MOSSBAUER SPECTROSCOPY AND CRYSTAL CHEMISTRY
OF AENIGMATITES
by
JIN
CHOI
BEOM
S.B.
Seoul National University (1979)
M.S.
Seoul National
Submitted
to
Un i v e r si t y (1981)
the
Department
of
Earth and Planetary Sciences
in
Partial
Fulfillment
Requirements
of
the
of the
Degree of
Master
of Science
at
the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
June,
0
1983
Jin Beom Choi
1983
The author hereby grants to M.I.T. permission to reproduce and to
distribute
copies of this thesis document
in whole or in part.
Signature of Author
Department of Earth and Planetary Sciences,
Certified
March,
1983
by
Thesis Advisor
Accepted
by
..Chairman,
Department Committee
IA$SACIiHSETSINSTiTUTE
OFTECHOLOGY
JUN 1 5 1983 Archives
LIBRARIES
2
ABSTRACT
VIII
VI
Na4
(Fe 2 +,Ti,Fe 3 +) 1 2 (SiFe
Aenigmatite,
common constituent
alkaline
of sodium-rich
3 +)
igneous
IV
12 0 40 ,
is a
rocks and is
classified as an open-branched single-chain silicate, whose
related to that of sapphirine.
is closely
structure
The first
Mossbauer spectra of three valid aenigmatite specimens were
recorded
and the detailed
the extreme
spectra,
the Mossbauer
crystal chemistry
were obtained.
In
overlap of peaks for Fe 2 + and
Fe3 + cations in the low velocity region made fitting difficult
and even made it impossible
to resolve
he liquid N 2 spectra.
Moreover, the Mossbauer spectra are complicated by the possible
occurrence
of iron cations in seven octahedral
tetrahedral sites.
were grouped
and four
However, the Fe2 +-bearing octahedral sites
into three
categories,
resulting
in three
quadrupole doublets. The Fe3 + cations in octahedral and
tetrahedral
result,
sites were accounted
the stringent
led to the resolution
three doublets
for by three peaks.
fitting of the room temperature
of nine peaks.
As a
spectra
Such peaks consist of
of Fe 2 +/oct and one combined
peak at low velocity
corresponding to two small peaks at high velocity which were
assigned
to Fe3+/tet
and Fe 3 +/oct.
Using the peak areas for
Fe 2 + and Fe3 + peaks, analytical data were reevaluated, including
the comparison of Fe 3 +/Fe2 + ratios determined by wet chemistry
and by Mossbauer spectroscopy.
As a result, electron microprobe
3
analyses
were recalculated
for FeO and Fe 2 0 3 proportions
crystal chemistry of aenigmatite was examined.
and the
Despite
experimental errors involved, the existence of significant
amounts of Fe3 + in tetrahedral coordination indicates that Fe3 +
has a preference
over A1 3 + for the tetrahedral
Thesis
Supervisor:
Title:
Professor
Roger
G. Burns
of Mineralogy
and Geochemistry
sites.
_
i
4
ACKNOWLEDGMENTS
The author wishes to express enormous thanks to his
academic and thesis advisor, Professor Roger G. Burns who
introduced him to the fantastic world of Mossbauer spectroscopy,
encouraged him, gave valuable advice, and partly supported him
throughout the research.
real scientist
treated
Besides, Dr. Burns showed what the
is and his enthusiasm
for research.
the author, who has been suffered
He always
from the unfamiliar
American life, especially communications, kindly, friendly,
generously,
and much more.
The author is personally
their
support during
indebted
his studies at M.I.T.,
to his parents
for
and to his wife,
Hye-Kyung who made it possible for him to complete this thesis
successfully.
This thesis is dedicated
author's
mind
as
an example
best friends forever.
of
to Roger who will stay in the
a scientist
as well
as
one
of
his
Z
5
TABLE OF CONTENTS
Page
ABSTRACT
2
ACKNOWLEDGMENTS
4
TABLE OF CONTENTS
5
LIST OF FIGURES AND TABLES
6
I.
8
INTRODUCTION
1.
What
2.
Crystal structure
3.
Parageneses
4.
Purpose and Method
II.
8
15
of Study
16
EXPERIMENTAL PROCEDURES
19
1. Aenigmatite Specimens
19
2.
Mossbauer Spectra
20
3.
Fitting Procedures
25
III.
IV.
is Aenigmatite?
RESULTS
43
1.
Mossbauer Parameters
43
2.
Crystal Chemistry of Aenigmatites
47
CONCLUSIONS
BIBLIOGRAPHY
53
56
6
LIST OF FIGURES AND TABLES
Page
LIST OF FIGURES
Figure
1.
Idealized polyhedral diagrams of
12
aenigmatite.
Figure
2.
Unfit Mossbauer spectra of aenigmatite
23
AEN1.
Figure
3.
The spectrum of AEN2 at room temperature
fitted with
Figure
4.
29
seven peaks.
The spectrum
of AEN2 at room temperature
31
fitted with nine peaks.
Figure
5.
Peak positions for Mossbauer spectra
35
at room temperature.
Figure
6.
The fitted
spectrum
of AEN1 at room
37
of AEN3 at room
39
temperature.
Figure
7.
The fitted
spectrum
temperature.
LIST OF TABLES
Table
1.
Unit cell parameters
sapphirine.
of aenigmatite
and
10
7
Page
Table
2.
Localities of aenigmatite specimens.
20
Table
3.
Uncorrected chemical analyses of
21
aenigmatite specimens.
Table
4.
Mossbauer parameters from two different
28
fitting schemes for AEN2.
Table
5.
Peak positions for Mossbauer spectra
of aenigmatite
specimens
34
at room
temperature.
Table
6.
Mossbauer parameters of AEN1 and AEN3
41
at room temperature.
Table
7.
Mossbauer parameters of synthetic
45
sapphirine and natural yellow sapphirine.
Table
8.
Ferric and ferrous chemical data of
49
aenigmatite specimens obatained from
the Mossbauer
Table
9.
analyses
and wet analyses.
Cation distribution in aenigmatite
specimens.
51
8
I.
1.
What
INTRODUCTION
Is Aenigmatite?
Aenigmatite,
discovered
by Breighaupt
in 1865, is
commonly found in sodium-rich alkaline igneous rocks of both
volcanic
and plutonic
parageneses
(Merlino,
1970).
It is
classified as an open-branched single-chain silicate (Liebau,
1980).
Because of its widely scattered occurrences, aenigmatite
was described by numerous mineralogists after its discovery.
However, uncertainties about its crystal chemistry and structure
led to considerable
the intensive
confusion
study by Kelsey
in the earlier
and McKie
literature,
until
(1964) clarified
structural and chemical properties of aenigmatite.
Although
aenigmas
the origin of its name is not clear,
of the mineral
aenigmatite
one of the
is that it has been called
many different names as a result of earlier uncertainty
regarding its composition.
Cossyrite, derived from the
pantellerite lavas of Pantelleria the ancient name for which was
Cossyra
(Fostner,
1881), was considered
but now the name is rarely used.
pseudomorph
of aenigmatite
to be the same mineral,
Kbingite,
by its discoverer
considered
as a
(Breithaupt,
1865),
was discredited as an arfvedsonite-aenigmatite intergrowth
(Brogger,
exists:
1890).
This confusion
over name and identity
two specimens made available
still
to us from the Harvard
collection (85123 & 85123A) turned out to be an arfvedsonite,
9
and the mineral
called aenigmatite
and Seifer, in press) appears
in a recent paper
(Steffen
to be riebeckite-arfvedsonite
rather than aenigmatite.
2. Crystal Structure
The crystal morphology
of aenigmatite,
studied
imperfectly by early workers, was determined to be triclinic
with
unit-cell
constants:
a = 96°59.51,
first detailed
a
:
= 96°49.5',
X-ray
study was done
(AEN1, Cambridge 38075)*.
1.
:
They reported
c
=
1.0050
:
112028? (Palache,
using a crystal of aenigmatite
Table
b
:
1933).
by Kelsey and McKie
from the Kola peninsula,
P.5862,
The
(1964)
U.S.S.R.
Unit-cell parameters are shown in
chemical
analyses
of twelve specimens
from eight different localities and characterized the chemical
2+
unit of aenigmatite
as the idealized
formula
Na 4 FeO
1 0 Ti 2 Si 1 20 4 0 .
They suggested the crystal structure of aenigmatite contains
silicate
chains of pyroxene
on distinct
octahedral
coordination number.
unsolved
type cross-linked
sites with Na on sites of higher
However, the complete structure remained
until two independent
were reported
by Ti 4+ and Fe 2 +
by Merlino
and simultaneous
(1970) and Cannillo,
determinations
et al. (1971) who
collected X-ray data from aenigmatite crystals from Naujakasik,
Greenland
(AEN3, Harvard
85124)*.
* Numbers refer to specimens described
in Table 2.
I
10
the crystal
structure
is closely
1973, 1980).
to that
related
of aenigmatite
structure
from the structure of sapphirine
could be deduced
Merlino,
of aenigmatite
In fact, the crystal
of sapphirine.
et al. (1971),
(1970) and Cannillo,
to Merlino
According
In spite of their different
(Moore, 1969;
chemical
compositions, there are significant crystallochemical
Table
1,
unit-cell
those of aenigmatite.
described
Table
Moreover,
in terms of a multiple
cell obtained
1.
by applying
are analogous
of sapphirine
parameters
As shown in
and sapphirine.
between aenigmatite
similarities
Unit cell parameters
cell can be
the triclinic
pseudomonoclinic
(four-fold)
the transforming
to
matrix
of
and sapphirine
of aenigmatite
Sapphirine
Aenigmatite
Cannillo et
Kelsey and
al. (1971)
McKie
pseudomonoclinic
triclinic
Merlino
(1973)
(1964)
pseudomonoclinic
triclinic
a A
12. 120
10.406
11.33
10.04
b A
29.63
10.813
29.08
10.38
c A
10.406
8.926
10.04
8.65
a
90°004'
104056'
aS
y
Z
96052'
12709 '
89044
8
'
125019?
2
90°
125023 '
900
8
107033'
95007'
123055
2
'
11
[100/T22/100]
VIII
VI
(Fe2 +,Ti,Fe 3 +) 6
Cannillo,
cell.
O0[(Si,Fe3+)
IV
6
(Mg,Fe,Al)0
6
(g,Fe,Al)
octahedra
along
8
occur
structure
along the z axis parallel
by chains of tetrahedra
(001) as well as by additional
however,
18], "walls" of
from each other in the y direction.
are interconnected
In aenigmatite,
is
IV
2 [(Si,Al) 6
run infinitely
to (100) and are separated
The walls
(1971)
018].
VI
In sapphirine,
et al.
formula of aenigmatite
that the ideal chemical
suggested
Na2
to the triclinic
A
octahedra
some modifications
because some octahedra
running
(Moore, 1969).
to the sapphirine
are replaced
by
Na-polyhedra of higher coordination, that is, distorted square
antiprisms.
The octahedral
walls,
continuous
(100) layers formed
antiprisms
(Fig. la) instead
by both (Fe,Ti) octahedra
of bands
chains are, as in sapphirine,
are build up of
therefore,
in sapphirine.
single
pyroxene-like
and Na-
Tetrahedra
chains, with
additional "wings" of corner-sharing tetrahedra, though the
chains are somewhat
sapphirine
(Fig.
less kinked
lb).
packed array, whereas
Oxygens
those
in aenigmatite
in sapphirine
in aenigmatite
than in
are in cubic close-
are too puckered
to
approximate close-packing.
In aenigmatite,
independent
groups:
there are six crystallographically
tetrahedral
)tetrahedra
sites which
can be divided
into two
[T(1) and T(4)] having one non-bridging
oxygen, and 2) tetrahedra
[T(2), T(3),
two non-bridging
however,
oxygens;
T(5), and T(6)] having
the difference
of Si-O
12
Fig.
1.
Idealized
aenigmatite:
of
lying in the (100) plane,
to z axis (pseudomonoclinic
b) layer of tetrahedral
connected
diagrams
a) layer of (Fe,Ti) octahedra
and Na-antiprisms
parallel
polyhedral
cell),
[Si 6 0 1 8 ] chains
by single octahedra.
shown in
two different layers represent same positions.
From Cannillo,
et al. (1971)
I
I
__
13
E
t%
U.0
Il
14
distances
range
between the two groups
of distances
It is difficult
tetrahedral
1.59
to
to determine
1.69
where
A (Cannillo,
et al.,
was assumed
to preferentially
occupy
1971).
in the six
ferric ions occur
sites but the small amount of Fe 3 + available
distribution
(Cannillo,
is
and the
is not significant
for
the T(3) site
et al., 1971).
Seven different octahedral sites exist in aenigmatite.
All have almost
the same range of metal-oxygen
to 2.17 A, except
the octahedron
to be preferentially
distances
to be occupied
The other
M1(7) which
2.10
is believed
by Ti 4 + and has a range of
occupied
of 1.84 to 2.09 A (average
also believed
cations.
around
distances,
1.98 A).
This M(7) site is
by some Fe 3 + and perhaps
six M sites are occupied
mainly
Fe 2 +
by Fe 2+
cations and minor amounts of Ca2 +, Mn2 +, Mg2 +, and Fe 3 + cations.
The M sites except M(7) can be divided
groups
with
in terms of types of bridging
two non-bridging
oxygens,
into three different
oxygens.
They are 1) M(5)
2) M(3), M(4),
and M(6) having
one non-bridging oxygen and "neutral" electrostatic balance, and
3) M(1) and M(2) having
similar non-bridging
oxygen
second group, but with "underbonded" oxygens.
basis for the fitting of the Mossbauer
described
spectra
as in the
This forms the
of aenigmatite
in Chap. II.
Minerals being isostructural with aenigmatite include
3+
rh6nite, Ca 4 (Mg,Fe2 +)8 Fe 2 Ti 2 (A16 Si6 0 40 ), which has the same
sapphirine-like monoclinic pseudo-cell (Cameron, et al., 1970);
15
krinovite Na2 Mg 2 CrSi3 01 0
Ca2 (MgAl)
(1972)
(Merlino, 1972); and serendibite
6 02i(Si,Al,B)6 01 8]
suggested
1974).
(Machin & Ssse,
that Cr in the krinovite
M(1), M(2) and M(8) sites, and Mg occurs
structure
Merlino
occupies
the
in the remaining
octahedra.
3.
Parageneses
Aenigmatite is a common consituent of sodium-rich
alkaline
igneous
rocks.
In volcanic
parageneses,
such as the
pantellerites, pantelleritic trachytes, and comendites of
Pantelleria
and Kenya Rift Velley,
the alkali
lavas of Oki
Island, Japan, and the c3mendites
of Mt. Nimrud, Armenia,
aenigmatite
of the groundmass,
occurs as a component
frequently as a phenocryst.
etc.,
and less
Chemical analyses indicate that
aenigmatite-bearing rocks are characterized by rather high
titanium contents and relatively low iron oxidation ratio
(Kelsey and McKie,
1964).
Aenigmatite is commonly associated with anorthoclase,
aegirine-hedenbergite
titanomagnetite
Australia.
and quartz
in peralkaline
in Pantelleria
trachytes
lavas and
in New South Wales,
Zoned aenigmatite with cores of titanomagnetite
indicates titanomagnetite reacts with sodium silicate liquid,
and then aenigmatite
crystallizes
at low oxygen
reaction may be represented as follows:
fugacity.
The
16
6 (TiFeO3 'Fe 3 04 ) + 12 SiO2 + 12 (NaFe3 +Si 2 0 6 + CaFe2+Si206)
titanomagnetite
aegirine - hedenbergite
2+
++
6 (Na 2Fe5 TiSi6 020 ) + 12 CaFe 2 +Si206 + 2 Fe3 04 + 5 02 +
aenigmatite
hedenbergite
magnetite
This kind of reaction may explain why analytical data of most
aenigmatites
show low Fe 3 +/Fe 2 + ratios.
The temperature
and
oxygen fugacity operating during the formation of the
aenigmatite
is estimated
10-13.7 atm, respectively
to
be
900
±+ 50
C and
(Deer, et al.,
10 - 1260
to
1978).
Another paragenesis of aenigmatite is plutonic and
includes nepheline and sodalite-syenites of Kangerdluarsuk, west
Greenland; a micro-syenite vein from South Boswell Bay, east
Greenland
pegmatites
(AEN2, Deer 3584)*; the foyaite
in Kola peninsula
and khibinite
(AEN1, Cambridge
38075)*;
monzonites and syenites from the Morotu River, Japan; and
aegirine-syenites
in Madagascar,
etc.
aenigmatite is commonly associated with
and hastingsite.
In these occurrences,
aegirine, riebeckite
Aenigmatite is sometimes replaced by albite
and riebeckite during later periods of albitization.
Astro-
phyllite and, less commonly, biotite also replace aenigmatite.
Aenigmatite also occurs in more silicic plutonic rocks,
such as nordmarkite
of Maine
and granite-complexes
of Nigeria,
where aenigmatite is usually associated with aegirine and
arfvedsonite-riebeckite.
* Numbers refer to specimens
described
in Table 2.
17
4.
Purpose
and Method
of Study
Although aenigmatite is easily recognized in common
igneous rocks and has received some attention from
mineralogists, its detailed crystal chemistry is poorly
understood
(Steffen
& Seifer, in press).
occupancies
in a recent
and was even misinterpreted
study
of cation site
The determination
is not easy because there are several different
crystallographic
positions in the aenigmatite structure.
Moreover, iron ordering over tetrahedral and/or octahedral sites
is also controvertial.
In this work,
distribution
the author
set out to determine
of Fe 2 + and Fe 3 + over different
the
cation sites by
In
measuring the Mossbauer spectra of several aenigmatites.
order
co accomplish
this goal,
certain
conditions
had to be met.
First, proper specimens of natural aenigmatite were carefully
collected.
In this step, all specimens
X-ray diffraction
as aenigmatite.
because
were identified
some amphiboles
Second, Mossbauer
by the
could be misidentified
measurements
were made on
specimens having different compositions at both room temperature
and liquid
N 2 temperature
(80°K).
Consistencies
in peak
positions as well as parameters among different specimens were
established to demonstrate the accuracy of fitting which should
be important for first Mossbauer works about aenigmatite.
Third,
distribution
of iron cations
over
he octahedral
and
tetrahderal sites were determined from the Mossbauer parameters,
18
and the crystal chemistry of aenigmatite was finally obtained.
X-ray diffraction analyses were made by a DIANO digital
model X-ray diffractometer using Fe-filtered Co radiation
(30KV/15mA), available in the X-ray Diffraction Laboratory in
the Department of Material Sciences, MI.T.
Mossbauer spectra
were recorded on a constant acceleration ASA (Austin Science
Associates)
Mossbauer
spectrometer,
using 512 channels
channel Nuclear Data multichannel analyser.
of a 1024
The spectra were
acquired from two different ends of a vibrator designed for dual
purpose;
one
end
of
the
vibrator
having
a
57
Co
source
in a
rhodium matrix (90-100 milicuries) was used for liquid nitrogen
spectra,
and the other end having
a
5 7 Co
source in a palladium
matrix (40-50 milicuries) was used for room temperature
measurements.
discussed
Sample preparation and fitting procedures will be
in Chap.
II.
19
II.
1.
EXPERIMENTAL PROCEDURES
Aenigmatite Specimens
Eight aenigmatite specimens with different parageneses
were collected
Cambridge
for this study.
University
Three of the samples came from
and were the very specimens
study by Kelsey and McKie (1964).
The remaining
used in the
aenigmatites
were obtained from the mineralogical collection at Harvard
University.
However,
two of them (Harvard
85123 & 85123A) were
disquailified as being arfvedsonites and were discarded from
this study.
After a preliminary examination of the Mossbauer
spectra of all the aenigmatites, three typical specimens were
selected for detailed Mossbauer experiments and computer
fitting.
Localities of specimens chosen in this study are
listed in Table 2 and chemical
analyses
are shown in Table 3.
20
Table 2.
Localities of aenigmatite specimens.
aenigmatite from Khibinite quarry, east of Kirovsk,
AEN1
Kola, USSR.
Univ.,
Obtained in powder form from Cambridge
#38075.
aenigmatite from South Boswell Bay, Kangerdlugsuak,
AEN2
East Greenland.
Cambridge
Univ.,
Obtained in powder form from
Specimen
W. A. Deer #3584.
aenigmatite from Naujakasik, Greenland.
AEN3
in the form of a hand specimen
Obtained
from Harvard
University #85124.
2. Mossbauer Spectra
Sufficient aenigmatite samples were weighed out so as
to give a total iron concentration
A recent
study (M.D. Dyar,
the optimum
bearing
pers. comm.,
is 5-10 mgFe/cm
in the Mossbauer
2
7.5 mg/cm 2 .
1982) confirmed
range of total iron concentrations
silicates
absorbance
of approximately
that
for most Fe-
to get a relatively
good
spectra and best statistics.
The
weighed sample was mixed with sugar, ground under acetone
(which helps to prevent ferrous iron from oxidation), mounted
in a round hole of 2.2 cm diameter
plastic
square plate,
and shielded
sealed by cellotape
in a
by a lead square plate with
21
Table 3.
Uncorrected chemical analyses of aenigmatite
specimens.
a
b
c
AEN1
AEN2
AEN3
SiO2
39.62
41.41
40. 24
TiO 2
9.66
8.30
7.52
A1 2 03
0.64
nil
1.31
Fe 203
4.64
4.46
FeO
33.92
35.87
Mn 0
2.46
1.78
0.83
Mg O
1.65
1.35
0.01
CaO
0.44
nil
0.13
*
0.17
}
*
ZnO
41. 13**
Na 20
7.20
6.87
7.75
K 20
0.04
0.04
0.06
H2 0+
0.05
nil
H20-
nil
n.d.
C1
0.02
n.d.
F
nil
n.d.
Z
100.34
100.08
*9.1
*
*
*
99.15
a: wet analysis by J.H. Scoon (S.O. Agrell, pers.
1982)
b: wet analysis by P.E. Brown & Mrs. Chadwick
(S.O. Agrell, pers. comm., 1982)
c: microprobe analysis by D.A. Nolet (R.G. Burns,
pers. comm., 1982)
*: element not analyzed
**:
determined as FeO
comm.,
22
hole.
the same diameter
disc
containing
suspended
runs,
For low temperature
the sample was attached
in a reservoir
to a copper
of liquid nitrogen
the velocity
range of
rod
(80°K).
spectra were run for 1-2 days to acquire more than
per channel within
the plastic
The
106 counts
5 mm/sec.
The Mossbauer spectra were calibrated relative to Fe
foil and transferred
the plotting
executed.
to a MINC PDP 11/23 minicomputer,
and the calculations
for curve-fitting
where
were
Examples of unfitted Mossbauer spectra of specimen
AEN1 at both 300°K (room temperature)
are
temperature)
shown
As illustrated
in Fig.
and at 80°K (liquid N2
2a and
2b.
in Fig. 2, the spectra show two
characteristic regions dominated by absorption by Fe2 +
components
at approximately
spectrum;
temperature
0.3 mm/sec
zero and 2.3 mm/sec
the high velocity
in the liquid N2
two peaks, the low velocity
the high velocity
low velocity
spectrum.
peak increases
Comparing
region has higher
one, resulting
Fe 3 + components.
in the room
by about
the area of
absorption
than
from overlap of Fe 2 + and the
The very small, broad peak at
approximately 0.9 mm/sec represents absorption by the high
velocity
peak of Fe 3 + components;
the peak is better resolved
N2 temperature
The strong overlap
in the liquid
spectrum.
Fe 2 + and Fe 3 + peaks in the low velocity
region
of
and the small
intensities of the Fe3 + components require very stringent
constraints in the fitting procedures.
Mossbauer spectra of
23
Fig. 2.
Unfit Mossbauer
aenigmatite AEN1.
spectra of
A) 300°K spectrum
(room temperature); B) 80°K spectrum
(liquid N 2 ).
The spectra
show two
regions dominated by absorption by
Fe 2 + ions at approximately
mm/sec.
zero and 2.5
The small peak at approximately
0.9 mm/sec originates from Fe3 + ions.
24
A) 300°K
lUl I
I#I-j'FnnTT
y7P'P'r
SI
-pi r
99 -
I
~~~~~~~~~~~~~~~~I
I
I
I
I
I
I
I
I
B) 80 °K
I
I
I
I
zZ
II
Ii
I
I
I
I
I
I
w,Ll 100 -
II
I
II
I
I
0
I
I
I
3;7-
I
t
I
I
I
I
I
I
i !I
co
0CO
I
I~~~~~~~~~~~
I
I
CO
!
Iro
I
Y
co
I
I
1!
II
1
I
II
I II
t
1
I
i
I
I
I
I
I
ti
1
I
!I
II
1
I1
I1
I1
I
I1
I I
i !!
!
94-
92
-s
i
I
_
-4
I!
-. 5
I-
I..
,
-z
I-
a
MM/SEC
I
I
z
' Il
3
'
i
9
s!
25
other
Fig.
3.
aenigmatie
specimens
are similar to that of AEN1
shown in
2.
Fitting Procedures
The Mossbauer spectra were fitted with a least-squares
program,
developed
by A. J. Stone
modified
by Huggins
computer
by Dr. K. M. Parkin
(Stone, et al.,
(1974), and adapted
fits a sum of Lorenzian
1969),
for the PDP11/23
and M. D. Dyar.
Stone's program
curves using the Gauss non-linear
regression method with appropriated constraints on the various
input parameters.
Prior to attempts
was necessary
to assess
to fit the aenigmatite
the number
Fe 2 + and Fe 3 + in the structure
discussed
in Chap.
sites, including
II, there
the M(7)
and type of cation
of aenigmatite.
site assigned
of peaks
sites is impossible,
to individual
assumption
was made that grouping
categories
could be made:
sites for
octahedral
to Ti 4+, and four
However, the
iron cations
so that it was necessary
sites into distinguishable groups.
it
As already
are seven different
different tetrahedral sites in aenigmatite.
resolution
spectra,
in each of these
to group
the
As noted earlier, the
of six M sites into three
[M(1) & M(2)],
M(5), and
[M(3), M(4)
& M(6)] leading to the resolution of three Fe 2 + doublets.
Although
ferric iron is assumed
site preferentially
(Cannillo,
to occupy
the tetrahedral
et al., 1971), the breadth
T(3)
of
26
the inner
peak suggests
of two Fe 3 + doublets.
the existence
While one doublet may originate from Fe 3 + cations in the T(3)
site, the other doublet
tetrahedral
may represented
site, or possibly
ferric
ferric ions in another
sites.
ions in octahedral
Another factor to be considered in the peak assignments
is whether the aenigmatite spectra show evidence for mixedvalence
iron cation
Burns
species.
(1981) suggested
that
silicates with infinite chains of edge-shared Fe2 +-Fe3 +
octahedra such as ilvaite, vivianite, deerite, and glaucophaneriebeckite, etc. permit electron delocalization between Fe2 +
and Fe 3 + which
leads
also classified
silicate
to a mixed-valence
aenigmatite
minerals
state of Fe2 .5+ .
as one of the mixed-valence
that had the potential
electron delocalization.
He
for exhibiting
The phenomenon of mixed-valence iron
cation species is very well detected by the Mossbauer effect
because the isomer shift parameter is intermediate between
values for discrete Fe2 + and Fe3 + cations (McCammon & Burns,
1980; Nolet & Burns,
1979).
Due to the extreme
overlap
in the low velocity
region
and relatively small intensities of ferric peaks, initial
attempts to fit aenigmatite spectra started with six peaks
consisting of two ferrous doublets and two ferric peaks at high
velocity
in order to determine
two more peaks were
the ferric peaks
the ferric
peak positions.
added in the low velocity
at high velocity
Then
region to match
but the fits diverged
due to
27
severe overlap of Fe 2 + and Fe 3 + peaks in the low velocity
to solve these problems,
In order
region.
small ferric peaks at low velocity was
convergence,
as well as constraining
low velocity
high velocity
wth
area values were applied
to the
Fe 3 + peak such that its area equalled
a chi-square
third Fe 2 + doublet
Fe 3 +
fixing peak positions
constraints
sum of two small ferric peaks at high velocity.
convergence
led to
peak did not
but the area of the combined
Next, additional
peak areas.
combined
introduced which
to the sum of two independent
correspond
combining of the two
was added in accord
This led to
50.
value of 750
the
Finally,
with the assumption
a
of
grouping Fe2 + sites into three distinguishable categories.
Then,
convergence
reasonable
value
was achieved
of 550
To illustrate
the spectrum
and the
the fitting procedures,
of AEN2 fitted
splitting
variations
to seven and nine peaks are shown
and their parameters
of isomer shift
in the seven-peak
2
are
(6), quadrupole
(A), and Fe 3 + peak areas between
spectra, while the
the results of
Table 4 shows that there are
in Table 4.
insignificant
was lowered to a more
50.
in Fig. 3 and Fig. 4, respectively,
summarized
2
the two fitted
value is decreased significantly from 760
fit to 552 in the nine-peak
fit.
However,
releasing constraints and permitting individual doublets to
attain separate width values made convergence difficult and
produced very broad and anomalous widths of ferric peaks
-- ·
28
0
CN
Le
CN
z
0
q4
0
*O
*
44
*
o
o
CD
cN
m
L
o*
N
A0
m. 00 OD
0*
o
uo
N
co
V
N
0_
0_
0)
m
m
c'
m
o0
co
--
Q0
wI
CD
.C:
Po
m
0
o0
4J
*
-1I
0
*
0'
0
0
44
u~
~
o
~
-
L
CN
0
CN
-
LA
w9
N0
0
*
0
o
O
40
.-0'
co
LA ,t' IrN
1r
'~
0
0
.
lq
co
0v
W
CD
w-
m
LA
e
O
(
*
O
0
N
N
0
N
0
44
0
r44
w
.,-!
5-,4i
la
0
4.
0
0
-H
rzn
14
54
n
0*
K
£0
V-
.-4
-
*
*
-~
w*
co
cc
CN
IV
0
O*
~-_l
I-
0
*
O
_-
0
-
or
cs
CN
IV'
*
0
*
0
O-
.-
4Q
0
N
.14'
w.
I
I
I
,
Ln
I
L)
N
v
k0
I
co
O
I
-cl
I
LA
I
I
N-
N-
10
ID
O
:E
0
H
c4
04
+
r4
N
0
0D
rz4
0()
H
H
+
N
0
P:4
~4J^
4-)
0
.4
+
m
0)
C34
CV
CD
N\
0
+
0
0
7
4
m
0
rz4
0)
-'-4
z
.
H
H
H
_
_
+
CN
0
4
+
H
4J
.4J
H
(D
C
H
4i
-
I.
+
N
N
w
0
r4
r
+
e
0
rX4
.,1-4
a)
0
r1,4
CD
CD
04
1,4
O
0cd
.-fo
0
"I
+
0
o4
0
4.4
H
H
1-4
0
o4
Cd
0
t~
E'
H-4
04
4D
5-I
0
m
0
rZ4
K
FK
29
Fig.
3.
The spectrum
of AEN2 at room
temperature fitted with seven peaks.
The doublets
1-2, 3-4 are dominated by
Fe 2+ components and the doublets
5-7 by Fe3 + components.
5-6,
30
In
n
N
J
l
I
l
N
0
¢f4
00
we
V4
33huomgSsb
31
Fig. 4.
The spectrum
temperature
The doublets
of AEN2 at room
fitted with 9 peaks.
1-2, 3-4, 5-6 are dominated
by Fe 2 + components
and the doublets
7-9 by Fe 3 + components.
7-8,
32
-
'e
N
"-4
Li
n
V-4
I
N
I
I
I
cV-0
0O4
cn
_
33
despite
the slight decrease
2
of the
in both fitting
steps.
of more than three ferrous doublets
The resolution
or
the resolution of two independent ferric doublets, even with
heavy
(that is, fixed position
constraints
so that the nine-peak
were unsuccessful
adopted
as the final procedure
and fixed areas),
scheme was
fitting
for the other
aenigmatite
spectra.
Attempts
and positions
not even be found.
No convincing
in isomer
is significant
but is negligible
for Fe 2 +
(usually
spectra,
This phenomenon
of closely overlapping
however,
the strong
peaks in the
In the case of
spectra.
region in most Mossbauer
the aenigmatite
of quadrupole
.2 - 0.5 mm/sec)
for Fe 3 + (0 - 0.03 mm.sec).
aids the resolution
low velocity
for
shift with decreasing
for both Fe 2 + and Fe 3+, the variation
temperature
usually
can be given
explanation
However, one possible reason is that while
there is a small increase
splitting
region could
peaks in the low velocity
of ferric
the divergence.
N2 spectra were unsuccessful
to fit the liquid
overlap of Fe 2 +
and Fe3 + in the low velocity region makes the resolution more
difficult with decreasing temperature.
decreasing
velocity
temperature
move outward
and the third ferrous
the positions
In other words, with
peaks at low
of ferrous
to lower velocity,
which
makes
peaks more closely overlap
the second
the ferric
34
+
.o
a
o
o
o_
o
o
o
r
o
oI
_t0
*
o
I
04
+
00
rue
04
a,
+O
04
,4
.1-i
v
rzu
0
0
.0
*n
-.
-,
o
N
o
0
oo
oI
*
In
*
.-o
0
*
O
I-
o
o
0
oN
N
oN
Ul
41.4
.I
08
(U
H
N
4)H
ru-
0
o0
H-
+
0to
to
U
a"
_
m
o
O
'-4
·
N0
0
'44
a,
14
t.4
r4
0
(
.0
o
0
(N
-W-
'4
N
0:4
H
! H
rz,_
I_
1-
o
o
(N
0-
*
oo
CS
(N
o
N
-'
(N
(d
'-4
4)
4)
04
4)
0)
0
4
o*
0
04
'U
N
CN
Cl
4)
'4
4-'
0a
N _
4 H
rz 4
-
o
(N
_
"I
o
CN
I)
o
N
I
14
(U
4)
9-4
N
.0
(U
SZ
ft
z
o*
3n
Ct
(
'-4
Id
35
Fig. 5.
Peak Positions
for Mossbauer
spectra at room temperature.
Note the
consistency
among
of peak positions
the different specimens.
(Numbers
represent different Fe 2 + groups and
T and
represent
tetrahedral
and
octahedral Fe3 + components, respectively.)
36
_ C)
0
_c
CN-
C"-
CO
CN
-_
C"-
-
O
CN
U
UI,
0Il-
0-
0-
m
_D
rT
I-
LI)
C_CM
0 -- -
-
-L
CN -
CN
-
_---
CD
L
0
I
z
LUL
O.. Ii
CN
CY)
UJ
LU
z
z
37
Fig. 6.
The fitted
at room temperature.
to
Fig.
4.)
spectrum
(Numbers
of AEN1
refer
38
J 1
I
I|
,,N
I
I
,
I
LA
I,
r
cr}
r 4
33HHGessage
LO
It
0)m
39
Fig. 7.
The fitted
at room temperature.
to Fig.
4. )
spectrum
of AEN3
(Numbers refer
40
an
P.
N
'y
tj
uW
U
NI
U)
I"
I1
q
c:
c
"I
a
N
(
ax
I
N
w
M)
0
0
r#
a
0)
a
N
0)
go
a
(0
4l
(0
0)
41
N
X
.)
,i
z$4EU
z
uLn
LD
tr
0
(a
0D
k
*
dP
Ln
0
m
O
N
%D
l
CN
'0
r"
,
0D
O'
a~
*
*
00
*
CO
0
Ln
*
A)
I'
UL'
*
t
0*
*
N
-
(A
M
CO
0
0*
4)
4J
..
0
*
a 0a
o
e
1
r
e1
4
*
6
*
O
O
O
.f-4
0
*
0
OV
O
*
*
o
.tv
L
0
H
O
LA
N
OD
!
0
urL
o
O~
-
L
r-
LO
0
0
4-
4)
*, t0
O
4
(Ua
'U
0
O
_
O
,
OD
0
41
o
CSCS
m
z
CD
to
44
0
a)
_
.,1
-
_
N
1.0
q-
OD
D
o
*
1
(U'
O
-
r
*
-
N
0
i-
(A
0
LA
0
-
0
0
0
o
-4I
O
4)
.,,
4)
.0
o*
~-
en
cn
x*
rC,
0
0
co
.U
0
En
504
oR
CN
I
r4
.n
I
I
I !
I
0
U'
D
I
I
(m-
O
LAn~
m
I
0
,1-I
H
1.0
.Q
it
+
C
ra,
-~ H
H
+
N
4
H
H
H
+
N
0
44)
41
+
rn
O
4C.
0
H
+
+
m
-.
H
H
H
+
N
Z
%
H
-H
+
N
44
V4)
41)
co
.H
'0
,)
,
4)
E
.0
(u
.-E-4
JJ
4)
4i
41
0
0
+
+
)
(A4
(A
C7
CX
r4
W
04
(U
r14
.-JJ
H
CN
z
Wr
(a
IC
42
combined
peak.
peak positions
This reiulted
in the failure of finding
for these three
components
the
in the liquid N 2
temperature spectra.
Having now obtained the number of peaks (actually
nine), peak positions, and widths, two other room temperature
spectra of aenigmatites AEN1 and AEN3 were fitted with similar
constraints
and nine-p
to those for AEN2.
ak fits converged
those of AN2.
Table
of the peak positions
established,
study.
of both the seven-
and are in good agreement
with
5 and Fig. 5 show that the consistencies
among different
spectra are well
which is one of the main criteria
sought in this
The fitted apectra of AEN1 and AEN3 are shown in Fig. 6
and Fig. 7, respectively
Table
The results
6.
and their parameters
are summarized
in
43
III.
1.
RESULTS
Mossbauer Parameters
According
final
to the procedure
spectra were fitted
doublets
ones.
Table
with nine peaks,
and a combination
Their Mossbauer
discussed
in Chapter
that is, three
peak corresponding
parameters
II, all
to two independent
are summarized
in Table 4 and
6.
The isomer shifts (6) obtained for Fe2 + in aenigmatite
are in the range
1.106 - 1.181 mm/sec
lower value of 1.087 mm/sec
except
for the doublet
for the slightly
1-2 of AEN1.
These
values agree well with the isomer shifts for Fe2 + in octahedral
sites in other
doublets
I
correlates
grouping
:
:
III
are
roughly
of the six octahedral
attributed
site,
II
The area ratios
with the multiplicity
the assignment
Fe 2 +(II)
silicates.
3
of sites
2
:
1,
which
in terms of the
sites in aenigmatite.
for the Fe 2 + doublets
to the absorption
:
for the three ferrous
Therefore,
is that Fe 2 +(I) is
by the sites M(3), M(4) and M(6),
by the sites M(1I) and M(2), and Fe 2 +(III)
by the M(5)
respectively.
In the assignment
of the Fe 3 + peaks,
the Mossbauer
parameters suggest different site occupancies from those deduced
from published chemical analyses done by Kelsey and McKie
(1964).
The isomer
shifts of Fe 3 + are divided
into two types;
44
one type has large values
of 0.488 - 0.510 mm/sec which are
consistent with ferric iron in octahedral coordination.
other
The
type has the smaller values of 0.28 - 0.33 mm/sec
corresponding to isomer shifts for tetrahedrally coordinated
ferric
ions. Actually,
Fe 3 + in tetrahedral
0.3 mm/sec.
the isomer
coordination
The isomer
with
shifts for most compounds
fall into the range of 0.2 to
shift is a function
of the s-electron
density at the nucleus, as compared with their density at the
source nucleus.
For iron cations, the isomer shift increases
with decreasing s-electron density.
ions in tetrahedral
environments
The isomer shifts of ferric
are strongly
dependent
on the
average metal-oxygen distances which affect the s-electron
density
at the nucleus
of ferric iron, that is, increasing
in
average metal-oxygen distances by substituting bigger cation
decreases in s-electron density at nucleus.
Thus, isomer shifts
increase with "larger" average bond lengths.
Such phenomenon
was confirmed by molecular orbital calculations of s-electron
density
for tetrahderal
As shown by the chemical
ferric irons
analyses
(Tang Kai, et al., 1980).
in Table
3, AEN2 is
aluminum-free, whereas AEN1 and AEN3 contain significant amounts
of aluminum which increase average bond length of the
tetrahedral sites.
This fact may explain why tetrahedral Fe 3 +
ions in AEN1 and AEN3 have slightly higher isomer shift values.
For a better correlation
of ferric
iron in tetrahedral
coordination, Mossbauer parameters of aenigmatites were compared
45
with those of the structrally related sapphirine (Steffen &
Seifert,
are
et al., 1968) Our results
in press; Bancroft,
consistent with their parameters for the second doublet
(Table 7) of synthetic
quadrupole
splitting
respectively.
mm/sec
sapphirine
(6) and
of 0.3 and 0.76 mm/sec,
(A) values
Their natural yellow sapphirine has 6 of 0.3
and A of 0.78 mm/sec.
tetrahedral
isomer shift
having
sites results
The assignment
in some aluminum
of Fe 3 + ions to
being in octahedral
coordination, which differs from conventional site occupancies
deduced from chemical analyses.
Table 7.
This will be discussed later.
Mossbauer parameters of synthetic sapphirine* and
natural yellow sapphirine**
Synthetic sapphirine
doublet
I
doublet
II
natural sapphirine
doublet
I
doublet
6
0. 29
0. 30
0. 27
0. 30
A
1.23
0. 76
1. 37
0. 78
H
0.53
0. 45
0. 73
0.52
L
0.48
0.49
0. 59
0.52
0. 36
0. 64
0. 42
0. 58
II
Width
%Area
* from Steffen and Seifert, in press
** from Bancroft, et al., 1968
values in mm/sec relative to Fe foil
H and
L refer
to the
high-
and
low-velocity
components
of
the doublets, respectively
Both doublets are assigned to ferric in tetrahedral sites
46
It becomes apparent from the fitting procedures and
derived Mossbauer parameters that no evidence was found for
mixed valence iron cation species resulting from electron
between Fe 2 + and Fe3 +.
delocalization
in aenigmatite,
an additional
limb of the high velocity
the position
mixed
valence
where
third peak.
species would be expected
of the low velocity
envelope
exist
peak might be located on the inner
envelope
of the present
If such phenomena
it could correspond
The matching
peak for
on the low velocity
so as to give a smaller
to
side
isomer
shift than the normal value for discrete Fe2 + cations (usually
lower than 1.00 mm/sec).
not be achieved
However,
that kind of fitting
and the present fitting
scheme indicated
could
that
the fitted
line matched
especially
in the limb areas of both high and low velocity
regions.
The author
the spectrum
suggests
effects.
crystal chemistry
section.
very well,
that the aenigmatite
studied here do not show any evidence
delocalization
envelope
of Fe2 + + Fe 3 + electron
The interpretation
of aenigmatite
specimens
of peak areas and
will be discussed
in the next
47
2.
Crystal Chemistry of Aenigmatites
Chemical
data for the aenigmatites
are shown in Table 3 which
summarizes
studied in this work
wet chemical
analyses
for
AEN1 and AEN2 and an electron microprobe analysis for AEN3.
Because of the structural
of several octahedral
cmplexities,
and tetrahedral
that is, the existence
sites and the possibility
of different site occupancies by minor cations (Fe3 + and A13+),
raw chemical data only cannot sufficiently describe the detailed
crystal chemistry of aenigmatite.
With the aid of additional
information obtained from Mossbauer spectroscopy, more precise
crystal
chemistry
effectively,
can be deduced.
we shall now discuss
In order
to do so
of Fe 2 + and
how proportions
Fe3 + can be determined from Mossbauer data and then compare the
Fe 3 +/Fe2 + ratios determined by the Mossbauer and wet chemical
teniques.
In an electron
microprobe
analysis,
usually expressed as weight percent of FeO.
ferric
iron is calculated
a lower weight
as the ferrous
determined
by the microprobe
by wet analysis.
considering
will
results in
as FeO instead
That is, total weight
be lower
Such discrepancies
the percent
This means that
state which
percent of iron (expressed
Fe 20 3 ) than the actual value.
total iron is
% of oxides
than the sum obtained
can be reduced
of the actual Fe 2 0 3
of
by
and multiplying
the
FeO content by the conversion factor 1.112 which is derived from
the ratio of formula weights
iron present
of (1/ 2 Fe 2 0 3 )/FeO.
as ferric or ferrous
The amount
can be determined
directly
of
48
from the peak areas in the fitted Mossbauer
as an Fe 3 +/Fe 2 + ratio.
expressed
AEN2
in Table 8 indicate
spectra
and can be
Such ratio data for AEN1 and
that Mossbauer
analyses
yield nearly
the same results as found by wet analyses.
Using the electron microprobe data of AEN3, the correct
weight percentages of FeO and Fe 2 03 are determined as follows;
raw probe data:
41.13 wt. % FeO.
Mossbauer analysis suggests that
Fe 2 +/Z
Therefore,
Fe
the amount
41.13
However,
4.960
= 0.1206.
state is
0.1206 = 4.960 wt. %.
as FeO, and it should be Fe 2 0 3 .
by multiplying
Fe
of iron in the ferric
this value is not correct because
considered
made
Fe 3 +/E
= 0.8794
by the factor
this iron is
The adjustment
is
1.112:
1.112 = 5.516 wt. % Fe 20 3.
The value
5.52 is the actual weight
corrected
value of FeO is determined
percent Fe 2 0 3
in AEN3.
as the percentage
The
of the
probe data:
41.13
The amounts
- 0.8794 = 36.17 wt. % FeO.
of Fe 3 + in tetrahedral
can be calculated
directly
different ferric peaks.
sites and octahedral
sites
from the area ratio of the two
Table 8 gives the comparison between
Mossbauer analyses and wet analyses.
49
Table 8.
Ferric and ferrous chemical data of aenigmatite
obtained
from the Mossbauer
peak areas and wet
chemical analyses.
AEN1
AEN2
Moss
wt. %
Fe 2 0 3
Wet*
AEN3
Moss
Wet*
Moss
4.69
4.64
4.74
4.46
5.52
33.87
33.92
35.62
33.92
36.17
tet
0.56
0.52
0.558
0.213
0.636
oct
0.433
0.47
0.457
0.741
0.556
Fe 2 +
8.065
8.066
8.473
8.537
8.698
Fe 3 +/Fe 2 +
0.12
0.12
0.12
0.11
0.14
FeO
Fe 3 +
{
* from Kelsey and McKie,
1964
Finally, the cation distribution among the structural
sites of aenigmatite
Mossbauer
analyses
can be described
completely
for ferric and ferrous
by using
the
iron and the
traditional crystal chemical approach.
In the assignment
of tetrahedral
sites, the discrepancy
between Mossbauer analysis and traditional crystal chemical
methods
arises
from the large amount
of ferric
iron in
tetrahedral sites found by Mossbauer spectroscopy.
According
to Kelsey and McKie (1964), all available Si 4+ and A1 3 + occupy
50
the tetrahderal
sites,
sites available
to ferric
the relative
leaving a small proportion
iron.
This results
of tetrahedral
in A1 3 + having
over Fe 3 + for the tetrahedral
enrichment
a
sites.
However, the Mossbauer experiments indicates that a higher
proportion of ferric iron occurs in tetrahedral coordination,
resulting
sites.
in an enrichment
The values
discrepancy
of Fe 3 + over A1 3 + for the tetrahedral
of Fe 3 +/tet and Fe 3 +/oct
between Mossbauer
in Table 8 show the
and wet analysis.
The site
occupancies yield 12 formula units of cations in tetrahedral
coordination,
based on the 40 anions
(C1-, OH-, 02-) in the
chemical formula.
The remainder of A1 3 + and Fe 3 + occupy octahedral sites
together with Mg 2 +, Ti 4+, and Fe 2+.
the octahedral
sites were filled
Remaining
deficiencies
in
by Mn Z +, Ca 2+, and even Zn 2 +
cations, yielding a total of 12 formula units of octahedral
cations.
amounts
Finally,
large
cations like Na+, K+, and the small
of Ca 2 + fill the eight-fold
amounting
to 8 formula
for the aenigmatite
units.
specimens
Cation distributions
coordination
The complete
sites,
cation distributions
are given in Table 9.
for AEN2 and AEN3 in Table
that there is an excess of cations
deficiency in octahedral sites.
in tetrahedral
9 show
sites and a
This indicates that Mossbauer
analysis overestimates the amount of ferric iron in tetrahedral
sites.
This experimental
error may be due to the problem
associated with fitting of low intensity ferric peaks which make
51
Table
Si4+
9.
Cation distributions
specimens
AEN1
AEN2
AEN3
11.264
11.785
11.571
0.558
0.636
12.343
12.207
A1 3+
0.176
Fe3+
0.560
4
in aenigmatite
12.000
A13+
0.039
Fe 3 +
0.433
0.457
0.556
Mg 2+
0.699
0.573
0.003
Ti4+
2.066
1.777
1.626
Fe 2+
8.065
8.473
8.698
Mn2+
0.593
0.429
0.202
Ca2+
0.105
0.444
0.040
Zn2 +
0.036
12. 000
11.709
11.605
Ca2 +
0.029
Na+
3.969
3.799
4.321
0.014
0.014
0.021
4.011
3.813
4.341
39.804
40.000
40.000
40.000
40.000
40. 000
C1-
0.10
OH-
0.096
£
--
52
accurate fitting very difficult.
Fe3 + are allocated
tetrahedral
distributions
remained
octahedral
coordination
to octahedral
are more reasonable.
in octahedral
appropriating
If the excess amounts of
sites.
However,
This may be offset
coordination
such assignment
may be questionable.
the cation
Even so, a deficiency
sites of AEN3.
some Na+ in higher
sites,
number
still
by
sites to
of Na+ in octahedral
53
IV.
The Mossbauer
CONCLUSIONS
study on three different
aenigmatite
specimens leads to the following conclusions:
1.
Aenigmatite
igneous rocks.
chemistry
identity
of sodium-rich
is a common consitituent
Earlier uncertainties about its crystal
and structure
which
alkaline
led to some confusion
still exists.
Through
in name and
this study, two
specimens from the Harvard collection (H85123 & H85123A)
were found
to be arfvedsonite
and even in a recent paper
(Steffer and Seifer, in press), amphibole was misinterpreted
as aenigmatite.
2.
Due to strong overlap
in the low velocity
region and small
intensities of Fe 3 + components, the fitting of aenigmatite
Mossbauer spectra was heavily constrained and only
All spectra
successful in the spectra at room temperature.
were fitted with nine peaks which consist of three doublets
and a combined
mm/sec)
peak at low velocity
corresponding
(approximately
(approximately
to two small independent
0.6 and 1.0 mm/sec)
0.0
peaks
in the high velocity
region.
3.
With isomer
shifts between
1.10 and 1.18 mm/sec,
three of the
thedoublets are assigned to Fe 2 + in octahedral coordination;
Fe 2 +(I) is attributed to M(3) + M(4) + M(6), Fe 2 +(II) is
54
attributed
M(5),
2 :
to M(1) + M(2), and Fe 2 +(III)
respectively.
1 which
Fe 3 + in tetrahedral
to
Their area ratio is approximately
agrees with the multiplicity
three cation groups.
A
is attributed
Two ferric doublets
coordination
0.66 - 0.76 mm/sec)
(6 = 0.49 - 0.51 mm/sec
3 :
of sites for the
are assigned
to
(6 = 0.28 - 0.33 mm/sec
and
and Fe 3 + in octahedral
coordination
and A = 1.07 mm/sec).
Fe 3 +/tet was
also confirmed by comparison with structurally related
sapphirine.
4.
Using the peak areas
for Fe 2 + and Fe 3 + peaks in the
Mossbauer spectra, the electron microprobe data for iron
were
recalculated
Fe2 03 .
to give weight
percentages
of FeO and
Comparisons of Fe 3 +/Fe2 + ratios determined wet
chemically and by Mossbauer spectroscopy demonstrated the
accuracy of Mossbauer analysis and yielded more precise
crystal
chemistry
of aenigmatite
than could be achieved
the traditional crystal chemical approach.
by
The imbalance of
total cation numbers in tetrahedral sites and octahedral
sites indicates
overestimate
of difficulty
that some experimental
of Fe 3 +/tet.
error occurred due to
Such an error may be the result
in the resolution
of very low intensity
peaks.
However, the existence of significant amounts of Fe 3 +/tet
suggests a different cation ordering in tetrahedral
coordination from that deduced by traditional methods, that
is, Fe 3 + has a preference
sites.
over A1 3 + for the tetrahedral
55
5.
Finally,
the spectra of aenigmatite
specimens
used in this
study show no evidence of electron delocalization suggested
by Burns
because
(1981).
However,
the three
samples
of all aenigmatites.
this conclusion
is not final
studied may not be representative
The author
suggests
that more
specimens with higher Fe3 + ion concentration and further
Mossbauer studies might yield evidence for electron
delocalization in aenigmatites.
56
BIBLIOGRAPHY
G. M.,
Bancroft,
Introduction
Spectroscopy:
1973, Mossbauer
to
Inorganic Chemists and Geochemists, McGraw-Hill, N. Y.
G. M.,
Bancroft,
G. Burns,
R.
of the Mossbauer
Stone,
Applications
mineralogy-II.
crystal
transitions
Iron
structures,
in mixed-valence
Ann. Rev. Earth &
of iron and titanium,
minerals
1968,
32, 547 - 559.
Burns, R. G., 1981, Intervalence
Cameron,
J.
and complex
Geoch. et Cosmo. Acta,
Planet.
A.
to silicate
effect
of unknown
silicates
and
Sci., 9, 345 - 383.
K. L., M.F. Carman,
and J. C. Butler,
from Big Bend National
1970, Rhonite
Park, Texas, Amer.
Mineral.,
55,
1587 - 1607.
Cannillo,
E.,
F. Mazzi,
J. H.
1971, The crystal
P.
D. Robinson,
structure of aenigmatite,
1962, Pantelleritic
I. S. E,
phenocrysts,
Min. Mag.,
Y. Ohya,
Amer.
liquids
1978, Rock-Forming
(2nd ed.): Vol. 2A, Single-Chain
John Wiley
& Sons, N. Y.
and their
33, 86 -113.
Deer, W. A., R. A. Howie, and J. Zussman,
Minerals
and
56, 427 - 446.
Mineral.,
Carmichael,
Fang,
Silicates,
57
Dollase,
W. A.,
1975, Statistical
spectral
fitting,
Dornberger-Schiff,
limitations
Amer. Mineral.,
K. and S. Merlino,
of Mossbauer
60, 257 -
264.
1974, Order-disorder
in
sapphirine, aenigmatite and aenigmatite-like minerals,
Acta Cryst., A30, 168 - 173.
Higgins,
J.B., P. H. Ribbe, and R. K. Herd,
1979, Sapphirine
I:
Crystal chemical contributions, Contr. Min. Petr., 68,
349 -356.
Higgins, J. B. and P. H. Ribbe,
1979, Sapphirine
II: A neutron
and X-ray diffraction study of (Mg-Al)VI and (Si-Al)IV
ordering in monoclinic sapphirine, Contr. Min. Petr.,
68, 357 -
368.
Huggins, F. E., 1974, Mossbauer
Pressure
Studies
of up to 200 Kilobars,
of Iron Minerals
Under
Ph.D. Thesis, M.I.T.,
Cambridge, Massachusetts.
Kelsey,
D. H. and
Min. Mag.,
D. McKie,
1964, The unit cell of aenigmatite,
33, 986 - 1001.
Liebau, F., 1980, Classification
ed., Orthosilicates,
Machin, M. P. and P. Ssse,
of silicates,
In P. H. Ribbe,
Rev. in Mineralogy
1974, Serendibite:
Vol.
5, 1 - 24.
a new member
of
the aenigmatite structure group. Neues Jahrb. Min.,
Mh., 435 - 441.
58
McCammon,
C. A. and R. G. Burns,
1979, The oxidation
mechanism
of vivianite as studied by Mossbauer spectroscopy,
Amer. Mineral.,
65, 361 -366.
Merlino, S., 1970, Crystal chemistry
Comm.,
of aenigmatite,
Chem.
20, 1288 - 1289.
i
Merlino, S., 1972, X-ray crystallography
Krist.,
136,
41,
in sapphirine,
fur Krist.,
Morse,
of sapphirine
54,
S. A. and
-
Tc, Zeit.
151, 91 -100.
P. B., 1969, The crystal structure
Mineral.,
Contr. Min.
23 -29.
Herlino, S., 1980, Crystal structure
Moore,
Z.
81 -88.
Merlino, S., 1973, Polymorphism
Petr.,
of krinovite,
34 -
of sapphirine,
Amer.
49.
J. H. Tal'oy,
1971, Sapphirine
reactions
in
deep-seated granulites near Wilson Lake, central
Labrador, Canada, Earth Planet. Sci. Letters, 10,
325 - 328.
Nolet,
D. A.
and
R. G.
Burns,
1979,
Ilvaite:
a study
of
temperature dependent electron delocalizatrion by
Mossbauer
Palache,
effect,
Phy. Chem. Mineral.,
C., 1933, Crystallographic
aenigmatite,
4, 221 - 234.
notes on anapaite,
and endidymite,
Z. Krist.,
86, 280 - 291.
59
Ruby,
S. L., 1973, Why MISFIT when you already have X 2? In:
Mossbauer
ed.),
Steffen,
Effect Methodology,
Vol. 8 (J. Gruverman,
Plenum Press, N. Y.
G. and F. Seifert,
1982, Ferric
iron in sapphirine:
A
Mossbauer spectroscopic study, in press.
Stone,
A.
J.,
H. J. Augard,
and
J. Fenger,
1969,
General
constrained non-linear regression for Mossbauer
spectra. publ. Danish Atomic Energy Comm., R150-M-1348.
Tang Kai, A., H. Annersten,
and T. Ericsson,
1980, Molecular
orbital (MSXa ) calculations of s-electron densities of
tetrahedrally coordinated ferric iron: Comparison with
experimental
isomer
shifts,
Phys. Chem. of Mineral.,
343 - 349.
Thompson,
R.
N. and
aenigmatite,
J. E. Chisholm,
Min. Mag.,
Yagi, K. and J. G. Souther,
1969,
Synthesis
of
37, 820 - 829.
1974, Aenigmatite
from Mt. Edziza,
British Columbia, Canada, Amer. Min., 59, 820 - 829.
5,