American Mineralogist, Volume 79, pages 411-425, 1994
Use of electron-energyloss near-edgefine structure in the study of minerals
LlunBNcn A. J. G.lnvrnr* AuN
J. Cnc'YEI'{
Department of Physicsand Astronomy, University of Glasgow,Glasgow Gl2 8QQ, U.K.
Rrx BnvosoN
Department of Materials Scienceand Engineering,University of Surrey, Guildford, Surrey GU2 5XH, U.K.
Arsrucr
High-resolution electron-energyloss near-edgefine structure (ELNES) recorded in a
scanningtransmission electron microscope (STEM) is shown to provide information on
the local structure and bonding of specific types of atoms in minerals. The Lr,, ELNES
from Fe (Fe'z+and Fe3*), Mn (Mn'*, Mn3+,and Mno*), and Cr (Cr3+and Cr6*) show
valence-specificmultiplet structures that can be used as valence fingerprints. In general,
the Z, edgefor a specific3d transition metal exhibits a chemical shift toward higher energy
losseswith an increasein oxidation state. Examplesof mixed valenceFe- and Mn-bearing
minerals are presentedwhere the presenceof multiple valencestatesis distinguished by a
splitting of the L, edge.The high spatial resolution that can be obtained using the STEM
allows variations in the relative proportions of the oxidation statesto be detected on a
scaledown to I nm2. This resolution is illustrated from a sample of hausmannitethat
shows different lr-edge shapesconsistent with variations in the Mn2+-Mn3+ ratio over
distancesof ca. 100 nm. Furthermore,spectraof many elementsexhibit ELNES shapes
characteristic of the nearestneighbor coordination, as is demonstrated for C in the carbonate anion and Si in the SiO" tetrahedral unit. The C K edge from CO3- is compared
with that for elemental forms of C that exhibit very different ELNES. Similarly the Si lr,,
ELNES from a rangeof SiOf -containing minerals all show the same near-edgeshapethat
is very different from Si and SiC. ELNES allows for its semiquantitative analysis,which
is illustrated by two theoretical techniques. Finally, the effectsof electron beam damage
are discussedin relation to the experimental changesobservedin the ELNES.
IxrnonucrroN
In a transmission electron microscope (TEM), the
high-energyincident electrons that have been elastically
scatteredmay be used to obtain images and diffraction
patterns from specific regions of the sample. This allows
important structural information to be deducedabout the
sample. The high spatial resolution offered by TEM has
been usefulfor studyingsubmicroscopictwins and exsolution lamellaein minerals.The TEM has also beenvery
useful in mineralogy for high-resolution structure imaging. The incident electrons also undergo inelastic interactionswith the sample(Joy, 1993)and can provide further information on the specimen.That has given rise to
a number of analyticaltechniques,such as electron-energy loss spectroscopy(EELS), energy-dispersiveX-ray
spectroscopy(EDX), and Auger emission spectroscopy
(AES).In EDX and AES it is the secondaryprocesses
of
electron and photon emission that are studied, whereas
in EELS the signal is due to the primary processof electron excitation. EELS can provide valuable information
about the material under study, such as a quantitative
* Presentaddress:Department of Geology,Arizona StateUniversity,Tempe,Arizona 85287-1404,U.S.A.
0003-o04x/94/ oso644 | l $02.00
chemicalanalysison the nanometerscale.More important, the electron-energyloss near-edgestructure(ELNES)
of a core-loss edge can also provide information about
the valency,coordination,and site symmetryof the atom
undergoing excitation (e.g.,Brydson et al., 1992a).
Recent advances in detector design, specifically the
availability of parallel EELS (PEELS)spectrometers(Krivanek et al., 1987; Krivanek, 1989),allows data to be
collected that are comparable with those obtained by
X-ray absorptionspectroscopy(XAS). Until recently,most
EELS systemsemployed serial detectors with which the
spectrum was acquired by scanning the signal across a
singledetector,one channelat a time. The development
of parallel detectors,specificallymultichannel arrays, has
allowed a large gain in the collection efrciency, typically
by a factor of 500, compared with earlier serial spectrometers. Parallel detection allows high-resolution spectrato
be acquired from materials that were too beam sensitive
to be collected by serial EELS at comparable spatial resolutions.
This paper demonstratesthe great potential of PEELS,
used in conjunction with a TEM, for providing information about the crystalchemicalenvironmentof atoms
in minerals.Clay minerals and clay-sizedmineralswere
chosen for this study, as they can be difficult to charac-
4tl
412
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
0
350
50
ELNES
400
Energy
Loss(eV)
100
D
450
Fig. l. A representative
EELSspectrumshowing(a)thezerolosspeakand low-lossregionand (b) a coreJossedgefrom the
mineralgaudefroyite.
To highlightthe fact that most electrons
do not participatein an inelasticscatteringevent,the zeroJoss
peakis shownto scalein a, togetherwith an expandedview
illustratingthe lowJossregion.Followingthe low-lossregionis
the monitonicallydecreasing
background.(b) The edgeonset
exhibitstwo intensewhite-linefeaturesdueto the CaLr.. edge.
The EXELFSmodulationsare very weakand only just visible
on thetail ofthe edge.
terize by conventional analytical techniques becauseof
their fine-grained and often heterogeneousnature. The
paper concentratesprimarily on two types of information
that can be gained from a qualitative analysis of the
ELNES spectra:(l) determination of the valency of 3d
transition elements, and (2) identification of local anion
and cation coordinations.The ability to model ELNES is
illustrated by two theoretical methods that provide semiquantitative information on the core-lossedgesand hence
about the atom undergoing excitation. Consideration is
also given to effectsinduced by the electron beam on the
sample as revealedby the ELNES. Only a brief introduction to EELS is given; more detailed accounts can be
found in Egerton(1986)and Buseckand Self(I992).There
are many papersthat deal with ELNES; introductory papers include Colliex et al. (1985),Brydson(1991),and
Brydson et al. (1992a).
Tgn EELS SPEcTRUM
An EELS spectrum(Fig. l) displaysthe electronintensity as a function ofenergy loss and can be divided into
several regions. The large zero-loss peak (ZLP) results
from transmitted electronsthat have undergoneelastic
and quasi-elastic(mainly phonon) interactions,i.e., electrons that have lost no, or very little, energy.Its width
depends primarily on the energy spread and stability of
the electron gun and determines the best resolution that
can be obtained at the core-lossedges.The region immediately following the ZLP and extending to energy
lossesof ca. 50 eV, called the low-lossregion,is an area
of the spectrumdominated by plasmons.Plasmonscan
be described as resulting from the collective excitations
of valenceelectrons.They can provide information about
the dielectricfunction (Buechner,1975),valenceelectron
densities,and, in suitable cases,the phasespresent in
alloys. Within the low-loss region are also found edges
causedby transitions from outer shell electrons.Extending beyondthe low-lossregionin a spectrumfrom a thin
sampleis a monotonically decreasingbackground,arising
predominantlyfrom plasmon and singleelectronexcitations, on which are superimposedthe coreJossedges.
Thesefeaturesresultfrom the transition ofcore electrons
to unoccupiedstatesin the conductionband; this requires
the energy transfer between the incident and a core electron to be greater than its binding energy.The coreJoss
edgesusually take the form of a step, characterizedby a
sudden increasein intensity (Fig. lb) that decreasesin
intensity with increasingenergyloss.This suddenrise in
intensity representsthe ionization threshold,the energy
of which approximately corresponds to the inner-shell
binding energyand henceis characteristicofthe element.
Close to the thresholds and for small scatteringvectors,
thesecore-lossedgetransitions are governedby the atomic dipole selection rules for the electronic transitions A/
: + l, and Aj : 0, +1, where I andj are the orbital and
total angular momentum quantum numbers of the subshell from which the electron has been excited. The edges
are classifiedaccording to the standard spectroscopicnotation, e.9.,the Ca Lr' edgesin Figure lb arisefrom transitions from the 2p core level.
In the first instance EELS can be used for elemental
analysis,sincethe edgesoccur at energylossesequivalent
to the binding energiesof the core electronsin the sample,
and the integrated intensity under the edge is related to
the number of atoms under the electron beam. Unlike
EDX, EELS can provide elemental data on all the light
elementsexceptH and He, although the latter has been
detectedunder certaincircumstances.
e.g.,in nanometersized He bubbles in irradiated Ni alloys (McGibbon,
r 9 9l ) .
Core-loss edgescan be divided into two regions (Fig.
lb), the ELNES, extending30-50 eV abovethe edgeonset, and the weaker extended electron-energyloss fine
structure (EXELFS). The equivalent XAS terms are
XANES (X-ray absorption near-edgestructure) and EXAFS (extendedX-ray absorption fine structure). In a solid, the core wave functions show very little change as
compared with an isolated atom. This is not the casefor
the outer bonding and antibonding orbitals that can in-
413
GARVIE ET AL.: ENERGY LOSS NEAR.EDGE FINE STRUCTURE
TABLE1. Name,simplifiedformula,valence,coordination,and sourceof the 3d transitionmetalcontainingminerals
Mineralname
Simolifiedformula*
Valenceand
coordination
Chromite
Siderite
Hematite
lron analogue of leucite*.
Chlorite (var. daphnite)
Cronstedtite
Nontronite
Fe-bearing minerals
r4tFe2+
(Fe,MgXCr,Al,Fe),04
r6tFe2+
(Fe,Mn)COg
16lFe3+
Fe,O"
r41Fe3+
KFeSi,O6
t61Fg2+
nH,O
(Fe,*,AD,,(Si,AD6O,o(OH),6'
l6iFe2+, {5lFe3+, I4lFe3+
(Fe'z*,Fe3*
nH,O
)6(Si,Al,Fe)1Olo(OH)6'
rsrFeF*
nH2O
Ctu e5Fe4(Si,Al)sOro(OHL
Ramsdellite
Asbolan
Norrishite
Bixbyite
Ganophyllite
Manganosite
Hausmannite
Mn-bealing minerals
t61Mn4+
MnO,
I6rMn1+
Mn(O,OH).(Co,N|),(OH),, nH,O
t6lMn3+
KLiMn4SisOr4
2 x I61Mn3+
(Mn,Fe),O3
16rMn2+
(K,Ca,Na),(Fe,Mn)"(Si,Al),,O,e(OH),.nH,O
I6lMn2+
Mno
l4lMn2+,I6lMn3+
c-Mn"Oo
Crocoite
Volkonskoite
Chromite
Cr-bearing
PbCrO.
Ca,,(Cr,Fe,Mg)o(Si,Al).O,0(OH)o'nHrO
(Fe,MgXCr,Al,Fe),O4
minerals
I61Cr6+
16ro13+
t6rcr3+
Source
Gebe lsland, Indonesia
lvigtut, Greenland
Langban,Sweden
Redruth, Cornwall, England
WhealJane, Cornwall,England
Lajitas,Texas
Pirika Mine, Hokkaido, Japan
Gebe lsland, Indonesia
HoskinsMine,N.S.W.,Australia
ThomasMt., Utah
Langban, Sweden
Nordmark, Sweden
KalahariMine Field,S. Africa
RedleadMine, Dundas,Tasmania
Glazov Region, Udmurtian
Gebe lsland. Indonesia
' Maior elements based on qualitative EDX analysis
't Synthetic Fe analogueof leucite
teract to form bands oforbitals that are governed by the
solid-state character of the specimen and can be appreciably modified by chemical bonding. The ELNES directly probes the unoccupied orbitals and therefore reflects the environment surounding the excited atom and
can provide information on bonding, valency, coordination, and site symmetry.Note that when discussingmicroanalysisthe edgeis taken to mean all transitions from
a given initial state,whereasin discussionsof ELNES or
XANES, the edge is taken to mean the structure in the
region closeto the threshold.
The EXELFS oscillations are very weak modulations
ofthe intensity extending several hundred electron volts
above the ELNES region. These oscillations are caused
by an ejectedelectron with a low kinetic energybehaving
similarly to a free electron. The ejectedelectron is backscatteredby the surrounding atom shells,which can provide information on the coordination number and nearest neighbor distances.EXAFS studies most commonly
involve analysis of K-edge modulations, such as those
from the transition elements.TheseK-edgesoccur at energy lossesof several thousand electron volts, where individual edgesare well separatedin energy. Unlike its
sistertechniqueEXAFS, EXELFS is unlikely to be widely
used, since the fine structure modulations often overlap
with other edgesin the energyrangeused to collect EELS
edges,typically <1000 eV.
ExpnnrprnNrAl. sETUP
The instrument used was a VG HB5 STEM operating
at 100 kV and equippedwith a cold field emissiongun
(FEG). With this microscope,a probe size of ca. l-nm
diameter was scannedover the sample.The microscope
was operated with a probe semiangleof I I mrad, a collection angleof 12.5mrad, and a probecurrent of ca.0.5
nA. The small collection semiangle,B, ensuresthat the
experiment is in the dipole regime for electronic transitions. A Garan 666 parallel electron-energyloss spectrometer (PEELS)is attachedto the top of the microscope
column (Krivanek et al., 1987; Krivanek, 1989).To exemplify the advantageof parallel detection, an 8-s acquisition on the PEELS would require just over I h on a
serial detectionsystem.As well as requiring patiencefrom
the operator, long counting times have other associated
problems,includingC contamination,lack of microscope
stability,and beam damageto the sample.The spectrometer was calibrated againstthe Ni I, edgein NiO that has
a peak maximum at 853.2 eV, as determinedusing synchrotron radiation studies(van der Laan et al.' 1986).
The samplesused in this study have come from a variety of sourcesor were collectedby one of us (L.G.) and
identified by powder X-ray diffraction and EDX. The
minerals containing the 3d transition metal elementsare
listed in Table l, together with their formulae, selected
crystallographicdata, and sources.Table 2lists minerals
used in the coordination fingerprint study. Materials
studied were either checked prior to study in the HB5
STEM by EDX in an SEM, or checked qualitatively by
EDX in the HB5 STEM. Crystalline samples were prepared by crushing selectedmineral grains in acetoneand
allowing a drop of the fine-grainedmaterial in suspension
to dry on a lacy C-coated Cu TEM edd. Alternatively,
clay sampleswere preparedby dispersinga small amount
of material in distilled HrO, and a small drop of the <2rrm clay fraction in suspensionwas dried on the above
grids. Lacy C films were used so that data could be recorded from thin electron-beam transparent grains projecting over the holes.Core-lossedgeswere recordedwith
a dispersionof 0.1 or 0.05 eV per channel in order to
record the detailed fine structure ofthe edges.The actual
energy resolution is limited by the energy spread of the
FEG, which is closeto 0.27 eV at low probe currents.At
4t4
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
TlsLe 2. Name,simplifiedformula,and sourceof C- and Si-containingminerals
Mineralname
Calcite
Siderite
Desautelsite'.
Hydrotalcite
Amorphous C**
Graphite
Diamond
Simplifiedformula'
C-containing minerals
CaCO"
(Fe,Mn)CO3
MguMn.(CO.)(OH)16
4H,O
Mg6Al,(C03XOH)16.
4H,O
c
LlanelweddQuarry, Wales
lvigtut, Greenland
Snarum, Norway
South Western Graphite Mine, Texas
South Africa
Si-containingminerals
Quartz
Talc
Nontronite
B silicon carbide-'
si--
c-SiO,
(Mg,Fe)6(Si,Al)sO,o(OH)4nH,O
Caoe5Fe4(Si,Al)so,o(OH)4
nHrO
sic
LlanelweddQuarry, Wales
LlanelweddQuarry, Wales
Lalitas, Texas
5l
' Major
elementsbasedon qualitativeEDX analysis
.- Synthetic material.
normal operating currents, the width increasesto a minimum of ca. 0.3 eV, which is further increasedto ca. 0.45
eV by residual interference.The EELS data weretypically
obtained with a scanned raster from areas in the range
I 30 x 100to I 3 x 10 nm, with acquisitiontimes ranging
from 0.5 to 8 s. With a samplethicknessof 20 nm and
an irradiatedareaof 13 x l0 nm2, this correspondsto a
volume of roughly 2600 nmr and an approximatemass
of 8 x l0 " g (with p :3 g/cm3).Since the analyzed
volume is so small, the areaprobed is almost invariably
part of a singlecrystal.In a typical EELSexperiment,the
dosegiven at the samplevariesfrom 1.25 x 1gse/At for
a 0.5-s Io 2 x l}e e/A' for an 8-s integrationtime. The
choiceof magnificationand acquisitiontime is governed
by the stability of the material under the electronbeam.
For eachcore-lossedgerecorded,three additional spectra
were collected:the dark current for the core-lossedge,the
low-lossregion,and its correspondingdark current. The
dark current arisesfrom thermally excited current in the
detector. The dark current is then subtracted from the
corresponding spectrum. Before presentingthe core-loss
edges, a background of the form AE-' was subtracted
from beneath the edge,and the effect of multiple scattering and asymmetry of the ZLP were deconvoluted from
the edge. The adjustable parametersA and r can be deducedby least-squares
fitting ofthe intensityofa specific
region directly preceding the core-lossedges.The topics
of background subtraction and ZLp deconvolution are
describedin detail in Egerton(1986).
Sample thickness is an important factor when collecting EELS spectra.In order to reducethe effectof multiple
scatteringthe thicknessshould be <50-100 nm for 100kV incident electrons(Egerton,1986).Ifthis condition is
not met, then there is an increasedprobability that the
transmitted electron will be involved in more than one
inelastic scattering event. Although the effect of a small
amount of multiple scatteringcan be removed using deconvolution techniques,it becomesrapidly more difficult
to detectedgesas the specimenthicknessincreases.
In the
samples studied, abundant thin areas could always be
producedby simply crushingthe materialin a mortar and
pestle.The phyllosilicatesare ideal materials for EELS
study becauseoftheir platy nature,which providesabundant electron-beamtransparentareas.
ELNES
Valency determination of 3d transition metals
All of the 3d transition metals are found in natural
phyllosilicates,and they can be a major chemical component. Examplesinclude roscoelite(V3*), volkonskoite
(Cr.* ), glauconite(Fe.* ), pennantite(Mn'* ), and pimelite (Nir* ). Fe is by far the most common 3d transition
metal in minerals.Many spectroscopictechniqueshave
been usedto determineboth the oxidation stateand coordination of 3d transitionmetalsin minerals:57FeMdssbauer spectroscopy(Cardile and Brown, 1988; Hawthorne, 1988), optical spectroscopy(Rossman, 1988),
XANES (van der Laan et al., 1992; Manceauet al., 1992:
Cresseyet al., 1993), electron paramagneticresonance
(Calas, 1988), photoelectronspectroscopy(Rao et al.,
1979), and high-resolutionX-ray emissionspectroscopy
(Kosterand Mendel, 1970;Wood and Urch, 1976).Analysis of K-edgeXANES has proved usefulin determining
the coordination and oxidation stateofselectedtransition
metals in minerals (Calasand Petiau, 1983;Manceauet
al., 1992).The feature that most reflectsthe crystal chemical environment of the transition elementsis a preedge
feature before the main K absorption edge.This has been
interpreted as arising from predominantly dipole-forbidden 1s- 3d transitionsand so is fairly weak in intensity.
In contrast,the dipole-allowedtransitionsthat give rise
to the Ir., edgesexhibit numerous strong sharp features
or multiplet structure. Fine structure is not readily observable on the K-edge prepeak becauseof the greater
core-hole lifetime broadening associatedwith the K edge
than with the Ir,. edges.The broadeningon the Ir., edges
of the 3d transition elementsis typically 0. 1-0.3 eV,
whereasit is three to four times larger on the corresponding K edge.The use of PEELSto determine the oxidation
4t5
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
state and, in certain circumstances,the coordination of
transition elementsis illustratedusingthe lr., edgesfrom
Fe. Mn. and Cr.
SeveralELNES and XANES studieshave shown that
lhe Lr.. edges from the 3d transition elements exhibit
structures that are sensitive to the valence state of the
metal (Otten et al., 1985;Otten and Buseck,1987;Rask
et al., 1987:Krivanek and Paterson,1990;Patersonand
Krivanek, 1990; van der Laan et al., 1992; de Groot et
al., 1993; Garvie and Craven, 1993, 1994). The ELNES
of the 3d transition melal Lr. edgesare characterizedby
two whiteline features(so called becausethey were first
identified as such on photographic plates) that are narrow, intensepeaks.These white lines arise from quasiatomic dipole-allowedtransitions from an initial state,
2p63d', to final statesof the form 2psJl'+t. The high
intensity of the 3d transition metal Lr,, edgesis due to
the high density of unoccupied3d states.The two separate edges,l,, and lr, arise from the two ways that the
spin quantum number, s, can couple to the orbital angular momentum, I giving the total angularmomentum,
j : I + s. That givestwo separatepeaksdue to transitions
from 2p,, (j -- lr), forming the L' edge, and 2pn U :
'/r), forming the I, edge. Final states of s character are
also dipole allowed, but they are lessintense becausethe
dipole-allowed2p - s transitionsare approximatelyan
order of magnitude less intense than 2p - d transitions'
This, coupled with the fact that s and p densitiesof state
are rather broad compared with d states,implies that the
Zr., edgesessentiallyrepresenttransitions to the vacant
3d orbitals. Hencethe ELNES reflectsthe local environment and charge state of the 3d transition metal ion in
question.
An isolated 3d transition metal ion such as Mn2* has
five degenerate3d orbitals. In a solid, the effect of the
field from the surrounding ligands is to split the 3d orbitals into two or more separateorbitals.In minerals,the
3d transition metals are predominantly octahedrally coordinated and exhibit a wide range of symmetries. For
high-spin compoundsthe effectof the ligandsis to split
the degenerate3d orbitals into the lower energy tr, and
higher energy e, orbitals. The energy difference between
these two sets of orbitals is defined as the crystal field
parameter,designatedas lODq (or A in some texts).As
will be seen later, the crystal field can have large effects
on the shapeof tllreLr.t edges.In the following sections
on the I- edges,the discussionswill be primanly concerned with the analysis of the Z. edges,which are more
intense and exhibit more multiplet structure than the
broader I, edges. The increased broadening of the It
edge,in comparison with the L, edge,is due to the availability of the extra Coster-Kronig Auger decay channel,
which reduces the lifetime of the final state and hence
increasesthe width of the L, multiplet lines.
Fe. Fe is found as both Fe2*and Fe3*in minerals,and
it is often the case that more than one valency and coordination is presentin the samemineral. Minerals containing Fe2* and Fe3* are most commonly octahedrally
Fe2+
Fe L2,3
I
705
720
715
710
Loss(eV)
EnergY
725
materials.Syntheticiron
Fig.2. Fe I- edgesfrom reference
)'
of leucite(torFet*
), siderite(t6lFe'?+
analogue
), hematite(r6rFe3+
and chromite(lolFet*). The spectraillustratedhereand in subsequentfiguresareshownon an arbitraryscale.
coordinated, although Fe in tetrahedral coordination is
sometimes found. Fe in minerals is expectedto be in the
high-spin state,as only strong field ligandssuch as CNcan cause spin pairing. Before the discussion of the Fe
l,., edgesfor severalphyllosilicates,the Fe Ir.. edgesfrom
a number of referencemineralscontainingFe2+and Fe3+
in both octahedral and tetrahedral coordination are dis)' hecussed.They are chromite (t4lFe2+
), siderite(l61Fe2+
and the Fe analogueof leucite (t41Fe3+).
matite (r6rFe3+),
The Lr.. edgesfrom the referencecompounds (Fig. 2)
all show different edgeshapes,as well as a distinct chemical shift between the divalent and trivalent Fe Z, edges
(Table 3). These basic edge shapesare consistent with
et al., 1993)and ELNES
publishedFe I, XANES (Cressey
(I(rishnan, 1990;Kurata et al., 1990)spectra.The l,, edges for the divalent Fe minerals arc at ca. 707.8 eV, and
the trivalent Fe Z, edgeis 1.7 eV higher,at ca. 709.5 eV.
The separation of the L, and Lt peak maxima, due to
spin orbit splitting, is ca. l l eV. The Fe I, edgesfrom
the referencecompounds illustrate that different site symmetries are recognizable by changes in the multiplet
structureson the Z3 edges.Hematite has a distinctive Iredge shape, with a prepeak leading the main L, peak,
whereasthe Fe analogueof leucite, which contains talFe3+'
416
TABLE3.
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
Energies (eV) of the ln and In edge peak maximum
for selected3d transitionmetal-containing
minerals
Mineral name
Siderite
Hematite
Manganosite
Norrishite
Bixbyite
Asbolan
Chromite
Crocoite
Element valence
and coordination
lo peak max.
I? peak max.
707.8
709.5
640.2
642.5
641.8
643.8
578.6
580.9
720.4
722.6
650.3
653.6
t6lFg2+
16lFe3+
t6lMn2+
16lMn3+
l6lMn3+
161Mn4+
I61Cf3+
r4tc16+
ocz.o
6s4.3
s86.6
589.5
Note: efior in energy values is +0.2 eV
shows an Z, edgewithout any pronounced fine structure.
fhs telpsz+L, edgefrom siderite shows several partially
resolvedmultiplets on both the low- and high-energysides
of the l,, peak maximum, whereastheseare not present
on the Z,3edgeof chromite. However,the peak at ca. 7l0
eV is much strongerin chromite than in siderite.The I,
edgesfrom hematite, the Fe analogueof leucite, and siderite correspondcloselywith the theoreticalatomic multiplet spectra calculated by van der Laan and Kirkman
(1992), which confirms the singlevalenceand site symmetry of the Fe in thesesamples.However,for chromite
the experimental edgehas a much larger subsidiary peak
aI ca.7I0 eV than the calculatededgefor lalFer+with the
appropriate 10Dq. This subsidiary peak is at approximately the correct energy for the L, edgefrom Fe3+.It
therefore appearsthat the chromite contains severalpercent of Fe3* substitutingfor I6tCr3+,as is often the case
for chromite (e.g.,Yang and Seccombe,1993).The presenceof Fe3+in octahedralsiteswas confirmed by EDX,
which showed that the majority of octahedral sites are
filled with Cr and Al, with approximatelyone out of every 16 octahedralsitesoccupiedby Fe.
The two peakson the I, edgein hematite are separated
by 1.5 eV, as they are in many other mineralscontaining
t6lFer*(unpublishedresults).Thesetwo peakshave been
interpreted in terms of a simple crystal field model
(Krishnan, 1990),where they have been assignedto the
tr, and e, orbitals. Although such a simplistic ligand field
approach appears to explain the features on the Fe I.
edge, as well as on the Mn I, edgesfrom minerals containing Mn3+ and Mn4+ (Garvie and Craven, 1993)and
compounds with the valence of the 3d transition metal
in the d0 state(Brydsonet al., 1987, 1989a),that is not
generallythe case(de Groot et al., 1990a, 1990b).The
value of the crystal field strength for hematite can be estimated by comparing its I, edge with the calculated
atomic multiplet spectra of van der Laan and Kirkman
(1992). Their best fit to the experimentalhematite I. edge
is to the spectrum calculated for Oo symmetry with a
1ODq value of 2 eY, which is closeto the value of 1.9 eV
determined by optical spectroscopyfor hematite (Marusak et al., 1980).Similarly, the values of the 10Dq can
be estimatedto be I and 0.5-l eV for the Fe analogueof
leucite and siderite, respectively.
705
7't0
715
720
EnergyLoss(eV)
725
Fig. 3. Fe Ir., edgesfrom threephyllosilicates.
Nontronite
(dominantlyt61Fe3+
(threedistinctFe sites,16lFe2+,
), cronstedtite
I6lFe3*,
r6lFez+
andtotFer*
(dominantly
) andthechloritedaphnite
with a few percentlolFer+
).
The Fe Zr.. edgesfrom a number of phyllosilicatescontaining various Fer+-Fe2+ratios were collected,and representative spectra are illustrated in Figure 3. The data
werecollectedfrom individual well-definedclay particles,
typically from areasmeasuring130 x 100nm2.Daphnite
has an Fe Lr., edge shape similar to that from siderite,
confirming both that the Fe is octahedrally coordinated
and divalent. The higher intensity of the partially resolved peak on the high-energy side of the I, edge, in
comparison with that on the siderite I. edge, may indicate the presenceof a small amount of Fe3*. That has
been confirmed for this sampleby 57FeM6ssbauerspectroscopy (Goodman and Bain, 1979).Cronstedtiteconr6tFe3+,
tains a number of Fe Sites:r4tFe3+,
and r6lFe2+
(Bailey, 1988). The presenceof both Fe3+and Fe2+is
visible from the double-peakedZ. edge, the two edges
being separatedby 1.7 eY. The partially resolvedpeak
on the low-energy side of the Fe2* L, edgeindicates that
it is dominantly octahedrally coordinated, as expected
from the crystal structure. Although the presenceof Fe3+
is confirmedby the peak 1.7eV higherin energythan the
Fe2* edge, it is not obvious from the L, edgethat both
I6lFe3+
and [alFe3+
arepresent.Nontronite givesan Fe lr.redgeshapeidentical to that from hematite, indicating that
the majority of the Fe is teJ['sr+,as expectedfor nontronite (Giiven, 1988).The splitting betweenthe two peaks
on the L, edgein nontronite is slightly lessthan in hematite, indicating a slightly smaller crystal field value.
4t7
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
This is consistentwith the l0Dq of 1.77eY for nontronite determined by optical spectroscopy(Bonnin et al.,
1985).As illustratedby comparingthe Z, edgefrom hematite with that from nontronite, it is possibleto obtain
a relative, and possibly quantitative, measureof the crystal field strength acting on the metal ion from the multiplet splittings.
Since the Fe2* and Fe3* Z. edgesare well separatedin
energy,it is possibleto determine Fe3+-Fe2+ratios by
fitting referencespectrato the experimental Fe Ir,, edges.
This hasbeendemonstratedby Cresseyet al. (1993)using
XANES from a number of Fe-containingminerals having
mixed valency and coordination. The same would also
be possibleusingEELS data, as long as a small collection
semiangle,B, was used in the microscope,which would
ensure that only dipole-allowed transitions were collected. Under this condition, EELS and XAS are equivalent.
The successof determining Fe3+-Fe2+ratios by fitting
referencespectrain part dependson the constancyofthe
Fe l,rr-edge shapesin different materials and environments. The spectra illustrated in Figures 2 and 3 and
published Fe I,., EELS and XAS spectra(Krishnan, 1990;
Krivanek and Paterson,1990; Kurata et al., 1990; Patersonand Krivanek, 1990;Cresseyet al., 1993)indicate
that the Fe2+and Fe3* Zr., edgesdo not show large deviations from the calculated Fe Lr., edge spectra of van
der Laan and Kirkman (1992), and so Fe3*-Fe2*ratio
determination should be possiblefor compounds containing Fe2* and Fe3* in roughly tetrahedral and octahedral coordination where the Fe to ligand bonds are
dominantly ionic. The crystallinity of the mineral should
also have little effect on the shapeof the Fe l., edges,if
one assumesthat the metal to ligand bonds are dominantly ionic, since the spectrafor Fe2* and Fe3+are dominated by quasi-atomictransitionsthat give rise to the Fe
Lr., edge,with only some modification due to the solid
state environment. Fe3*-Fe2*ratio determinationis expected to be more difficult when a multivalent or multisite compound containssitesthat are highly distorted from
To or Oo symmetry, as in iron phthalocyanine (Koch et
al., 1985),which containsFe in a squareplanar site, or
when the characterofthe ligandsbondedto the Fe causes
the Fe to assumethe low-spin state,as in FeS,(Thole and
van der Laan, 1988)and the iron isocyanates(Kurata et
al., 1990).Although Cresseyet al. (1993)have shownthat
ratio in a numit is possibleto determinethe Fe3+-Fe2+
ber of materials,suchas spinels,amphiboles,Fe-containing glass,and amethyst,this implies that one can determine the Fe3+-Fe2+ratio at the nanometer scale using
EELS, assumingthe energy resolution of the EELS/TEM
setup is -0.5 eV or better. Such an energyresolution is
usually only obtained with an ultra-high vacuum TEM
fittedwith a cold field emissiongun.
Mn. Mn is anotherelementcommon in the geological
environment,and chemicallyit can exhibit a largenumber of oxidation statesand coordinations. In contrast to
Fe, Mn is found in the three oxidation statesII, III, and
IV in nature, although Mn in sedimentary and surficial
635
640
645
650
Energy
Loss(eV)
655
Mn compounds
Fig. 4. Mn Zr., edgesfrom representative
showingthe effectof differentoxidationstateson the shapeand
ganophyllite
energyof the L,, edges.Manganosite1tet14nz*;,
(r6lMnr+),
bixbyite(twodistinct[6rMnr*sites),norrishite(single
t6lMn3+
(t6lMna+
(tullt4no*;.
site),asbolan
), andramsdellite
Mn minerals is more likely to be Mn3* and Mna+. The
coordinationof Mn is almostalwaysoctahedral,although
t41Mn2+
is found in some minerals such as jacobsite,galaxite, and hausmannite.Mn minerals are often finegrained heterogeneous
mixtures containing more than one
oxidation state. That makes the determination of their
and chemoxidation stateby conventionalspectroscopic
ical techniques very difficult and imprecise, as it only
gives a summed averageof the oxidation statesfrom the
bulk sample. The high spatial and energy resolution offered by PEELS in a FEG-STEM is ideal for obtaining
such information.
Figure 4 comparesthe Mn Zr., edgesfrom a number of
Mn-containingminerals.Therearethreeobviouschanges
in the spectra with an increasing oxidation state of the
Mn: a move to higher energylossesof the I, peak maximum (Table 3), a decreasein the L.-L, white line intensity ratio, and a characteristic shape change of the L,,
edges.Manganositeillustratesa typical Mn2* Ir., edge
with an I. peak maximum at ca. 640 eV; many other
high-spinMn2* compoundshave beenstudiedprevious-
418
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
Mn2*
| "i'.
Mn L2,3
Mn L3
Mn2.
Hausmannite
635
640
645
650
(eV)
Energy-Loss
Fig. 5. Mn L, edgesfrom a well-crystallized
specimenof
hausmannite.
The Z, edgeis split into two peaksbecause
of the
(lalMn2+
two Mn sitesin hausmannite
andt6lMnr+
).
ly, and all exhibit the same basic edge shape with little
variation in coordinationnumber or site symmetry(Rask
etal., 19871'
Garvie and Craven, 1993).The Mn l,, edges
from ganophyllite have the same shapeand I, energy as
in manganosite,indicating that the Mn is dominantly
Mn2+, as is commonly assumedfor this mineral (Dunn
et al., 1983).The Mn Ir.. edgesshown hereare typical of
high-spint6lMn2+.Low-spin t6lMn2+is not thought to occur in natural minerals, although it has a characteristic
edgeshape(Crameret al., l99l) that can be usedto distinguish it from high-spin t6lMn2+.Two minerals are illustratedthat contain Mn3*, bixbyite, and the oxybiotite
norrishite (Tyrna and Guggenheim,1991).The Mn Zr.,
edgesin both minerals are very similar in shape,with a
broad I, edgeand a maximum atca.642 eV. In bixbyite
the I, peak maximum is split into two peaks separated
by ca. 0.5 eV, and thereis further evidenceofunresolved
multiplet splitting on the low-energyside of the I, peak.
In norrishite only partially resolved multiplets are observed;this was observedfor severalMn3* Z. edges(Garvie and Craven, 1993)wherethe Mn sits in a Jahn-Teller
distorted environment due to the da electron configuration. The last two examplesin Figure 4 are from the Mna *
mineralsasbolanand ramsdellite(Miura et al., 1990).As
with the Mn2* and Mn3* Ir., edges,the Mna Ir.. edges
have distinct edgeshapes,with an l, peak maximum at
ca. 644 eV. All Mna+ Lr., edgeshave similar shapes,with
a broad L, edgeand a double peaked L, edge.The two
peaks on the Z, edge have been interpreted in terms of
crystal field theory and assignedto the tr, and e, orbitals,
respectively(Garvie and Craven, 1993).Similar assignments were made on the Mn2* I. edge.Interpretation of
the Mn3+ I. edge is more complicatedbecauseof the
tetragonaldistortion ofthe octahedroncausedby the JahnTeller effect, which further splits the tr, and e, orbitals.
The Mn Z,. edge from manganite was assignedorbitals
basedon Do, symmetry in Garvie and Craven (1993).
636
638
640
642
644
Energy
Loss(eV)
646
648
Z, edgesfrom anothersampleofhausFig.6. Representative
from top to bottomwith an
mannite.The .L,edgesarearranged
to an inincreasein intensityof the Mn'* peakcorresponding
in the relativeamountof Mnr+.
crease
The ability of PEELS to observe variations in oxidation state at high spatial resolution was studied in two
samplesof hausmannite.Both samplesoccurredas small,
octahedral crystals of - I mm in size but were from differentlocalities.The first samplegavedistinct and similar
Mn lr.. edgeshapesfrom all the areasstudied under the
microscope.Figure 5 illustratesa typical lr., edgefrom
the first sample. A similar edgeshapehas been recorded
by XANES (Cresseyet al., 1993)and ELNES (Paterson
and Krivanek, 1990).The Mn I, edgefrom hausmannite
can be interpreted by consideration of its crystal structure. Hausmannite can be representedby the formula
r4tMn2+r6tMn3*Oo
with an ideal Mn3+-Mn2+ratiOOf 2:1,
although experimentally synthesizedhausmannite was
found to have Mn3+-Mn2+ratios varying from 2 to 3.5
(Kaczmarekand Wolska, 1993).The L. edgeshowstwo
distinct peaksseparatedby 1.6eV. The first one is at 640
eV, consistentwith Mn'?+,and the secondone is ca. 1.5
eV higherand is from Mn3*. The secondsampleof hausmannite exhibited a range of Mn L.-edge shapes.Representativespectraare illustrated in Figure 6. It is evident
from these spectra that there are large variations in the
Mn3+-Mn2+ratio, rangingfrom almost pure Mn2* (at the
top of Fig. 6) to edgesthat resemble ideal hausmannite
(middle of Fig. 6) to edgesthat are dominated by Mn3t
4t9
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
(bottom of Fig. 6). Thesevariations can occur over distancesas short as 100 nm.
Cr. In nature, Cr occursin two oxidation states,III and
VI. The Cr6+ion occursexclusivelyin the chromateand
dichromate anions,as in crocoite,PbCrOo,and has not
been found in clay minerals. The best-known Cr-bearing
clay is the smectitevolkonskoite, which hasa bright green
color due to the high percentageof Cr residing in the
octahedrallayer (Giiven, 1988).the Cr lr., edgesfor three
minerals are illustrated in Figure 7. The Cr Ir., edges
from chromite show the samemultiplet structure and basic edgeshapesas the theoretical atomic multiplet spectra
of van der Laan and Kirkman (1992) for Cr3* in O, symmetry and 10Dq valuesof 2-2.5 eV. This rangeof l0Dq
valuesis consistent\Miththe crystal field strengthof I6lCr3+
compounds bonded to O that have optically determined
lODq valuesof 2.1-2.5eV (Lever,I 984).The Cr l, edge
for chromite is at 578.6 eV. When one comparesthe Cr
Ir., edgesfrom volkonskoite and chromite, it is evident
that the various featureshave the same shapeand energy
positions.This directly confirms the trivalent nature of
Cr in the volkonskoite. The featureson the I, edgein the
volkonskoite are slightly broader than those from the
chromite, which can possibly be attributed to the more
distorted octahedral environment in the former mineral.
Similar edge shapeshave been recordedpreviously for
the Cr Ir., edgesfrom CrrO, (Kurata et al., 1988; Kdvanek and Paterson,1990).For comparison,the Cr Lr'
edgefrom crocoite is illustrated in Figure 7. This edgeis
consistent with lower energy resolution PEELS work on
KrCrOo (Kurata et al., 1988).The Z, ar.d L2 edgesare
resolved into two separatepeaks, and there is a shift of
the l,, peak maximum to a higher energyloss (Table 3),
consistentwith the higher effectivechargeon the Cr atom.
The I4rCr6+Lr:-edge shape is characteristicof tetrahedrally coordinatedXOX anions(X: Ti, V, Cr, and Mn)
in general(Brydson et al., 1993) and provides a fingerprint for these species.
575
580
585
590
Energy
Loss(eV)
595
Fig. 7. Cr Zr.. edgesfor chromite (r6rcr3+
), volkonskoite
(l6lcrr+
), andcrocoite(tolCru*
).
Severalallotropes of C exist, the most well known being diamond, graphite, and the fullerenessuch as Cuo.
Betweenthe extremesof sp3-bondeddiamond and sp2bonded graphite are a whole range of carbonaceousmaterials that have intermediate bond types, such as glassy
C and evaporatedamorphousC (Robertson,1986;Robertson and O'Rielly, 1987). In minerals, inorganically
Anion and cation coordination fingerprints
bonded C is found almost exclusivelyin the carbonate
Many EELS spectraexhibit a near-edgestructure char- anion, COI-. Its structureconsistsof a central C atom
acteristicof the arrangementand number of atoms in the that is bonded to three planar trigonal O atoms. The
first coordination shell surroundingan atom. This gives COI- anion is isostructuraland isoelectronicwith BOIrise to the concept of a coordination fingerprint (Brydson and NO; anions, which all exhibit similar K ELNES
et a1.,I 988; Saueret al., I 993).A coordinationfingerprint shapes.The C K ELNES from the minerals containing
ariseswhen the excitedatom is surroundedby atomsthat the carbonateanion exhibit an edgeshapeconsistingof
are strong electron backscatterers,such as O'? and F
a sharp initial peak a| 290.2 eV and a broader,less inIn that casethe near-edgestructure is dominated by scat- tense,featurewith a maximum at 301.3 eV (Fig. 8). These
tering eventswithin the first coordination shell surround- features may be attributed to transitions to the unoccuing the atom. The ELNES structure may then be related pied zr* and o* antibonding molecular orbitals of a
to the molecular orbital (MO) structure of the excited COI- cluster, respectively.The C K edgefrom the COIatom and its nearestneighbors(Saueret al., 1993).Some anion is illustrated for calcite, siderite, and two members
examplesof ELNES coordination fingerprints include the of the hydrotalcite group, hydrotalcite and desautelsite
B K edgefrom BO, and BOo (Saueret al., 1993),the S (Fig. 8). The extra structure on the o* peak from calcite
and P K edgesfrom SO?- and POI- anions (Hofer and is probably due to scatteringfrom shells surrounding the
Golob, 1987),the Lr., and,K edgesfrom talAl and t6lAl carbonatecluster. For comparison,Figure 8 illustratesthe
(Brydsonet al., 1989b;Ildefonseet al., 1992),and the K C K edgefor evaporated amorphous C, graphite, and diamond. Their edgesare consistentwith previously pubedgesof t4tMgand I6rMg(Tafto and Zhu, 1982).
420
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
o*
cK
Amorphous
carbon
100
280
290
300
Energy
Loss(eV)
310
Fig.8. C Kedgefrom mineralscontainingthecarbonate
anion (calcite,siderite,desautelsite,
and hydrotalcite)and three
allotropesof C (amorphous
C, graphite,and diamond).
lished XANES and ELNES for these materials (Kincaid
et a1.,1978;Rosenberget al., 1986;Bergeret a1.,1988;
Weng et al., 1989a;Bernatowiczet al., 1990;Batsonand
Bruley, 1991; Katrinak et al., 1992).The fullerenesCuo
and Cro also have C K edge shapessimilar to that from
graphite but with differencesdue to the molecular nature
of the Cuoand Cro molecules(Terminello et al., l99l;
Shinoharaet al., l99l). Amorphous C and graphitehave
basically the same edge shapes,with an initial peak at
285 eV and a secondmore intense feature at ca. 290 eY',
the similarities can be explained in terms of the similarities in their bonding. Diamond has a different edgeshape
from graphite,with a peak maximum at 292.6 eV. The
differencesbetween the C K edgesin graphite and diamond may be explained by the presenceof sp' bonding
in graphite,which resultsin a peak at 285 eV, identified
as transitions to the zr* molecular orbital, and a second
peak at 290 eY, due to transitions to o* orbitals (Weng
et al., 1989a).In diamond the bonding betweenthe C
atoms can be explained in terms of tetrahedrally directed
sp3hybrid orbitals, and the first peak is identified as arising from transitions to molecular orbitals of o* character
(Wenget al., 1989a).
110
130
120
EnergyLoss(eV)
140
Fig. 9. Si Ir., edgesfrom threesilicatescontainingSiO"tetrahedra(qtartz, talc, and nontronite).For comparison,the Si
Z, edgesfrom p-SiCand Si arealsoshown.
The presenceof Si in minerals is almost exclusively
associatedwith the SiOo tetrahedron. These tetrahedra
may join together to form a large number of crystalline
silica polymorphs (Keskar and Chelikowsky, 1992), as
well as amorphous and glassy phases.The silicates are
also characterizedby the SiOo tetrahedron: these tetrahedra are present in simple isolated tetrahedra, as in the
neosilicates,but alsojoin togetherto form a huge number
of structures(Liebau, 1985). A common cation fingerprint observed in minerals is the Si Ir., edge from totSi.
The Si lr., edgesfrom quartz, talc, and nontronite are
illustrated in Figure 9, togetherwith the Si Ir., edgesfrom
Si and SiC. The Si Ir.. edgesfrom quartz and the two
clay minerals exhibit very similar edgeshapes,characterized by two initial sharp peaks,peaksA (106.3) and B
(108.4 eV), followed by a sharp peak C (115.0)and a
broad peak D (ca. 133 eV). Peak A, from q:uarlz,is split
into two peaksseparatedby 0.45 eV; a similar splitting
has been recorded by high-resolution XANES (Li et al.,
1993) and EELS (Garvie, unpublished data). The observed structures have been assignedto the unoccupied
molecular orbitals of a SiOX- tetrahedron (McComb et
al., l99l;, Li et al., 1993).Thus basedon the molecular
orbital assignmentsin McComb et al. (1991) and Li et
al. (1993), the four peaks from the quartz Si Zr., edge
have been assignedto the following unoccupied molec-
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
ular orbitals: peak A : a, (sJike), peak B : t, (pJike),
peak C : primarily e (d-like), and peak D : tz (dJike).
This Si L,r-edge shapeis seenin the majority of silicates,
although a recent study of the neosilicates(McComb et
al., 1991, 1992)illustratesthat the fingerprinttechnique
must be used with care. Relative to quartz, the Si I*
ELNES from zircon (McComb et al., 1992) shows a distinctly different edgeshape,which is related to the highly
distorted SiOi tetrahedra. Octahedrally coordinated Si,
which rarely occursin minerals,is presentin stishovite.
The Si l,,., absorptionedgefrom stishovite(Li et al., 1993)
is considerablydifferent in shapefrom the Si Ir,3 ELNES
in minerals containing regular or distorted SiOotetrahedra, which implies that the edgeshapemay be used as a
guide to the coordination of the Si. The different Si [.,
ELNES from minerals containingregularor distorted SiOo
tetrahedra could possibly be used as a guide to the site
symmetry of the Si, although further experimental work
will be neededto confirm this. For comparison,the Lr.
edgesfrom SiC and Si are illustrated in Figure 9. The
differencesin the edge shapesbetween the Si Ir,. edges
from the silicatesand the Si and SiC can be understood
in terms of the differing unoccupied densities of states
(Weng et al., 1989b, 1990), whereasthe different positions ofthe edgeonsetsare relatedto the differing effective chargeson the Si atoms in the different materials.
Mounr-rNc
coRE-r-ross EDGES
A host of theoretical techniques have been used to
model EELScore-lossedges,which can be divided roughly into two types(de Groot et al., 1993):the local density
approximation (LDA) and the ligand field multiplet (LFM)
approaches.The former includesmethods such as band
structure, multiple scattering,and molecular orbital calculations(seeBrydson, 199I , for an introduction to these
methods as applied to ELNES data). These are suitable
for weakly correlatedsystemswhere the interaetion of the
excited electron with the core hole created by the excitation processis small. The LFM method explicitly takes
accountof electron-electron
and electron-corehole interactionsin stronglycorrelatedsystemsand has beenmost
successfullyused to model the 3d transition metal lr.,
edges(de Groot et al., 1990a, 1990b;van der Laan and
Kirkman, 1992;Cresseyet al., 1993)and M,., edges(van
der Laan, l99l). Two contrastingmethodsfor modeling
the different core-loss edgesin MnO are illustrated: the
multiple scattering method to model the O K edge and
the atomic multiplet method for the Mn Zr., edge.
Multiple scattering calculations
The multiple scattering calculations were performed
using the ICXANES computer code of Vvedenskyet al.
(1986) using full multiple scattering.Multiple scattering
theory is basedon the interferencebetweenthe outgoing
ejected electron wave and the returning backscattered
electronwave. Briefly, the transition rate ofexcitation is
calculatedwithin the dipole approximation,using essentially an atomic contribution modified by multiple elastic
530
540
560
550
EnergyLoss(eV)
421
570
Fig. 10. Experimental
O K edgefrom MnO comparedwith
the resultsof rhe multiple scatteringcalculationsfor differenr
clustersizes.The resultsfor shells3 and 5 arenot illustrated,as
theyhad little effecton the final plot.
scattering of the excited electron by the surrounding atoms. This approach,while formally equivalentto a Korringa-Kohn-Rostockerband structurecalculation,is considerablymore flexible,allowing the inclusionof excited
atom potentialsto accountfor the effectsofthe excitation
process.The scatteringpropertiesof the various atoms
are described by phase shifts that may be calculated by
numerical integration of the Schrddinger equation, assuming a muffin-tin form for the atomic potentials in the
crystal, obtained by the simple superpositionof neutral
atomic chargedensities.The multiple scatteringcalculation is performed in real space by dividing the cluster
into shellsof atoms surroundingthe centralexcitedatom.
The multiple scatteringis solved within eachshell in turn,
the final
and the resultsare then combinedto reassemble
cluster, use being made of any structural symmetry.
Many studieshave illustrated the applicability of the
multiple scattering approach in modeling ELNES and
XANES spectra(Oizumi et al., 1985;Lindner et al., 1986;
Brydsonet al., 1988, 1992b;McComb et al., 1992;Sipr,
1992;Kurata et al., 1993).The calculationsfor the O K
edgein MnO (Fig. l0) were performedfor sevenshells,
extendingto a cluster radius of I 1.86 au (where I au :
0.52918 A) around an O atom. The core hole was accounted for by the use of the (Z + l)* approximation for
422
Exoerimental
Mn L3edge
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
The sharp intensepeak c arisesbecauseof intrashell multiple scattering within the first O coordination shell and
reflects transitions to unoccupied O 2p stateshybridized
with the more delocalizedMn 4s and 4p states.Peaksd
and f are adequately reproduced after the addition of a
third O coordination shell (shell 6) and arise from mainly
intershell multiple scattering.Finally, the largebroad peak
e is modeledusingthe first O coordinationshelland may
be adequately reproduced using solely single scattering.
Such a basic resonantfeature is known as a multiple scattering resonance,and its energy position follows the relationship E e t/nz(where ,E is the energy of the feature
above the edgeonset and R is the O-O distance).Thus
such multiple scattering resonancesprovide a means of
determining bond lengths and bond length variations
(Kurata et al., 1993).Multiple scatteringresonancesarise
since certain ions (notably the oxide anion) are relatively
strong backscatterersand give rise to a potential barrier
for electron scattering.
Atomic multiplet calculations
The 3d transition metal L23edgescannot be modeled
in terms of a one particle density of statesbecauseof the
636
638
640
642
644
646
large electron correlation effectsin the partially filled 3d
Energy
Loss(eV)
orbitals and the strong interaction with the 2p core hole
Fig. I l. Comparisonof the Mn L, edgefrom manganosite createdduring the excitation process.The LFM approach
(r61Mn'?+
Mn d5atomicmultipletspectrafor has been usedto model the Ir., edgesin a number of 3d
) with the calculated
Mn2+in O, symmetrywith crystalfield strengths
of 0.5-2.0eV. transition metalsand compounds(de Groot et al., 1990a,
1990b;van der Laan and Kirkman, 1992).The theoretical Mn2* Lr., data presentedhere were calculatedby van
the central O atom in the cluster. In agreementwith the der Laan and Kirkman (1992\ and take account of the
findingsofLindner et al. (1986)and Rez et al. (1991)on following effects in order to calculate the 2p63d' MgO, the dominant scatterersin oxide structuresare the 2psJfl'+t dipole transitions:2p-3d and 3d-3d Coulomb
O atoms, and it is possibleto reproducemany of the and exchangeinteractions, 2p and 3d spin-orbit interacmajor featuresusinga clusterthat ignoresthe Mn atoms. tions, and the crystal field acting on the 3d states.AlThe following cluster surrounding a central O atom was though transitions2p - 4s are dipole allowed,they have
used:shell I : six Mn atomsat 4.19 au, shell2 : 12 O been neglectedin the calculationsbecausethe 2p - 36
atoms at 5.93 au, shell 3 : eight Mn atoms at 7.26 au, transitionsdominateover the transitionsto 4s states.The
shell4 : six O atomsat 8.38au, shell5 :24 Mn atoms 3d-3d and 2p-3d two particle interactions define the
aI 9.37 au, shell6:24 O atomsat 10.21au, and shell7 ground stateof the transition metal ion and split the EELS: 12 O atoms at 11.86 au. By varying the cluster size, XAS final stateinto a largenumber of configurations.The
the featuresin the experimental O Kedge can be assigned calculatedatomic multiplet spectrum for the 3d" comto specific scatteringevents. The six main featuresof the pounds (r > 0) typically consistsof hundredsof unreexperimental O K edge, labeled a-f, are successfullyre- solvablefinal statesspreadover typically 15 eV for the
producedby the multiple scatteringcalculation(Fig. 10). Mn Ir., edges.In order to compare the multiplet spectra
The first peak, a, can be adequately reproduced using with experimentalones,the multiplet linesare broadened
solely the first coordination shell (shell l) of Mn. The to take into account effects such as lifetime and vibradominant scatteringevents are d-like scatteringfrom Mn tional broadening and hybridization.
The validity of the atomic multiplet method to model
and pJike scattering from O. Hence this feature reflects
transitionsto unoccupiedO 2p stateshybridizedwith the the experimental Lr., edge is illustrated in Figure I l. In
narrow metal 3d band that is highly localized around the this figure the theoretical multiplet spectra for Mn2+ in
metal atom sites. This region therefore reflects the co- O, symmetry with crystal field valuesof 0.5, l, 1.5, and
valencein the transition metal oxides. This is in agree- 2 eV are comparedwith the Mn I, edgefrom MnO. The
ment with the previous work on the O K edgesof 3d agreementbetweenthe experimental L, edgeand the thetransitionmetal compounds(de Groot et al., 1989;Kura- oretical L, edgewith a crystal field value of I eV is exIa eI al.,1993). Peak b appearsto arise only after inclu- cellent. All the featureson the experimental edgeare resion ofa third O shell(sixth coordinationshell)indicating produced in the theoretical spectrum. This crystal field
that this region is also sensitive to longer range effects. value of ca. I eV, determinedby comparingthe experi-
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
423
oxidamental L, edgewith the theoretical ones, is also in good mobile underthe electronbeam,and subsequent
agreementwith the l0Dq of 1.2 eY (Pratt and Coehlo, tion of the metal.
1959)determinedfor MnO by optical spectroscopy.
The
good agreementbetweenthe theoretical and experimental
Acxxowr-nncMENTs
Mn I. edge for MnO is in part due to the dominantly
We are all g.ratefulto the following for providing samples:G.C. Allen
ionic bonding between Mn and O. For more covalent (Interface Analysis Centre, University of Bristol, U.K.) for the synthetic
compoundssuchas Mn-O bonding in pyrolusite,the ex- jacobsite; D.C. Bain (Macaulay Land Use ResearchInstitute, U.K.) for
perimental and theoretical edges show quite different the samplesof volkonskoite and daphnite; J. Faithfull (Hunterian Mustructures(Garvie and Craven, 1993, 1994).Most theo- seum, U.K ) for samplesof ganophyllite (M9609), hausmannite(M8097),
(Royal Veterinary and Agricultural
reticalatomic multiplet spectracalculationsassumesole- and silicon carbide; H.C.B. Hansen
University, Denmark) for the synthetic desautelsite;C.M.B. Henderson
ly ionic-type bonding, although introducing terms in the (Department of Geology, University of Manchester,U.K.) for the syncalculationsfor covalencecan be used successfullyto thetic iron analogueof leucite; S. Guggenheim(University of Illinois at
model more covalent compounds (van der Laan et al., Chicago)for the sample of norrishite; and H. Miura (Department of Geology and Mineralogy, Hokkaido University, Japan) for the ramsdellite.
1986),but this is more difficult.
Errncrs
TNDUCEDBy rHE ELEcTRoN BEAM
In minerals, the main effectsof electron beam irradiation are elementmobility and amorphization(Champnessand Devenish, 1992).Theseeffectsare most notable
during EDX in a TEM, whereelementmobility and loss
is often a major problem, particularly with the long
countingtimes requiredfor EDX. Theseproblemscan be
especiallyacute for the clay minerals (van der Pluijm et
a1.,1988),which often contain mobile elementssuch as
K, as well as HrO and OH. In many instancesthe effects
induced by the electron beam changethe nature of the
solid and thereforealso the environment of specificatoms. Thesechangescan be followed using an atom-specific probe such as PEELS. PEELS has been used to follow the changesoccurring during the electron irradiation
of solids, such as the transformation of calcite to CaO
(Wallsand Tenc6,1989;Murookaand Walls, 1991)and
the transformationof BOoto BO. in hydroxyborates(Sauer
et al., 1993).Egertonet al. (1987)provided a good overview of PEELS-relatedstudiesof chemical changesinduced by the electron beam. In materials containing 3d
transition elements, the changesin the edge shape are
consistentwith changesin the valency of the transition
metal atom. This hasbeenstudiedin detail in the mineral
asbolan(Garvieand Craven, 1994),whereit waspossible
to follow the reductionof the Mn4* to Mn2+ with intense
electron beam irradiation. During irradiation the Mn .L..,
edgechangesits shapeand position, consistentwith increasingproportionsof Mnr+ and finally Mn'?+.
Although the majority of materials containing Mn in
the oxidation statesofIII or IV werefound to be reduced
to lower oxidation states,the opposite has occasionally
been observed.In sussexite,Mn2+BO,(OH), and ganophyllite, the Z, edgeinitially changedin shapeand energy
from an Mn2+ to an Mn3* I, edge.Further irradiation
causedthe Mn3* edge to revert back to an Mn2* edge.
Similar behavior was observedin all the phyllosilicates
that containedFe2*. After a short period of irradiation,
the Fe2+component was completelytransformedto an
The
L, edgehaving the sameenergyand shapeas I61Fe3+.
oxidation of the Fe'7*in the phyllosilicatesand the Mn'?+
in sussexiteand ganophylliteis presumablylinked to the
loss of H in the OH anions,which is expectedto be very
We are grateful to G. van der laan (SERC Daresbury Laboratory, U.K.)
for the Mn lr 3multiplet data The authors would like to thank the SERC
for providing the Gatan 666 PEELS and for a postdoctoral research assistantship(L.A.J G.) under grant GWGI2832.
RrnnnBNcns crrED
Bailey, S.W. (1988) Structures and compositions of other trioctahedral
l:l phyllosilicates.Mineralogical Society of America Reviews in Mineralogy,19, 169- 188.
Batson,P.8., and Bruley,J. (1991)Dynamic screeningofthe coreexciton
by swift electronsin electron-energyJossscattering.Physical Review
l€tters, 67, 350-353
Berger,S.D., McKenzie,D.R., and Martin, P J. (1988)EELS analysisof
vacuum arc-deposited diamond-like films. Philosophical Magazine
Irtters. 57. 285-290.
Bernatowicz,T.J., Cibbons, P.C., and kwis, R.S. (1990) Electron energy
loss spectrometryof interstellar diamonds The AstrophysicalJournal,
3s9,246-255
Bonnin, D., Calas,G., Suquet,H , and Pezerat,H. (1985)Sitesoccupancy
of Fer* in Garfield nontronite: A spectroscopicstudy. Physics and
Chemistryof Minerals, 12,55-64.
Brydson, R. (1991) Interpretation of near-edgestructure in the electron
energyJossspectrum Bulletin of the Electron Microscopy Society of
America. 21. 57-67.
Brydson,R , Williams,B.G., Engel,W., Sauer,H,Zf-itler, E., and Thomas, J.M. (1987) Electron energyJossspectroscopy(EELS) and the electronic structure of titanium dioxide. Solid State Communications, 64,
609-6r2.
Brydson,R., Williams, B.G., Engel,W., Lindner, Th, Muhler, M, Schliigl,
R, Zeitler, E., and Thomas, J.M. (1988) Electron energy-losssp€ctroscopy and the crystal chemistry of rhodizite. II. Near-edgestructure.
Journal ofthe Chemical Society,FaradayTransactionsl, 84, 631-646.
Brydson, R., Sauer,H., Engel,W., Thomas, J M., Znitler, E., Kosugi, N.,
and Kuroda, H. ( I 989a)Electron energylossand x-ray absorptionspectroscopy of rutile and anatase:A test of structural sensitivity. Joumal
ofPhysics:CondensedMatter, l, 797-812.
Brydson,R., Sauer,H., Engel,W., Thomas,J.M., and Zeitler,E. (1989b)
Co-ordination fingerprints in electron loss near-edgestructures:Determination of the local site symmetry of aluminium and beryllium in
ultrafine minerals. Journal of the Chemical Society, Chemical Comm u n i c a t i o n sI.5 . l 0 l 0 - l 0 l l .
Brydson, R., Sauer,H , and Engel,W . (1992a)Electron energyloss nearedge structure as an analytical tool: The study of minerals. In M.M.
Disko. C.C. Ahn. and B. Fultz, Eds.,Transmissionelectronenergyloss
spectrometry in materials science,p. l3l-154 The Minerals, Metals
and Materials Society,Warrendale,Illinois.
Brydson, R., Sauer,H., Engel,W., and Hofer, F. (1992b) Electron energyloss near-edgestructuresat the oxygenK-edgesoftitanium (IV) oxygen
compounds.Journal of Physics:CondensedMatter, 4, 3429-3437.
Brydson,R., Garvie, L.A.J , and Craven,A.J. (1993)Zr,-edgesof tetrahedrally co-ordinated do transition metal oxoanions XOI . Journal of
Physics:CondensedMatter, 5, 9379-9392.
Buechner, U. (1975) The dielectric function of mica and quartz deter-
424
GARVIE ET AL.: ENERGY LOSS NEAR-EDGE FINE STRUCTURE
mined by electronenergylosses.Journal ofPhysics C: Solid StatePhysics, 8, 2781-2787.
Buseck,P.R., and Sel{ P (1992) Electron energyloss spectroscopy(EELS)
and electron channelling(ALCHEMI). Mineralogical Societyof America Reviewsin Mineralogy,27,l4l-180.
Calas,G. (1988) Electron paramagneticresonance.Mineralogical Society
of America Reviewsin Mineralogy,18, 513-571.
Calas, G., and Petiau, J. (1983) Coordination of iron in oxide glasses
through high resolution K-edgespectra:Information from the pre-edge.
Solid State Communications- 48. 625-629.
Cardile,C.M., and Brown, IW.M. (1988)An 5?FeMdssbauerspectroscopic and X-ray diffraction study of New Zealand glauconite. Clay
Minerals. 23. 13-25.
Champness,P.E., and Devenish, R.W (1992) Radiation damagein silicate minerals: Implications for AEM. In A. I-6pez-Galindo and M.I.
Rodriguez-Garcia,Eds., Electron microscopy 92, vol. II: Materials sciences,p. 541-545. Secretariadode Publicacionesde la Universidad de
Granada, Granada, Spain.
Colliex, C., Manoubi, T., Gasgnier, M., and Brown, L.M. (1985) Near
edgefine structureson electron energyloss spectroscopycoreJossedges ScanningElectron Microscopy, Il,489-512
Cramer, S.P.,de Groot, F.M F., Ma, Y, Chen, C.T., Sette,F., Kipke,
C.A., Eichhorn,D M., Chan,M.K., Amstrong, W.H., Libby, E., Christou, G., Brooker,S., McKee,V , Mullins, O.C , and Fuggle,J.C. (1991)
Ligand field strengths and oxidation states from manganeseZ-edge
spectroscopy.Journal of the American Chemical Society, I 13, 79377940.
Cressey,G., Henderson,C.M.B., and van der l-aan, G (1993) Use of
Z-edge X-ray absorption spectroscopyto characterisemultiple valence
statesof 3d transition metals: A new probe for mineralogicaland geochemical research.Physicsand Chemistry of Mineral, 20, I I I - I 19.
de Groot, F.M.F, Grioni, M., Fuggle,J.C., Ghijsen,J., Sawatzky,G.A.,
and Petersen,H. (1989) Oxygen ls x-ray-absorption edgesof transition-metal oxides. Physical Review, B40, 5715-5723.
de Groot, F.M F, Fuggle,J.C.,Thole, B.T., and Sawatzky,G.A. (1990a)
L,. x-ray-absorption
edgesofd0 compounds:K*, Ca'z*,Sc3*,and Ti4*
in O, (octahedral)symmetry. Physical Review, B4l, 928-937.
-(1990b)
2p x-rcy absorption of 3d transition-metal compounds:
An atomic multiplet description including the crystal field. PhysicsReview, B42, 5459-5468.
de Groot, F.M F, Abbate, M., van Elp, J., Sawatzky,G.A., Ma, Y.J.,
Chen, C.T., and Sette, F. (1993) Oxygen ls and cobalt 2p x-ruy absorption of cobalt oxides. Journal of Physics: CondensedMatter, 5,
2277-2288.
Dunn, P.J., Peacor,D.R., Nelen, J.E., and Ramik, R.A. (1983) Ganophyllite from Franklin, New Jersey;Pajsberg,Sweden;and Wales:New
chemical data. Mineralogical Magazine,47, 563-566.
Egerton, R.F (1986) Electron energy-lossspectroscopyin the electron
microscope,410 p. Plenum, New York
Egerton,R.F, Crozier, P.A., and Rice, P. (1987) ElectronenergyJoss
spectroscopyand chemical change.Ultramicroscopy, 23, 305-312.
Garvie, L.A J., and Craven, A.J (1993) Determination of the crystal field
strengthin manganesecompoundsby parallel electronenergylossspectroscopy. In A J. Craven, Ed., Proceedingsof the Institute of Physics
Electron Microscopy and Analysis Group, vol. 138, p. 3l-34. Institute
ofPhysics ConferenceSeries,Bristol, U.K
-(1994)
Electron beam induced reduction of Mna* in manganese
oxides as revealedby parallel EELS. Ultramicroscopy, in press.
Goodman, B.A., and Bain, D.C (1979) Miissbauer spectra of chlorites
and their decompositionproducts.Proceedingsofthe InternationalClay
Conference1978,p. 65-74.
Gtiven, N. (1988) Smectites.Mineralogical Society of America Reviews
in Mineralogy.19. 49'7-559.
Hawthorne, F C (1988) Miissbauer spectroscopy.Mineralogical Society
of America Reviewsin Mineralogy,18,255-340.
Hofer, F., and Golob, P. (1987) New examplesfrom near-edgefine structures in electron energyloss spectroscopy Ultramicroscopy,21,379384.
Ildefonse, P., Calas, G., Kirkpatrick, R.J., Montez, B., Flank, A.M, and
Lagarde, P. (1992) Local environment of aluminium in amorphous
alumino-silicatesby XANES and MAS NMR. In Y.K. Klaraka and
A.S. Maest, Eds., Proceedingsofthe 7th International Symposium on
Water-Rock Interaction, p 153-158. Balkema, Rotterdam, Germany.
Joy, D.C. (1993) Electron specimeninteractions.In A G Fitzgerald,B.E.
Storey,and D. Fabian, Eds., Quantitative microbeam analysis,p. 213245 SUSSPand IOP, Edinburgh and I-ondon.
Kaczmarek,J., and Wolska, E. (1993) Cation and vacancydistribution in
nonstoichiometric hausmanite.Journal of Solid State Chemistry, 103,
387-393.
Katrinak, K.A., Rez, P., and Buseck,P.R. (1992) Structural variations in
individual carbonaceousparticlesfrom an urban aerosol.Environmental Scienceand Technology,26,1967-1976.
Keskar, N.R., and Chelikowsky, J.R. (1992) Structural propenies ofnine
silica polymorphs. Physical Review B, 46, l-13.
Kincaid, B.M., Meixner,A.E., and Platzman,P.M. (1978)CarbonKedge
spectroscopy.Physical
in graphite measuredusing electron-energyJoss
Review lrtters, 40, 1296- 1299.
Koch, E.E., Juenet,Y., and Himpsel, FJ. (1985) High-resolutionsoft
x-ray excitation spectraof 3d-metal phthalocyanines.Chemical Physics
I€tte$.116.7-11.
Koster, A.S., and Mendel, H. (1970) X-ray KB emission spectraand energy levels ofcompounds of3d-transition metals: I Oxides Journal of
Physicsand ChemistryofSolids, 3l,25ll-2522.
Knshnan, K.M. (1990) Iron 2,, near-edgefine structurestudies Ultramicroscopy,32,309-311.
Krivanek, O.L. ( I 989) Improved parallel-detectionelectron-energy-loss
spectrometer Ultramicroscopy, 28, I 18-121.
Krivanek, O.J., and Paterson,J H (1990) ELNES of3d transition-metal
oxides. I. Variations across the periodic table. Ultramicroscopy, 32,
313-318.
Krivanek, O.L., Ahn, C.C., and Keeney, R B (1987) Parallel detection
electron spectrometerusing quadrupole lenses.Ultramicroscopy, 22,
1 0 3 - l1 6 .
Kurata, H., Ishizuka, K., and Kobayashi, T. (1988) Near edge structure
in electron energyloss spectraof chromium trioxide intercalated into
graphiteand somechromium oxides.Bulletin of the Institute of Chemical Research,Kyoto University, 66, 572-57 9.
Kurata, H., Nagai, K., Isoda, S., and Kobayashi,T. (1990) ELNES of iron
compounds. In L.D. Peacheyand D.B. Williams, Eds., Proceedingsof
the XIIth Intemational Congressfor Electron Microscopy, vol. 2: Analytical Sciences,p- 28-29. San Francisco Press,San Francisco,California.
Kurata, H., LefBvre,E., Colliex, C., and Brydson, R. (1993) ELNES structures in the oxygeniK-edgespectraoftransition metal oxides. Physical
Review. 847. 13763-13768
I-ever, A.B P (1984) Inorganic electronic spectroscopy:Studiesin physical and theoretical chemistry, vol. 33 (2nd edition), 863 p. Elsevier,
Amsterdam
Li, D., Bancroft,G M., Kasrai, M., Fleet,M.E., Feng,X H., Tan, K.H.,
and Yang, B.X. (1993) High-resolution Si K- and Z,r-edge XANES of
a-quartz and stishovite. Solid State Communications, 87, 613-617 .
Liebau, F. (1985) Structural chemistry ofthe silicates:Structure,bonding,
and classification,347 p. Springer-Verlag,Berlin.
Lindner, Th., Sauer,H., Engel,W., and Kambe, K. (1986) Near-edge
structure in electron-energyloss spectra of MgO. Physical Review B,
33,22-24.
Manceau,A , Gorshkov, A.I., and Drits, V.A. (1992)Structural chemistry
of Mn, Fe, Co, and Ni in manganesehydrous oxides. L lnformation
from XANES spectroscopy.American Mineralogist, 77, 1133-1143.
Marusak, L.A., Messier, R., and White, W.B. (1980) Optical absorption
spectrum of hematite, a-Fe,O, near Ir to UV. Journal of Physicsand
Chemistryof Solids,41, 981-984.
McComb, D.W , Hansen,P L., and Brydson,R. (1991)A studyof silicon
ELNES in neosilicates.Microscopy, Microanalysis,Microstructures,2,
5 6l - 5 6 E .
McComb,D.W., Brydson,R., Hansen,P.L., and Payne,R.S.(1992)Qualitative interpretation ofelectron energy-lossnear-edgestructurein natural zircon. Journal of Physics:CondensedNfalter, 4, 8363-8374.
McGibbon, A.J. (1991) The application of PEELS on a STEM to the
problem ofinert gas bubbles in solids. In F.J Humphreys, Ed., Pro-
GARVIE ET AL.: ENERGY LOSS NEAR.EDGE FINE STRUCTURE
ceedingsof the Institute of PhysicsElectron Microscopy and Analysis
Group conference,p. 109-1 12. Institute of PhysicsConferenceSeries
no 119, Institute of Physics,Bristol.
Miura, H., Kudou, H., Choi, J H , and Hariya, Y. (1990) The crystal
structure of ramsdellite from Pirika mine. Joumal of the Faculty of
Science,Hokkaido University, SeriesIV, 22, 6ll-617.
Murooka, Y., and Walls, M.G. (1991)Beam damageon anisotropicmaterials (CaCO,) in STEM. In F.J Humphreys, Ed., Proceedingsof the
Institute of PhysicsElectron Microscopy and Analysis Group conference,p. 337-340. Institute ofPhysics ConferenceSeriesno. ll9, Institute of Physics,Bristol.
Oizumi, H., Fujikawa,T , Ohashi,M., Maezawa,H., and Nakai, S. (1985)
F-K XANES studiesofalkaline-earth fluorides.Journal ofthe Physical
Societyof Japan,54, 4027-4033.
Otten. M.T.. and Buseck.P R. (1987) The oxidation state of Ti in hornblende and biotite determined by electron energy loss spectroscopy,
with inferencesregardingthe Ti substitution Physics and Chemistry
of Minerals. 14. 45-51
Otten, M.T., Miner, B., Rask,J.H , and Buseck,P R (1985)The determination of Ti, Mn and Fe oxidation states in minerals by electron
energy-lossspectroscopy.Ultramicroscopy, 18, 285-290.
Paterson,J.H.. and Krivanek, O J. (1990) ELNES of 3d transition-metal
oxides. II. Variations with oxidation state and crystal structure. Ultramicroscopy, 32, 3 19-325
Pratt, G.W.J., and Coehlo, R. (1959)Optical absorptionof CoO and MnO
above and below the Neel temperature.PhysicsReview, I I 6, 28 I -286.
Rao, C N R., Sarma,D.D., Vasudevan,S , and Hegde,M.S. (1979)Study
of transition metal oxides by photoelectronspectroscopyProceedings
ofthe Royal SocietyLondon, 4367,239-252.
Rask, J H , Miner, B A., and Buseck,P.R. (1987) Delermination of manganeseoxidation statesin solids by electron energyJossspectroscopy.
Ultramicroscopy
, 21, 321-326.
Rez, P., Weng, X, and Ma, H. (1991) The interpretationof near edge
structure. Microscopy, Microanalysis, Microstructures, 2, | 43-l 5 l.
Robertson,J. (l 986) Amorphous carbon Advances in Physics,35, 3 17374.
Robertson, J., and O'Rielly, E.P. (1987) Electronic and atomic structure
ofamorphous carbon. Physical Review, B35, 2946-2957.
Rosenberg,R.A., Love, P.J., and Rehn, V. (1986) Polarization-dependent
C(K) near-edgex-ray-absorptionfine structureofgraphite. PhysicalReview B. 33. 4034-4037
Rossman, G.R. (1988) Optical spectroscopy.Mineralogical Society of
America Reviewsin Mineralogy,18,2O7-254.
Sauer,H., Brydson,R., Rowley, P N , Engel,W., and Thomas,J.M. (1993)
Determination of coordinations and coordination-specificsite occupanciesby electron energyJossspectroscopy:An investigation of boron-oxygencompounds. Ultramicroscopy, 49, 198-209.
Shinohara,H., Sato, H., Saito, Y., Tohji, K., Matsuoka, I., and Udagawa,
Y ( I 99 I ) Carbon r(-edgeX-ray absorptionnear-edgestructuresof solid
C,n.ChemicalPhysicsI-etters,183, 145-148.
Sipr, O. (1992) Real-spacemultiple-scatteringanalysisof x-ray absorption
near-edgeK spectraof Cu,O and CuO. Journal of Physics:Condensed
Matter. 4. 9389-9400
'fafts,
J., and Zhu, J. (l 982) ELNES, a potential techniquein the studies
of focal atomic arrangements.Ultramicroscopy, 9, 349-354.
425
Terminello,L.J., Shuh,D.K., Himpsel,F.J.,Lapiano-Smith,D.A , Stdhr,
J , Bethune,D.S., and Meijer, G. (1991)Unfilled orbitals of Cuoand
C," from carbon K-shell X-ray absorption fine structure. Chemical
PhysicsIrtters, 182,49 l-496.
Thole. B.T., and van der l-aan, G. (1988) Branching ratio in x-ray abPhysicalReview,838, 3158-3171.
sorptionspectroscopy.
Tyrna, P.L., and Guggenheim,S. (1991) The crystal structure of nomshite, KIiMnl*Si,O,.: An oxygen-richmica American Mineralogist, 76,
2 6 6 - 2 7| .
van der laan, G. (1991) M23 absorption spectroscopyof 3d transirionmetal compounds. Journal of Physics: CondensedMatter, 3, 14437454.
van der Laan. G., and I(irkman, I.W. (1992) The 2p absorption spectra
of 3d transition metal compounds in tetrahedral and octahedralsymmetry. Journal ofPhysics: CondensedMatter,4, 4189-4204.
van der Laan, G.,Zaanen, J., Sawatzky,G.A., Karnatak, R., and Esteva,
J.-M. (1986) Comparisonof x-ray absorption with x-ray photoemission
ofnickel dihalidesand NiO. PhysicalReview,B33, 4253-4263.
van der l-aan, G., Pattrick, R.A.D, Henderson, C.M B., and Vaughan,
D.J.,1992]' Oxidation state variations in copper minerals studied with
Cu 2P X-ray absorption spectroscopy.Journal of Physicsand Chemistry ofSolids, 53, I 185-l 190.
van der Pluijm, B.A., I-ee,J.H., and Peacor,D.R. (1988) Anall'tical electron microscopy and the problem of potassium diffusion. Clays and
Clay Minerals,16. 498-504
Vvedensky, D.D., Saldin, D.K., and Pendry, J.B. (1986) An update of
DLXANES, the calculation of X-ray absorption near-edgestructure.
Computer PhysicsCommunications, 40, 421-440.
Walls, M G., and Tenc6, M.G. (1989) EELS study of beam-induceddecomposition of calcite in the STEM. In P J. Goodhew and H Y. Elder,
Eds., Proceedingsof the Institute of PhysicsElectron Microscopy and
Analysis Group and Royal Microscopical Society Conference,p. 255258. Institute ofPhysics ConferenceSeriesno 98, Institute ofPhysics,
Bristol.
Weng,X., Rez, P, and Ma, H (1989a)Carbon K-shellnear-edgestructure: Multiple scattering and band-theory calculations. Physical Review.840. 4175-4178.
Weng, X , Rez, P., and Sankey,O.F (1989b) Pseudo-atomic-orbitalband
theory applied to electron-energy-lossnear-edgestructures. Physical
Review, B40, 5694-5704.
Weng, X., Rez, P., and Batson, P E. (1990) Single electron calculations
for the Si L.. near edge structure. Solid State Communications, 74,
l0l3-1015.
Wood, P R., and Urch, D.S. (1976) Valence-bandX-ray emission spectra
of manganesein various oxidation states.Journal of the Chemical Society, Dalton Transactions,P art 3, 2472-247 6.
Yang, K., and Seccombe,P K. (1993) Chemical variation of chromite in
the ultramafic cumulatesof the Great SerpentineBelt, Upper Bingara
to Doonba, New South Wales, Australia. Canadian Mineralogist, 31,
7 5-87 .
MaNuscnrm REcErwDJut-v 2, 1993
Meroscnrvr AccEr'rEDDpcevssr 29, 1993