.475(6)
X!O(4) 63 2.792(3) 2.780(4) 2.786(2) 2.807(3) 2.786(2) 2.785(2) 2.800(2) 2.802(3) 2.802(5)
X!O(5) 63 2.727(3) 2.711(4) 2.722(2) 2.750(3) 2.716(2) 2.716(2) 2.735(2) 2.740(3) 2.736(4)
<X!O> 2.671 2.665 2.670 2.678 2.668 2.669 2.671 2.672 2.671
Y!O(1)
o
1.940(3) 1.927(4) 1.927(2) 1.959(3) 1.9249(15) 1.927(2) 1.957(2) 1.961(3) 1.966(5)
Y!O(2)
o
62 1.960(2) 1.955(3) 1.967(2) 1.993(2) 1.961(1) 1.961(1) 1.962(1) 1.967(2) 1.969(3)
Y!O(3)
o
2.150(3) 2.133(4) 2.176(2) 2.234(3) 2.160(2) 2.154(2) 2.155(2) 2.172(3) 2.177(5)
Y!O(6)
o
62 1.958(2) 1.950(3) 1.979(2) 2.026(2) 1.968(1) 1.963(1) 1.971(1) 1.995(2) 1.994(3)
<Y!O> 1.988 1.978 1.999 2.039 1.990 1.988 1.996 2.010 2.012
Y!O(1)
d
1.81(1) 1.79(2) 1.79(1) 1.80(1) 1.798(8) 1.81(1) 1.787(6) 1.785(9) 1.77(1)
Y!O(1’)
d
2.009(8) 2.00(1) 2.002(6) 2.055(7) 1.992(5) 1.991(7) 2.051(4) 2.059(6) 2.080(9)
Y!O(2)
d
62 1.882(4) 1.883(8) 1.886(4) 1.902(4) 1.885(3) 1.883(4) 1.876(3) 1.874(5) 1.869(7)
Y!O(2’)
d
62 2.042(4) 2.029(8) 2.049(4) 2.087(5) 2.040(3) 2.042(4) 2.051(3) 2.062(5) 2.073(7)
Y!O(3)
d
2.151(3) 2.134(4) 2.177(2) 2.235(3) 2.161(2) 2.155(2) 2.156(2) 2.173(3) 2.178(5)
Y!O(6)
d
62 1.958(2) 1.951(3) 1.979(2) 2.026(2) 1.968(1) 1.963(1) 1.971(1) 1.995(2) 1.995(3)
<Y!O> 1.978 1.968 1.987 2.027 1.981 1.980 1.983 1.996 1.997
Z!O(3) 1.959(1) 1.964(2) 1.961(1) 1.960(1) 1.9621(7) 1.9607(9) 1.9585(9) 1.962(1) 1.959(2)
Z!O(6) 1.866(2) 1.870(3) 1.865(1) 1.852(2) 1.869(1) 1.868(1) 1.861(1) 1.859(2) 1.850(3)
Z!O(7) 1.888(2) 1.882(2) 1.884(1) 1.884(2) 1.882(1) 1.883(1) 1.885(1) 1.879(2) 1.884(3)
Z!O(7’) 1.942(1) 1.936(2) 1.943(1) 1.958(2) 1.942(1) 1.942(1) 1.943(1) 1.949(2) 1.946(3)
Z!O(8) 1.882(2) 1.879(2) 1.881(1) 1.879(2) 1.880(1) 1.879(1) 1.881(1) 1.881(2) 1.875(3)
Z!O(8’) 1.897(2) 1.898(3) 1.900(1) 1.912(2) 1.899(1) 1.898(1) 1.900(1) 1.901(2) 1.906(3)
<Z!O> 1.906 1.905 1.906 1.908 1.906 1.905 1.905 1.905 1.903
o, d
see Table 2
TABLE 4. Site scattering values (e.p.f.u.*) derived from SREF and EMPA.
—— Xsite —— —— Ysite —— —— Tsite ——
SREF EMPA SREF EMPA SREF EMPA
SHM1 7.7(1) 7.2 30.6(3) 28.7 81.9(4) 80.8
SHM2 8.0(2) 7.6 32.1(5) 30.3 80.0(8) 77.9
SHM3 8.7(1) 8.4 37.4(3) 35.7 82.5(3) 78.9
SHM3a 10.34(9) 9.3 47.9(3) 47.1 80.2(4) 82.4
SHM3e 8.16(9) 8.3 34.85(7) 34.2 78.2(3) 76.1
SHM5 8.3(1) 9.4 33.9(1) 34.1 77.7(3) 75.3
SHP1 7.5(1) 7.5 29.9(2) 29.2 79.7(3) 82.3
SHP2 8.5(2) 8.5 37.3(3) 35.9 80.5(5) 80.6
SHP3 8.7(2) 7.7 38.6(3) 35.3 80.9(5) 81.5
<dev.> —— <0.5> —— —— <1.5> —— —— <1.6> ——
* electrons per formula unit
752
A. J. LUSSIER ET AL.
probe operating under the following conditions in
wavelength-dispersion mode: excitation voltage:
15 kV; specimen current: 20 nA; beam size:
5mm; peak count-time: 20 s; and background
count-time: 10 s. The following crystals and
standards were used for KaX-ray lines: TAP: Al,
kyanite; Mg, forsterite; Na, albite; F, fluoro-
riebeckite; PET: Si, diopside; P, VP
2
O
7
; K,
orthoclase; Ca, diopside; LiF: Ti, titanite; V,
VP
2
O
7
; Cr, chromite; Mn, spessartine; Fe,
fayalite; Zn, gahnite. Ten points on each crystal
were analysed and backscattered-electron images
of the crystals were examined to check for
compositional zoning; none was observed.
Chemical compositions are listed in Table 5.
57
Fe M˛ssbauer spectroscopy
For Mo¨ssbauer spectroscopy, ~4 mg of three
samples from the black core of sample SHM were
mixed with sucrose and ground carefully under
acetone. The mixture was then loaded into a Pb
ring (2 mm inner diameter) and covered by tape
on both sides. Mo¨ssbauer spectra were acquired in
transmission geometry using a
57
Co(Rh) point
source. The spectrometer was calibrated with the
room-temperature spectrum of a-Fe. The spectra
were analysed by a Voigt-function-based quadru-
pole-splitting distribution method using the
RECOIL
1
software package; fitting parameters
are given in Table 7.
TABLE 5. Chemical composition (wt.%) and unit formulae* (a.p.f.u.) for Mogok tourmalines.
SHM1
b
SHM2
b
SHM3
b
SHM3a
b
SHM3e
b
SHM5
b
SHP1
b
SHP2
b
SHP3
b
SiO
2
36.65 34.86 35.06 36.24 34.40 33.54 37.68 36.06 36.90
TiO
2
!! 0.47 0.60 0.39 0.87 0.04 0.68 0.35
B
2
O
3
** 12.63 13.94 13.32 11.26 14.02 14.13 11.98 12.59 10.80
Al
2
O
3
43.03 44.10 41.61 36.40 42.87 41.81 42.17 40.44 40.91
CaO 0.27 0.62 0.56 0.23 0.56 0.93 0.16 0.40 0.20
MnO 0.57 0.70 0.99 1.92 0.87 1.08 1.00 3.24 3.26
FeO !! 2.28 6.94 1.43 1.18 0.19 0.56 0.43
Li
2
O** 1.80 1.58 1.47 1.21 1.46 1.52 1.89 1.59 1.60
Na
2
O 1.94 1.72 1.99 2.43 1.95 1.88 2.11 2.15 2.14
H
2
O** 3.83 3.89 3.86 3.59 3.83 3.66 3.68 3.74 3.87
F 0.14 0.08 0.02 0.26 0.16 0.43 0.43 0.23 0.00
O=F !0.06 !0.03 !0.01 !0.11 !0.07 !0.18 !0.18 !0.10 !0.00
Total 100.80 101.45 101.62 100.99 101.88 100.87 101.15 101.58 101.91
T: Si 5.642 5.324 5.436 5.859 5.283 5.210 5.812 5.618 5.723
B 0.358 0.676 0.564 0.141 0.717 0.790 0.188 0.382 0.279
ST6.000 6.000 6.000 6.000 6.000 6.000 6.000 6.000 6.000
B 333333333
Z:Al 666666666
Y: Al 1.810 1.938 1.603 0.936 1.759 1.655 1.666 1.423 1.477
Ti !! 0.055 0.073 0.045 0.102 0.005 0.080 0.041
Fe
2+
!! 0.296 0.938 0.184 0.153 0.025 0.073 0.056
Mn
2+
0.074 0.091 0.130 0.263 0.113 0.142 0.131 0.428 0.428
Li 1.116 0.971 0.916 0.788 0.899 0.948 1.173 0.996 0.998
SY3.000 3.000 3.000 3.000 3.000 3.000 3.000 3.000 3.000
X: Na 0.579 0.509 0.598 0.762 0.581 0.566 0.631 0.649 0.643
Ca 0.045 0.101 0.093 0.040 0.092 0.155 0.026 0.067 0.033
SX0.624 0.610 0.691 0.802 0.673 0.721 0.657 0.716 0.676
OH 3.932 3.961 3.989 3.867 3.922 3.789 3.790 3.897 4.000
F 0.068 0.039 0.011 0.133 0.078 0.211 0.210 0.113 0.000
* Formulae normalized to 31 anions
** B
2
O
3
, Li
2
O and H
2
O calculated by stoichiometry: B = (9 !Si); Li = 15 !(T+B+Z+Y); OH + F = 4 a.p.f.u.
CRYSTAL CHEMISTRY OF MUSHROOM ELBAITE
753
Magic-angle-spinning nuclear magnetic resonance (MAS
NMR) spectroscopy
The
11
B MAS NMR spectra were acquired at
14.1T (n
L
= 192.4 MHz) on a Varian Inova 600
device, using a 3.2 mm double-resonance probe
(Varian-Chemagnetics) spinning at 12!24 kHz. A
short 0.5 ms pulse of nutation frequency n
rf
= 56
kHz was used in order to allow quantification of
the
[3]
B and
[4]
B signals. Recycle delays varied
significantly (0.8!20 s) over the samples.
Depending on the quality of the spectra obtained
(i.e. peak lineshapes, resolution etc.), quantification
was done either by full simulation of the line-
shapes, taking into account all transitions, line
fitting (using WINNMR, Bruker) or by direct
integration of the peak intensities. In the latter two
methods, the line intensities were corrected for
contributions from the satellite transitions to the
integrated signals as described by Massiot et al.
(1990). The simulations were done using STARS
(Skibsted et al., 1992, 1993) within the Varian
VnmrJ environment. The simulations treat the
quadrupole interaction to second order and include
contributions from all transitions (central plus
satellites). The quadrupole coupling constant and
the associated asymmetry parameter (C
Q
and Z,
respectively) for the
[3]
B site were obtained from
the lineshape of the central transition. For the
[4]
B
site, the extent of the satellite transition spinning
sidebands of select samples (i.e. where sufficient
S/N allowed measurement) was used and C
Q
was
fixed at this value (0.3 MHz) for all other samples.
27
Al (n
L
= 156.3 MHz) MAS spectra, were
acquired with a short (0.3 ms, 7º) pulse. Recycle
delays varied from 0.8 to 20 s. Spinning rates of
20 kHz were used for all samples. Spinning rates
up to 23.5 kHz were also used for selected
samples, but no change in lineshape was
observed.
Results and discussion
11
B MAS NMR spectra
Where the chemical shifts of isotopes in different
coordinations are sufficiently different, MAS
NMR can be used to detect different coordination
numbers in solids. With regard to tourmaline, we
are interested in B (
11
B) and Al (
27
Al);
[3]
B and
[4]
B are typically observed in the ranges 14 to
20 ppm and !2 to 4 ppm, respectively, and
[6]
Al
and
[4]
Al are typically observed in the ranges !5
to 20 ppm and 40 to 60 ppm, respectively.
11
B MAS NMR has been used to detect small
amounts of
[4]
B in tourmaline (e.g. Tagg et al.,
1999; Schreyer et al., 2002; Marler and Ertl,
2002). One restriction is that the contents of
paramagnetic species in the sample must be low,
as paramagnetic species quench the signal. In the
tourmaline crystals examined here, the paramag-
netic contents are quite low (Table 5) and we
were able to get both
11
B and
27
Al MAS NMR
spectra for all samples except the most Fe-rich
fraction of SHM (black).
Spectra for the SHM sample (Fig. 2) show
strong peaks centred about 14 ppm corresponding
to
[3]
B, and a weaker sharper peak at ~0 ppm
corresponding to
[4]
B. In principle, the ratio of the
intensities of the signals from
[3]
B and
[4]
B gives
the relative amounts of these species (Bray, 1999;
Michaelis et al., 2007), provided there is minimal
interaction with paramagnetic species. In tourma-
line, the amount of
[3]
B is known (3.0 a.p.f.u.) and
hence the amount of
[4]
B may be derived. Spectra
for mushroom tourmaline SHP (Fig. 3) show
significant line-broadening relative to spectra of
the SHM tourmaline. Resolution is significantly
decreased and the
[3]
B peak extends well into the
region where
[4]
B signals might be expected. This
very broad
[3]
B signal may be obscuring any small
signal arising from
[4]
Binthesesamples.
Attempts to do
11
B multiple-quantum MAS
NMR were thwarted by the very short spin-spin
(T
2
) relaxation times of the samples.
FIG. 2.
11
B MAS NMR spectra of mushroom tourmaline
SHM. The positions of the peaks corresponding to
[3]-coordinate B and [4]-coordinate B are marked;
Y
T
M
= sum of transition metals assigned to the Ysite.
754
A. J. LUSSIER ET AL.
Simulation of the
11
B MAS spectra allowed
extraction of the NMR parameters; isotropic
chemical shift, d
iso
, quadrupole coupling, C
Q
and its associated asymmetry parameter, Z
(Table 6). The quadrupole-coupling constants
and isotropic chemical shifts for the
[3]
B sites
are 2.7!2.8 MHz, typical values for borates (Kriz
and Bray, 1971; Kroeker and Stebbins, 2001). The
asymmetry parameters are very small (near zero),
indicative of an axially symmetric environment
for B, consistent with the crystal structure. In
addition, the simulations allowed estimation of
the populations of the two sites. The
[4]
B signal, in
those samples which were suitable for simulation
(i.e. which were not too degraded by para-
magnetic constituents), varied between 6 and
13% of the total intensity.
27
Al MAS NMR spectra
Figure 4 shows selected
27
Al MAS NMR spectra.
There is no signal in the 40!60 ppm region
characteristic of tetrahedrally coordinated Al, and
hence these tourmalines do not contain
[4]
Al. The
signals centred on ~0 ppm are due to octahedrally
coordinated Al, and show great variation in the
shape of the overall envelope. As discussed in
detail by Lussier et al. (2008), the peak centred at
0 ppm is a composite of signals due to Al at both
the Yand the Zsites. The sequence of spectra
SHM1 to SHM2 to SHM3 shown in Fig. 4
corresponds to gradually increasing paramagnetic
content of the samples. Thus there is considerable
paramagnetic broadening of the signals from Al at
both the Y- and the Z-sites.
Formula calculation
Usually, the unit formula for tourmaline is
normalized to 31 anions. If analysis for H, B and
Li is not done, assumptions are necessary to
calculate the formulae on the basis of 31 anions:
[OH + F] = 4 a.p.f.u., B = 3 a.p.f.u., and Li =
FIG. 3.
11
B MAS spectra of (a) SHP1 (b) SHP2 and
(c) SHP3 acquired at 14 T. These spectra are much
broader than those from the SHM tourmaline. Spinning
sidebands are denoted by asterisks, and the position of
the peak corresponding to
[4]
B is marked by the dashed
line.
TABLE 6.
11
B NMR parameters for Mogok tourmalines.
Sample Site d
iso
(ppm) C
Q
(MHz) ZPop. (%)
SHM1
[3]
B 18!1 2.75!0.1 0.1!0.1 90!3
[4]
B!0.1!1 0.3
a
1.0
a
10!3
SHM2
[3]
B 18!1 2.75!0.1 0.1!0.1 91!3
[4]
B 0.1!1 0.3
a
1.0
a
9!3
SHM3
[3]
B 12!2
b
n.d. n.d. n.d.
[4]
B n.d. n.d. n..d n.d.
SHP1 13!2 n.d. n.d. !
SHP2 13!2 n.d. n.d. !
SHP3 14!2 n.d. n.d. !
a
Values fixed for simulation.
b
Due to broadening, centre of mass taken in place of isotropic shift.
CRYSTAL CHEMISTRY OF MUSHROOM ELBAITE
755
15!S(Y+Z+T) a.p.f.u. (Henry and Dutrow,
1996; Hawthorne and Henry, 1999). For the
tourmalines examined here,
27
Al MAS NMR
spectra show that all Al is [6]-coordinated, and
11
B MAS NMR spectra show that there is
significant
[4]
B present at the Tsite. Thus we set
T= Si + B and normalized the formulae based on
31 anions with [OH + F] = 4 a.p.f.u., Si + B =
9 a.p.f.u., and Li = 15 !S(Y + Z + T) a.p.f.u. The
resulting formulae are given in Table 5; all
correspond to elbaite (Hawthorne and Henry,
1999), despite the great variation in colour.
Recently, Ertl et al. (2007) reported a crystal-
structure refinement of an olenitic mushroom
tourmaline from Momeik, Myanmar. All the
structures refined in the present study are elbaite,
and no olenite was encountered in our work.
Site populations
Site-scattering values (Table 4), unit formulae
calculated from electron-microprobe analysis
(Table 5), mean bond-lengths (Table 3), and
indications of the presence or absence of
[4]
B
and
[4]
Al from MAS NMR spectroscopy allow us
to derive site populations for these tourmalines.
The T s ite
The
27
Al MAS NMR spectra (Fig. 4) show that
there is no
[4]
Al present, and the
11
B MAS NMR
spectra (Fig. 2) show that there is significant
[4]
B
present in these crystals. Thus the Tsite is
occupied by Si and B only, and we expect a
negative correlation between <T!O> and the
amount of
[4]
B at the Tsite as indicated from the
formula calculations (Fig. 5a). The line in Fig. 5a
was drawn through the expected distance for
complete Si occupancy of the Tsite (1.620 A
˚,
MacDonald and Hawthorne, 1995) and the grand
<
[4]
B!O> distance of 1.475 A
˚given by
Hawthorne et al. (1996) for borate minerals. The
data are in accord with this ideal line in that no
point deviates from it by >2.7 standard deviations.
The Z site
The Zsite was initially assigned as completely
occupied by Al in the refinement process, and the
equivalent isotropic-displacement parameters
(U
eq
) are in accord with equal scattering at the Z
site in all of the structures refined here (i.e.
0.077(2)!0.094(1) A
˚
2
). The <Z!O> distances
vary from 1.903(3) to 1.908(2) A
˚, and hence show
no significant variation in these structures, in
accord with the lack of variation in U
eq
. In accord
with these results, we chose not to refine the site
scattering at the Zsite in these structures as
FIG. 4.
27
Al MAS NMR spectra of mushroom tourmaline
SHM; the positions of the signals corresponding to
[4]
Al
and
[6]
Al are marked.
TABLE 7. Mo¨ssbauer parameters for Mogok
tourmaline.
d(mm/s) D(mm/s) Rel. area (%)
SHM3a
Fe
2+
1.09(2) 2.55(5) 32(2)
Fe
2+
1.10(6) 1.57(5) 10(9)
Fe
2n+
0.8(2) 1.7(5) 13(11)
Fe
2n+
0.6(4) 1.4(7) 13(13)
Fe
3+
0.4(2) 0.9(4) 32(5)
SHM3e
Fe
2+
1.08(1) 2.57(1) 58(3)
Fe
2+
1.06(2) 1.41(3) 21(3)
Fe
2n+
0.6* 1.6(3) 6(3)
Fe
3+
0.35* 0.86(8) 16(2)
SHM5
Y
Fe
2+
1.08(1) 2.57(2) 63(4)
Y
Fe
2+
1.07(3) 1.44(6) 17(5)
Y
Fe
2n
0.6* 1.6(4) 5(4)
Y
Fe
3+
0.35* 0.8(1) 15(3)
* constrained
756
A. J. LUSSIER ET AL.
variation of too much of the scattering in a
structure can lead to erroneous refined site-
scattering values. The grand <Z!O> distance is
1.905 A
˚, in accord with complete occupancy of
the Zsite by Al as assigned in Table 5.
The Y sit e
The Y-site populations were assigned from the
unit formulae given in Table 5, and the agreement
with the refined site-scattering values is shown in
Table 4. The variation in <Y!O> distances with
the aggregate radius of the constituent Y-site
cations is shown in Fig. 5b. A well-developed
linear correlation is exhibited: <Y!O> =
1.482(24) + 0.808(38) <r[Y]>, R= 0.992, standard
error of estimate = 0.0022 A
˚, supporting the
assigned site-populations and the estimation of Li
content during calculation of the unit formulae
from the electron-microprobe analyses.
The X site
The X-site populations were assigned from the
unit formulae of Table 5, and are reasonably
compatible with the refined site-scattering values
(Table 4). As is usual with elbaite, the Xsite is not
filled: in these samples, the occupancy varies
between 0.61 and 0.72 a.p.f.u.
Charge transfer (IVCT) in SHM
Intervalence charge-transfer can occur in minerals
where adjacent sites of similar stereochemistry
are occupied by transition elements of different
valence state, such that on absorption of an
incident photon, an electron moves from one atom
to another (for a more detailed explanation, see
Burns, 1981; Hawthorne, 1988). The absorption
of visible-light photons during the charge-transfer
process will often result in colouration that is
more intense than would normally be expected for
(d!d) transitions.
In tourmaline, homo-atomic and hetero-atomic
IVCT mechanisms such as Mn
2+
!Ti
4+
(brown),
Fe
2+
!Ti
4+
(yellow-green) and Fe
2+
!Fe
3+
(dark
brown, black, red) have been reported (e.g. Smith,
1978a,b; Smith and Strens, 1976; Faye et al.,
1974; Mattson and Rossman, 1984, 1987, 1988;
Taran et al., 1993; Taran and Rossman, 2002). In
contrast to other minerals, where Fe
2+
!Fe
3+
intervalence charge-transfer leads to a new
absorption band in the visible spectrum, in
tourmalines the Fe
2+
optical absorption band
may be drastically intensified by the Fe
2+
!Fe
3+
transition without the presence of any additional
bands (Smith, 1978a,b; Mattson and Rossman,
1987). It has been shown that Mo¨ssbauer spectro-
scopy is sensitive to Fe
2+
and Fe
3+
in different
short-range arrangements in tourmaline (Dyar et
al., 1998) and can identify the presence of
Fe
2+
!Fe
3+
clusters where there is electron
delocalization (Hawthorne, 1988; Ferrow, 1994;
Dyar et al., 1998; De Oliveira et al., 2002;
Eeckhout et al., 2004).
FIG. 5. (a) Variation in <T!O> distance as a function of
[4]
B in SHM and SHP mushroom tourmaline crystals;
the circles represent the amounts of
[4]
B derived from
the formula-unit calculation; the line is drawn through
the point <Si!O> = 1.620 A
˚and projects to a <B!O>
distance of 1.475 A
˚, the grand <
[4]
B!O> distance for
borate minerals given by Hawthorne et al. (1996).
(b) Variation in <Y!O> distance as a function of the
aggregate radius of the cations at the Ysite in mushroom
tourmaline crystals. The black line drawn through the
data is a regression line: <Y!O> = 1.482(24) +
0.808(38) <r[Y]> with R= 0.992 and a standard error
of estimate of 0.0022 A
˚.
CRYSTAL CHEMISTRY OF MUSHROOM ELBAITE
757
In SHM tourmaline, the black coloration of the
core region of sample SHM3 indicates the
possibility of both heteronuclear and homonuclear
intervalence charge-transfer. In SHM1 and
SHM2, Mn is the only transition-element
present, whereas SHM3 contains a similar
amount of Mn and significant Fe, and hence the
black colour of SHM3 suggests an IVCT process
involving Fe. This is confirmed by the Mo¨ ssbauer
spectra of SHM3 shown in Fig. 6. The optimal fit
was achieved using a four-site model, labelled A
to D(Table 7). Assignment follows that of Dyar
et al. (1998) who analysed nearly 50 different
tourmaline specimens by Mo¨ssbauer spectroscopy
and identified six common doublets:
Y1
Fe
2+
(d=
1.09; D= 2.47) ,
Y2
Fe
2+
(d= 1.09; D= 2.19),
Y3
Fe
3+
(d= 1.07; D= 1.60),
T
Fe
3+
(d= 0.17; D=
0.51),
Y/Z
Fe
3+
(d= 0.43; D= 0.82) and Fe
2+
!Fe
3+
(d= 0.77; D= 1.21 mm/s), where Y1, Y2, and Y3
are arbitrarily named and are interpreted as short-
range clusters.
Short-range order/disorder involving Fe
2+
and Fe
3+
In the spectra of Fig. 6, isomer shift and
quadrupole-splitting parameters coincide with
Y1
Fe
2+
,
Y3
Fe
2+
,
Y/Z
Fe
3+
,andFe
2+
!Fe
3+
.
However, on the basis of structure-refinement
data, it has been shown that no appreciable
amount of Fe
3+
is present at the Zsite (which is
occupied exclusively by Al), precluding the
possibility of any Y>Z charge transfer; thus
all Fe
2+
and Fe
3+
occurs at the Ysite (Table 5).
The relative intensities of the peaks in the
Mo¨ssbauer spectra give the amounts of isolated
Fe
2+
,isolatedFe
3+
and adjacent Fe
2+
!Fe
3+
(Table 7): e.g. Fig. 6a: 42% (0.394 a.p.f.u.),
32% (0.077 a.p.f.u.) and 26% (0.086 a.p.f.u.),
respectively. Splitting the adjacent Fe
2+
!Fe
3+
pairs into their component cations gives Fe
2+
=
0.516 and Fe
3+
= 0.423 a.p.f.u. The probability of
Fe
3+
occurring adjacent to Fe
2+
for a random
short-range distribution is 160.42/3 = 0.14. As
this cluster contains two Fe atoms, the total
amount of Fe occurring as Fe
2+
!Fe
3+
pairs for a
random short-range distribution is 0.28 a.p.f.u., in
close accord with the observed amount of Fe
involved in Fe
2+
!Fe
3+
: 0.24 a.p.f.u. Thus there is
no short-range order involving Fe
2+
and Fe
3+
.
Acknowledgements
We thank Barb Dutrow and George Harlow for
their significant comments on this paper. This
work was funded by University of Manitoba
Graduate Fellowships to AJL, SH and VKM, an
NSERC PGS-D to AJL, by a Canada Research
Chair in Crystallography and Mineralogy, and by
Major Facilities Access grants to FCH, by
Research Tools and Equipment, and Discovery
Grants to FCH and SK from the Natural Sciences
and Engineering Research Council of Canada, and
by Canada Foundation for Innovation Grants to
FCH and SK.
References
Aurisicchio, C., Ottolini, L. and Pezzotta, F. (1999)
Electron- and ion-microprobe analyses, and genetic
inferences of tourmalines of the foitite-schorl solid
FIG. 6. The Mo¨ssbauer spectra of SHM: the positions of
the material examined are shown on the sketch of the
mushroom and are identified as (a), (b) and (c).
758
A. J. LUSSIER ET AL.
solution. European Journal of Mineralogy,11,
217!225.
Bloodaxe, E.S., Hughes, J.M., Dyar, M.D., Grew, E.S.
and Guidotti, C.V. (1999) Linking structure and
chemistry in the schorl-dravite series. American
Mineralogist,84, 922!928.
Bosi, F. and Lucchesi, S. (2004) Crystal chemistry of the
schorl!dravite series. European Journal of
Mineralogy,16, 335!344.
Bosi, F., Lucchesi, S. and Reznitskii L. (2004) Crystal
chemistry of the dravite!chromdravite series.
European Journal of Mineralogy,16, 345!352.
Bosi, F., Andreozzi, G.B., Federico, M., Graziani, G.
and Lucchesi, S. (2005) Crystal chemistry of the
elbaite!schorl series. American Mineralogist,90,
1784!1792.
Bray, P.J. (1999) NMR and NQR studies of boron in
vitreous and crystalline borates. Inorganica Chimica
Acta,289, 158!173.
Burns, P.C., Macdonald, D.J. and Hawthorne, F.C.
(1994) The crystal chemistry of manganese!bearing
elbaite. The Canadian Mineralogist,32, 31!41.
Burns, R.G. (1981) Intervalence transitions in mixed-
valence minerals of iron and titanium. Annual
Review of Earth and Planetary Sciences,9,
345!383.
Ca´ mara, F., Ottolini, L. and Hawthorne, F.C. (2002)
Crystal chemistry of three tourmalines by SREF,
EMPA, and SIMS. American Mineralogist,87,
1437!1442.
De Oliveira, E.F., Castaneda, C., Eeckhout, S.G.,
Gilmar, M.M., Kwitko, RR., De Grave, E. and
Botelho, N.F. (2002) Infrared and Mo¨ ssbauer study
of Brazilian tourmalines from different geological
environments. American Mineralogist,87,
1154!1163.
Dyar, M.D., Taylor, M.E., Lutz, T.M., Francis, C.A.,
Robertson, J.D., Cross, L.M., Guidotti, C.V. and
Wise, M. (1998) Inclusive chemical characterization
of tourmaline: Mo¨ ssbauer study of Fe valence and
site occupancy. American Mineralogist,83,
848!864.
Eeckhout, S.G., Corteel, C., Van Coster, E., De Grave,
E. and De Paepe, P. (2004) Crystal-chemical
characterization of tourmalines from the English
Lake District: electron-microprobe analyses and
Mo¨ ssbauer spectroscopy. American Mineralogist,
89, 1743!1751.
Ertl, A. and Hughes, J.M. (2002) The crystal structure of
an aluminum-rich schorl overgrown by boron-rich
olenite from Koralpe, Styria, Austria. Mineralogy
and Petrology,75,1!2, 69!78.
Ertl, A., Hughes, J.M., Brandsta¨ tter, F., Dyar, M.D. and
Prasad, P.S.R. (2003a) Disordered Mg-bearing
olenite from a granitic pegmatite from Goslarn,
Austria: A chemical, structural, and infrared spectro-
scopic study. The Canadian Mineralogist,41,
1363!1370.
Ertl, A., Hughes, J.M., Prowatke, S., Rossman, G.R.,
London, D. and Fritz, E.A. (2003b) Mn-rich tourma-
line from Austria: structure, chemistry, optical
spectra, and relations to synthetic solid solutions.
American Mineralogist,88, 1369!1376.
Ertl, A., Pertlik, F., Dyar, M.D., Prowatke, S., Hughes,
J.M., Ludwig, T. and Bernhardt, H.-J. (2004) Fe-rich
Olenite with tetrahedrally coordinated Fe
3+
from
Eibenstein, Austria: Structural, chemical, and
Mo¨ ssbauer data. The Canadian Mineralogist,42,
1057!1063.
Ertl, A., Rossman, G.R., Hughes, J.M, Prowatke, S. and
Ludwig, T. (2005) Mn-bearing ‘oxy-rossmanite’
with tetrahedrally coordinated Al and B from
Austria: Structure, chemistry, and infrared and
optical spectroscopic study. American Mineralogist,
90, 481!487.
Ertl, A., Hughes, J.M., Prowatke, S., Ludwig, T.,
Brandstatter, F., Korner, W. and Dyar, M.D. (2007)
Tetrahedrally coordinated boron in Li-bearing
olenite from ‘Mushroom’ tourmaline from
Momeik, Myanmar. The Canadian Mineralogist,
45, 891!899.
Faye, G.H., Manning, P.G., Gosselin, J.R. and
Tremblay, R.J. (1974) The optical absorption spectra
of tourmaline: importance of charge-transfer pro-
cesses. The Canadian Mineralogist,12, 370!380.
Ferrow, E.A. (1994) Mo¨ ssbauer effect study of the
crystal chemistry of tourmaline. Hyperfine
Interactions,91, 689!695.
Francis, C.A., Dyar, M.D., Williams, M.L. and Hughes,
J.M. (1999) The occurrence and crystal structure of
foitite from a tungsten-bearing vein at Copper
Mountain, Taos County, New Mexico. The
Canadian Mineralogist,37, 1431!1438.
Grice, J.D. and Ercit, T.S. (1993) Ordering of Fe and Mg
in the tourmaline crystal structure: the correct
formula. Neues Jahrbuch fu
¨rMineralogie
Abhandlungen,165, 245!266.
Grice, J.D., Ercit, T.S. and Hawthorne, F.C. (1993)
Povondraite, a redefinition of the tourmaline
ferridravite. American Mineralogist,78, 433!436.
Hawthorne, F.C. (1988) Mo¨ ssbauer spectroscopy. Pp.
255!340 in: Spectroscopic Methods in Mineralogy
and Geology (F.C. Hawthorne, editor). Reviews in
Mineralogy, 18, Mineralogical Society of America,
Washington, D.C.
Hawthorne, F.C. (1996) Structural mechanisms for light-
element variations in tourmaline. The Canadian
Mineralogist,34, 123!132.
Hawthorne, F.C. (2002) Bond-valence constraints on the
chemical composition of tourmaline. The Canadian
Mineralogist,40, 789!798.
Hawthorne, F.C. and Henry, D.J. (1999) Classification
CRYSTAL CHEMISTRY OF MUSHROOM ELBAITE
759
of the minerals of the tourmaline group. European
Journal of Mineralogy,11, 201!215.
Hawthorne, F.C., Burns, P.C. and Grice, J.D. (1996) The
crystal chemistry of boron. Pp. 41!115 in: Boron:
Mineralogy, Petrology and Geochemistry (E.E.
Grew and L.M. Anovitz, editors). Reviews in
Mineralogy, 33, Mineralogical Society of America,
Washington, D.C.
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.
Henry, D.J. and Dutrow, B.L. (1992) Tourmaline in low-
grade clastic sedimentary rocks: an example of the
petrogenetic potential of tourmaline. Contributions
to Mineralogy and Petrology,112, 203!218.
Henry, D.J. and Dutrow, B.L. (1996) Metamorphic
tourmaline and its petrogenetic applications. Pp.
503!557 in: Boron: Mineralogy, Petrology and
Geochemistry (E.E. Grew and L.M. Anovitz,
editors). Reviews in Mineralogy, 33, Mineralogical
Society of America, Washington, D.C.
Henry, D.J. and Guidotti, C.V. (1985) Tourmaline as a
petrogenetic indicator mineral: an example from the
staurolite-grade metapelites of NW Maine. American
Mineralogist,70,1!15.
Hia, K., Themelis, T. and Kyaw, T. (2005) The
pegmatitic deposits of Molo (Momeik) and Sakan-
gyi (Mogok). Australian Gemologist,22, 303!309.
Hughes, J.M., Ertl, A., Dyar, M.D., Grew, E.S., Shearer,
C.K., Yates, M.G. and Guidotti, C.V. (2000)
Tetrahedrally coordinated boron in a tourmaline:
boron-rich olenite from Stoffhu
¨tte, Koralpe, Austria.
The Canadian Mineralogist,38, 861!868.
Hughes, J.M., Ertl, A., Dyar, M.D., Grew, E.S.,
Wiedenbeck, M. and Brandsta¨ tter F. (2004)
Structural and chemical response to varying
[4]
B
content in zoned Fe-bearing olenite from Koralpe,
Austria. American Mineralogist,89, 447!454.
Kalt, A., Schreyer, W., Ludwig, T., Prowatke, S.,
Bernhardt, H.-J. and Ertl, A. (2001) Complete solid
solution between magnesian schorl and lithian
excess-boron olenite in a pegmatite from the
Koralpe (eastern Alps, Austria). European Journal
of Mineralogy,13, 1191!1205.
Keller, P.C. (1983) The rubies of Burma: A review of
the Mogok stone tract. Gems & Gemology,6,
209!219.
Kriz, H.M. and Bray, P.J. (1971) The
11
B Quadrupole
Interaction and non-bridging oxygens in crystalline
borates. Journal of Magnetic Resonance,4, 76!84.
Kroeker, S. and Stebbins, J.F. (2001) Three-coordinated
boron-11 chemical shifts in borates. Inorganic
Chemistry,40, 6239!6246.
Lussier, A.J., Hawthorne, F.C., Aguiar, P., Michaelis, V.
and Kroeker, S. (2008) The occurrence of tetra-
hedrally coordinated Al and B in tourmaline: A
11
B
and
27
Al MAS NMR study. American Mineralogist
(in press).
Macdonald, D.J. and Hawthorne, F.C. (1995) The
crystal chemistry of Si-Al substitution in tourmaline.
The Canadian Mineralogist,33, 849!858.
Marler, B. and Ertl, A. (2002) Nuclear magnetic
resonance and infrared spectroscopic study of
excess-boron olenite from Koralpe, Styria, Austria.
American Mineralogist,87, 364!367.
Marschall, H.R., Ertl, A., Hughes, J.M. and
McCammon, C. (2004) Metamorphic Na- and OH-
rich disordered dravite with tetrahedral boron
associated with omphacite, from Syros, Greece:
chemistry and structure. European Journal of
Mineralogy,16, 817!823.
Massiot, D., Bessada, C., Coutures, J. P. and Taulelle, F.
(1990) A quantitative study of
27
Al MAS NMR in
crystalline YAG. Journal of Magnetic Resonance,
90, 231!242.
Mattson, S.M. and Rossman, G.R. (1984) Ferric iron in
tourmaline. Physics and Chemistry of Minerals,11,
225!234.
Mattson, S.M. and Rossman, G.R. (1987) Fe
2+
-Fe
3+
interactions in tourmaline. Physics and Chemistry of
Minerals,14, 163!171.
Mattson, S.M. and Rossman, G.R. (1988) Fe
2+
-Ti
4+
charge transfer in stoichiometric Fe
2+
,Ti
4+
-minerals.
Physics and Chemistry of Minerals,16, 78!82.
Michaelis, V.K., Aguiar, P.M. and Kroeker, S. (2007)
Probing alkali coordination environments in alkali
borate glasses by multinuclear magnetic resonance.
Journal of Non-Crystalline Solids,353, 2582!2590.
Neiva, A.M.R., Manuela, M., Silva, V.G. and Gomes,
M.E. (2007) Crystal chemistry of tourmaline from
Variscan granites, associated tin-tungsten- and gold
deposits, and associated metamorphic and metaso-
matic rocks from northern Portugal. Neues Jahrbuch
fu
¨r Mineralogie Abhandlungen,184, 45!76.
Nova´ k, M. and Povondra, P. (1995) Elbaite pegmatites
in the Moldanubicum: a new subtype of the rare-
element class. Mineralogy and Petrology,55,
159!176.
Nova´ k, M., Selway, J., C
ˇerny
´, P., Hawthorne, F.C. and
Ottolini, L. (1999) Tourmaline of the elbaite-dravite
series from an elbaite-subtype pegmatite at Bliz
ˇna´,
southern Bohemia, Czech Republic. European
Journal of Mineralogy,11, 557!568.
Povondra, P. and Novak, M. (1986) Tourmalines in
metamorphosed carbonate rocks from western
Moravia, Czechoslovakia. Neues Jahrbuch fu
¨r
Mineralogie Monatshefte,1986, 273!282.
Schreyer, W., Wodara, U., Marler, B., Van Aken, P.A.,
Seifert, F. and Robert, J.-L. (2002) Synthetic
tourmaline (olenite) with excess boron replacing
silicon in the tetrahedral site: I. Synthesis conditions,
760
A. J. LUSSIER ET AL.
chemical and spectroscopic evidence. European
Journal of Mineralogy,12, 529!541.
Selway, J., C
ˇerny
´, P. and Hawthorne, F.C. (1998)
Feruvite from lepidolite pegmatites at Red Cross
Lake, Manitoba. The Canadian Mineralogist,36,
433!439.
Selway, J.B., Nova` k, M. C
ˇerny
´, P. and Hawthorne, F.C.
(1999) Compositional evolution of tourmaline in
lepidolite-subtype pegmatites. European Journal of
Mineralogy,11, 569!584.
Selway, J.B., Nova´ k, M., C
ˇerny
´,, P. and Hawthorne,
F.C. (2000a) The Tanco pegmatite at Bernic Lake,
Manitoba. XIII. Exocontact tourmaline. The
Canadian Mineralogist,38, 869!976.
Selway, J.B., C
ˇerny
´, P., Hawthorne, F.C. and Nova´k, M.
(2000b) The Tanco pegmatite at Bernic Lake,
Manitoba. XIV. Internal tourmaline. The Canadian
Mineralogist,38, 877!891.
Selway, J.B., Smeds, S-A., C
ˇerny
´, P. and Hawthorne,
F.C. (2002) Compositional evolution of tourmaline
in the petalite-subtype Nyko¨ pingsgruvan pegmatites,
Uto¨ , Stockholm Archipelago, Sweden. GFF,124,
93!102.
Sheldrick, G.M. (1998) SADABS User Guide, University
of Go¨ ttingen, Germany.
Skibsted, J., Nielsen, N. Chr., Bildsøe, H. and Jakobsen,
H.J. (1992)
51
V MAS NMR spectroscopy: determi-
nation of quadrupole and anisotropic shielding
tensors, including the relative orientation of their
principal-axis systems. Chemical Physics Letters,
188(5!6), 405!412.
Skibsted, J., Nielsen, N. Chr., Bildsøe, H. and Jakobsen,
H.J. (1993) Magnitudes and relative orientation of
51
V quadrupole coupling and anisotropic shielding
tensors in metavanadates and KV
3
O
8
from
51
V MAS
NMR spectra.
23
Na quadrupole coupling parameters
for a- and b-NaVO
3
.Journal of the American
Chemical Society,115, 7351!7362.
Smith, G. (1978a) A reassessment of the role of iron in
the 5000!30,000 cm
-1
range of the electronic
absorption spectra of tourmaline. Physics and
Chemistry of Minerals,3343!373.
Smith, G. (1978b) Evidence for absorption by exchange-
coupled Fe
2+
-Fe
3+
pairs in the near infrared spectra
of minerals. Physics and Chemistry of Minerals,3,
375!383.
Smith, G. and Strens, R.G. (1976) Intervalence transfer
absorption in some silicate, oxides, and phosphate
minerals. Pp. 583!612 in: The Physics and
Chemistry of Minerals and Rocks (R.G.J. Strens,
editor). Wiley, New York.
Tagg, S.L., Cho, H., Dyar, M.D. and Grew, E.S. (1999)
Tetrahedral boron in naturally occurring tourmaline.
American Mineralogist,84, 1451!1455.
Taran, M.N. and Rossman, G.R. (2002) High-temp-
erature, high-pressure optical spectroscopy study of
ferric-iron-bearing tourmaline. American
Mineralogist,87, 1148!1153.
Taran, M.N., Lebedev, A.S. and Platonov, A.N. (1993)
Optical absorption spectroscopy of synthetic tour-
malines. Physics and Chemistry of Minerals,20,
209!220.
Taylor, M.C., Cooper, M.A. and Hawthorne, F.C. (1995)
Local charge-compensation in hydroxy-deficient
uvite. The Canadian Mineralogist,33, 1215!1221.
Themelis, T. (2007) Gems and Mines of Mogok: The
Forbidden Land. A & T Press, Bangkok, Thailand.
Zaw, K. (1998) Geological evolution of selected granitic
pegmatites in Myanmar (Burma): constraints from
regional setting, lithology, and fluid-inclusion
studies. International Geology Review,40,
647!662.
CRYSTAL CHEMISTRY OF MUSHROOM ELBAITE
761