SECONDARY ION MASS SPECTROMETRY (SIMS) AND ITS
APPLICATION
TO CHEMICAL
WEATHERING
William Shotyk
GeologicalInstitute
Universityof Berne
Jim B. Metson
Berne, Switzerland
Auckland, New Zealand
Department of Chemistry
Universityof Auckland
Abstract. Secondaryion mass spectrometry(SIMS)
is the massspectrometryof atomic specieswhich are
emitted when a solid surface is bombardedby an
energeticprimary ion beam. By continuallybombarding the surface of the sample with the ion beam, the
atomsmakingup the materialbeingstudiedare sputtered away. The secondaryions emitted from the surface are continually analyzed and their intensities recorded over time. The secondary ion intensities are
proportional to the concentration of elements in the
sample,thereby producinga semiquantitativeconcen-
tration depth profile. The depth profile provides an
illustrationof the chemicalcompositionof a sampleas
a function of depth through the surface. The SIMS
techniquehas been appliedto a wide variety of surface
analyticalproblemsand can easily be used to analyze
reacted glass and mineral surfaces which have been
exposed to weathering solutions. Traditional experimental studies of chemical weathering were based
primarily on the analysesof aqueous solutionsgenerated during leaching experiments. Such studies have
provided valuable information concerning rates and
stoichiometry of mineral dissolution reactions but
have led to some confusion and much speculation
regardingthe mechanismsof surfaceprocesses.SIMS
analysesof the surfacesof dissolvingglassesand plagioclase feldspars have recently been used to help
resolve a number of unansweredquestions.For example, SIMS analysesof dissolvingfeldsparshave shown
how the chemical compositionof reacted surfacesand
depth of attack vary, dependingon the compositionof
the mineral, the p H of the leaching solution, and the
presenceof dissolvedsalts and complex-formingorganic ligands.
1. INTRODUCTION: SIMS, CHEMICAL
WEATHERING, AND SOILS
For example,considera typical SIMS depth profile of
fresh (unweathered)labradorite (An54)feldspar with
the followingbulk composition(wt %): SiO2, 55.3! -+
0.21; A1203, 28.84 _ 0.20; CaO, 10.95 +- 0.10; Na20,
1.1.
SIMS
4.72 +_0.07; K20, 0.19 +_0.02. Shown on the ordinate
Secondary ion mass spectrometry (SIMS) is the
in Figure l a is the logarithm of secondary ion intenmassspectrometryof atomic specieswhich are emitted when a solid surfaceis bombardedby an energetic sity, which is proportionalto the atomic concentration
primary ion beam [Vickerman, 1987]. Continually of elementsin the sample.The most intensesignalsare
bombardingthe surfaceof the samplewith the primary 27A1and28Si,followedby 4øCaand23Na,reflecting
in the feldspar.The 39Kand 88Sr
ion beamis a processknown as "sputtering." Most of their abundance
the particles which are sputteredaway are uncharged, signals are less intense, reflecting their lower abundance in the mineral (i.e., trace elements). Shown on
the
abscissais the sputtering time in seconds. Meaions.Either the positiveor negative"secondary"ions
surements
of crater depth as a function of sputter time
ejectedfrom the sample are separatedaccordingto
their mass/chargeratio and analyzed in a mass spec- have shown that the primary ion beam penetrates
at a rateof approximately
1-2 ]t/s. Thus
trometer.Thus the atomsmakingup the materialbeing labradorite
studied are continually ejected and analyzed, with the SIMS depth profile of the fresh labradorite shows
secondary ion intensities recorded as a function of that the concentrations of A1, Si, Ca, and Sr do not
time. The secondaryion intensitiesare proportionalto change significantlyfrom the surface of the mineral
the concentrationof elementsin the sample,thereby over the distance profiled (approximately 500-1000
producinga semiquantitativeconcentrationdepth pro- A).
file.
The SIMS depth profile of the same labradorRe
leached
in HC1 at p H 4 reveals a significantly different
The depthprofilerevealsthe chemicalcomposition
picture.
The intensitiesof AI, Ca, and Sr are signifiof a sampleas a functionof depththroughthe surface.
but a small fraction
consists of atomic or molecular
Reviewsof Geophysics,32, 2 / May 1994
pages 197-220
Paper number 93RG02329
Copyright1994by the AmericanGeophysicalUnion.
8755-1209/94/93RG-02329515.00
ß 197 ß
198 ß Shotyk and Metson' SECONDARY ION MASS SPECTROMETRY
32, 2 / REVIEWSOF GEOPHYSICS
zone of the leached mineral) and then to control specimens (no leaching). These ratios show the percent
change in AI/Si and Ca/Si of the reacted feldspar with
depth (Figures l c and l d). Aluminum and Ca are
clearly depleted relative to Si, from the surface of the
mineral toward the interior. The SIMS depth profiles,
therefore, provide direct evidence that an AI- and
Ca-depleted
layerapproximately
$00-1000f• thickhas
4
S'
1
Do
200' 4•0 ' 66¸ ' 860'1000
SputterTime (s)
C
r- • -lOO
I
I,
I
•
.....
'
Q) 0 _100,...•.,
Q-
0
,
,
,
,
...
400 800 1200 16002000
Depth(A)
Figure 1. (a) SIMS depth profile of fresh labradoritc feldspar
(An54). Notice that the secondary ion intensitiesof AI, Si,
Ca, and Sr do not vary significantlywith sputteringtime. The
intensitiesof 23Naand 39K are higherat the surfaceof the
mineral than in the interior. This feature is unique to SIMS
depth profiles of Na and K and does not accurately reflect
the true metal concentration profiles. Instead, both profiles
are affectedby the charge-inducedmigrationof Na + and K +
in responseto the negative primary beam [Streit et al., 1986].
(b) SIMS depth profile of labradoritc leached at pH 4.0 in
10-4 M HC1for 72 daysat roomtemperature.
Theintensities
of AI and Ca are significantly lower at the beginning of the
profile than they are after 500 s. (c) Percent change in AI/Si
of labradoritc (An54) leached for 72 days in HCI. Four spots
on the surface of the sample were depth profiled. (d) Same
as Figure 1c except for Ca/Si.
cantly lower at the surface of the mineral than in the
interior(Figure1b). For example,the 27A1ion intensitiesat thesurfaceof thefeldspar(5 x 103countsper
second) are 1 order of magnitude lower than the inten-
sitiesat 1000s (5 x 104countsper second).The
relativechangein 4øCaion intensityfrom surfaceto
formed on the surface of the dissolving feldspar.
The formation of such cation-depleted leached layers on the surface of feldspars dissolving in HCI was
first predicted by Correns and colleagues[Correns and
von Engelhardt, 1938a, b; Correns, 1940] more than
50 years ago. However, the possible existence of such
layers could not be confirmed and remained controversial for a long time (see Petrovic et al. [1976],
Holdten and Berner [1979], Aagaard and Helgeson
[1982], Helgeson et al. [1984], Tole e! al. [1986], Muir
et al. [1989], Mogk [1990], and Schweda [1990] for
reviews). A number of investigators used SIMS depth
profiling to identify significant hydrated, cation-depleted layers on the surface of dissolving glasses (see
the reviews by Gossink [1980] and Lodding et al.
[1985]). Once the value of the SIMS for depth-profiling
dissolving glasses was realized, it was successfully
applied to aluminosilicate weathering. In the last few
years, direct evidence for the formation of leached
layers on the surface of dissolving feldspars has been
obtained, to a large extent as a result of using SIMS.
The value of SIMS analyses for helping to understand
chemical weathering as it takes place at the solid-water
interface
is now well established.
The SIMS has found a variety of applications in the
Earth sciences [Shimizu and Hart, 1982; Reed, 1984;
Havette, 1985; Metson et al., 1985; Williams, 1985].
Silicates, carbonates, sulphides, biominerals, and biological materials have all been analyzed for major and
trace elements from H to U, including the platinum
group elements and rare earth metals (Table 1). The
emphasisof this review is on the use of SIMS to study
processes taking place at the surface of dissolving
glasses and minerals.
The purpose of this review is threefold: to outline
some of the outstandingproblems in chemical weathering, including an introduction to their historical development; to explain the SIMS technique and compare it with other methods of surface analyses; and to
show how the SIMS may be used to help answer some
of the unresolved questions concerning chemical
weathering in soils and sediments.
intensities
of 28Siare eitherunchanged
or perhaps
1.2. Chemical Weathering of Minerals in Soils
The long-standing interest in the rates and mecha-
slightly higherat the surfaceof the mineral in compar-
nisms of mineral
the bulk specimen is even greater. Notice that the
ison to the interior.
dissolution
in soils is rooted
in the
need to understand the development and maintenance
In order to emphasizethe changesin relative abun- of soil fertility. For example, the classic textbooks on
dancewith depth throughthe samples,the ratios AI/Si agriculture from the eighteenth and nineteenth centuand CaJSimay be normalized to bulk values(unaltered ries by Home [1762], Davy [1815], Liebig [1843],
32, 2 / REVIEWSOF GEOPHYSICS
TABLE 1.
Shotykand Metson: SECONDARYION MASS SPECTROMETRY
ß 199
Some Recent Applicationsof SIMS in Earth Sciences
Material
Silicates
Application
REE
major, trace, REE
REE, Th, U
Ni
H
oxygen isotopes
(Fluid inclusions)
Carbonates
Sulphides
Na, K, Ca
Fe, Mg, Mn, Sr
Au and Pt
Ag, Cu, Ni, Zn
Ag and Au
sulphurisotopes
Biominerals
Biological materials
Reference
Crozaz and Zinner [1986]
MacRae [1987]
Muir et al. [1987]
Bottazzi et al. [1991]
Vanucci et al. [ 1991]
Bottazzi et al. [1992]
MacRae et al. [1993]
Yurimoto et al. [1989b]
Gijbels [1987]
Yurimoto et al. [199!]
Herrig and Williams [1988]
Yurimoto et al. [1989a]
Lorin et al. [1990]
Hervig [1992]
Hervig et al. [1992]
Lorin [1992]
Diamond et al. [1990]
Swart [1990]
Lindsay and Sellschop [1988]
Vandentop et al. [1989]
Chryssoulis[1990]
Mcfarlane and Shimizu [1991]
Arehart et al. [1993]
Lodding et al. [1990]
Linton and Goldsmith [1992]
Ripoli et al. [1992]
REE denotes rare earth element.
[ !938] summarizesthe sequence of relative stability of
primary mineralsin soils.For mafic rocks the seriesis
addition to the importance of chemical weathering to o!ivine < augite < hornblende < biotite and for felsic
plant nutrition, we now realize that chemicalweather- rocks the series is calcic plagioclase < alkali plagioing regulates the chemical composition of natural wa- clase < K feldspar < muscovite < quartz. When a
ters [Stumm and Morgan, 1981; Berner and Berner, granite gneiss weathers, for example, the essential
1987; Drever, 1988] and provides the long-term acid- minerals of the fresh rock (oligoclase, orthoclase, mineutralizing capacity of soils [Sverdrup, 1990]. Al- crocline, quartz, biotite, and hornblende) are altered at
though pedology and geology continue to drive re- different rates: plagioclase and hornblende are the
search efforts, much of the interest in chemical least stable primary minerals, biotite, microcline, and
weatheringtoday is stimulatedby two global environ- orthoclaseare moderately stable, and quartz and zirmentalproblems:the desireto understandthe effectof con are the most stable; the residual clays are comacid rain on soils, natural waters, and building mate- posedchiefly of kaolinite [Goldich, 1938].
An example of a weathering profile developed on
rials [Drever, 1985; Colman and Dethier, 1986; Kittrick, 1986], and the need to predict the long-termfate granite is shown in Figure 2. In comparison to the
compositionof fresh rock, oligoclaseand biotite have
of vitrified nuclear waste [e.g., Brookins, 1984].
Chemicalweatheringis the alterationprocesswhich completelydisappearedfrom the upper layers of the
rocksand mineralsundergoas they approachchemical s0il profile,and microclineand quartzhave declined
equilibrium with their surroundings [Merrill, 1921; significantly.In contrast, kaolinite becomes increasChesworth, 1992]. Under the conditions of their for- ingly important toward the top of the profile. The
mation, all mineralsare thermodynamicallystablein thicknessof sucha weatheringprofile and the extent of
relationto their reactants. For example, igneousmin- alteration dependon factors such as temperature, rate
erals are stable under conditionsof high pressureand of water movement,p H, oxidation potential, the reactemperature.At the surfaceof the Earth theseprimary tive surface area of the reacting particles, and the
mineralsare thermodynamicallyunstableand so they specificnatureof the mineralbeingaltered [Jacksonet
decompose,releasingsolutesand producingsecond- al., 1948;Loughnan, !969; Carroll, •970].
Chemical transformations. The minera!ogical
ary minerals which are more stable under these new
conditions.
changeswhich take place during weatheringresult in
Johnston [1847], Dana [1858], and Thaer [!860], all
discuss the chemistry of soil mineral weathering. In
Mineralogicaltransformations. The "stability se- the losses of alkali and alkaline earth metal cations and
ries of mineral weathering" proposed by Goldich silica and the accumulation of A! and Fe oxides and
200 ß Shotyk and Metson- SECONDARYION MASSSPECTROMETRY
32, 2 / REVIEWSOF GEOPHYSICS
spondingrates measured in the lab [e.g., Paces, 1973,
1983; Velbel, 1985, 1986, 1989]. Are field and labora-
tory mineral weathering rates comparable [Brantley,
1992; Casey et al., 1993]?
First, the reacting surface area (i.e., the surface
area actually in contact with the weathering solution)
in natural soils may be much less than the total mineral
1.8II
surface area available for reaction [Drever and Swo-
2.0
boda-Colberg, 1989; Swoboda-Colberg and Drever,
1992]. In stirred reactors, for example, the entire mineral surface area is available
2.2
2.4
2.6
fresh rock
0
2o
40
Volume
60
80
lO0
Percent
Figure 2. Soil profile and underlying saprolite derived from
the weathering of Stone Mountain granite, Georgia [from
Grant, 1963]. The friable, white saprolite is dominated by
kaolinite, quartz, muscovite, and microcline; plagioclaseand
biotite have disappeared. When the base of the soil B horizon is encountered, the original rock texture is replaced by a
soil texture, with an increasein bulk density and the appearance of iron staining. (Reprinted with permission.)
hydroxides [Polynov, 1937; Bear, 1964]. Goldich
[1938], for example, noticed that the rate of loss of
cations relative to AI from the Morton gneiss decreasedin the order Na > Ca > Mg > K > Si. In fresh,
young soils the sum of the concentrations of N a, K,
Mg, and Ca are high in relation to that of AI or Fe,
whereas highly weathered tropical soils consistessen-
for reaction.
In a soil the
reactive surface area is probably a very small percentage of the total surface area, depending on the soil
texture, pore size distribution, and rate of water movement [Swoboda-Colberg and Drever, 1993; Velbel,
1993]. Second, in soils of the temperate zone, average
annual temperatures are low enough to account for
differences in rates of dissolution of up to 5 times
[Velbel, 1990]. Third, batch techniques which include
stirring or shakingmay give artificially high weathering
rates, probably due to mechanical abrasion [van Grinsven and van Riemsdijk, 1992]. Finally, lab experiments
typically omit the wetting and drying cycles characteristic of soil profiles [Inskeep et al., 1993].
In addition to physical differencesbetween leaching
conditions in the lab and the field, there are usually
1.4
1.2
tially of AI20 3 and Fe20 3 (Figure 3). The degreeof loss
of alkali and alkaline
earth cations relative
to AI or Fe
may be used to calculate a quantitative index of weathering intensity [Jackson and Sherman, 1953; Ruxton,
1968;Parker, 1970;Nesbitt and Young, 1.982;Harnois,
1988]. Of particular interest today is the rate at which the
mineralogicaland chemical transformationstake place.
Weathering rates. Rates of mineral dissolutionin
weathering profiles may be calculated from the chemical composition of natural waters [Johnson et al.,
1968; Katz et al., 1985; Stauffer and Wittchen, 1991],
from the mineralogical and chemical composition of
soil profiles [April et al., 1.986;Kirkwood and Nesbitt,
1991], or from experimental studies of the rate of
dissolution of whole soils [Brown and Lund, 1991].
The book by Sverdrup [1990] provides an excellent
introduction to the kinetics of mineral dissolution in
soils and summarizes
most of the recent literature.
Rates of mineral dissolution in the laboratory are usually calculated from the measured concentrations of
dissolvedconstituentsin the output solutionsof continuously stirred, flow-throughreactors [Rimstidt and
Dove, 1986]. Mineral dissolution rates measured in the
field may be 10-1000 times slower than the corre-
0.6
0.4
0.2
!
1
I
-'"oo•
-.
,,I,,
2
(1)
o
O
•
I
3
4
'•
•
5
o
F:
----
o
Figure 3. The depletion of alkali and alkaline earth metals
relative to A1 (or Fe, or Si) increases with increasing weathering intensity. Ratios such as these have been used in a
variety of ways to produce chemical indices of weathering
intensity. (After Jenny [1941].)
32, 2 / REVIEWSOF GEOPHYSICS
Shotykand Metson- SECONDARYION MASSSPECTROMETRY
ß 201
significantchemical differencesbetween the two environments. For example, the compositionsof the leach-
z
ing solutions used in lab studies are simple, often
consistingonly of HCI. Natural soil solutioncompositions are much more complex. Consider the distribution of cationic charge in forest soil solutions, for
example.Even at pH 4, H + may accountfor lessthan
20% of the total cationiccharge(Figure4). The pres-
u
enceof othercations,especially
A13+andits hydrolysisproducts, may significantly reduce dissolutionrates
[Chou and Wollast, 1984; Casey et al., 1988; Drever
and Swoboda-Colberg, 1989; Swoboda-Colbergand
Drever, 1992]. Also, the anionic charge in forest soil
solutionsmay be completely dominated by organic
ligands which may accelerate or retard rates of dissolution, dependingon their size and charge [Stumm et
al., 1987]. However, the possible effects of dissolved
cations and organic anions on rates and mechanismsof
mineral dissolution
are not well known.
In section2 of the paper an introductionis provided
to some chemicalfactors affecting the rates and mechanisms of silicate mineral dissolution, including the
effect of dissolved salts and complex-formingorganic
ligands.Until recently, suchstudiesand their interpretation were basedprimarily upon chemical analysesof
the leaching solutionsbefore and after reacting with a
solid phase, with inferences then made about surface
reactions. Today, however, we are able to study dissolution and precipitation processesdirectly as they
take place at the solid-solution interface [Hochella and
Figure 5. Brongniart' s [ 1841] apparatus for the electrolysisof
dissolvingK feldspar. Finely ground feldspar was placed in
the tube, with a strip of copper in one arm and a strip of zinc
in the other. The metal strips were connected by a wire.
After 15 days the liquid on the zinc side was still opaque
whereas that on the copper side had settled and was quite
clear. The experiment was continued for 2 years. By this
time the zinc electrode was heavily coated with a residue of
leached feldspar. The copper electrode remained clean and
the copper arm contained a strong solution of alkaline potassium carbonate. (From Cushman and Hubbard [ 1907].)
15
pH
4.55
25
pH 3.76
E
White, 1990]. SIMS is a powerful analytical technique
for studying this region. In section 3 the principles of
the SIMS method are described, and a comparison is
made with other methods of surface analysis. In section 4, examples are given of successfulSIMS applications to selected problems in chemical weathering.
pH
4.O4
55
105
2.
I
o
I
20
I
I
40
I
I
,I
60
I
80
I
LABORATORY
STUDIES
OF MINERAL
I
100
Cationic Charge (%)
• H+
• Na+
• K+
• Mg2+
•,• Ca2+
• AI3+
Figure4. Cationic chargebalancein a Cryptopodzolicforest
soil solutionat Coperain CantonTicino, Switzerland.Even
at pH 3.8, H + accountsfor only 20% of the total cationic
charge.
DISSOLUTION
The traditional approach to quantifying mineral dissolution rates and identifying reaction mechanisms is
basedupon the chemical analyses of aqueous solutions
before and after reaction with a solid phase. Brongniart [1841] showed that during electrolysis potassium
feldspar dissolvesincongruentlyin water, with K +
becoming separated from the negatively charged, residual colloid (Figure 5). This process was first quantified by Daubrde [ 1879]when he found that K feldspar
202ß Shotykand Metson:SECONDARY
ION MASSSPECTROMETRY
32, 2/REVIEWSOF GEOPHYSICS
dissolving in distilled water released to solution 10 increasinglysiliceouslayers favored in more acidic
times as much K as either A1 or Si.
solutions.Assumingthat dissolutionwas uniform over
the surface of the mineral, leached layer thicknesses of
2.1. The "AmorphousPrecipitateHypothesis"
The first systematic studies of rates of feldspar
dissolutionin closed systemswere presentedin a series of papers by Cushman [Cushman, 1905a, b,
1907; Cushman and Hubbard, 1907, 1908]. These
studiesincludeda range of feldsparcompositionsand
showed that the rate of dissolution was highly depen-
dentupontheir compositionandparticlesize(reactive
surface area). In closed systemsthe dissolutionrates
were relatively rapid initially, but soon the minerals
were found to stop dissolvingaltogether.This led to
the suggestionthat the rate of dissolutionof feldsparin
water was controlledby protectivecoatingsformed by
amorphoussolidson the surfaceof the mineral;these
coatingsformed by precipitationfrom saturatedsolutions in closed systemsand were seen under the microscope[Cushman, 1905a]. Under the conditionof
the experimentsthe removal of insolubledecomposition productswas controllingthe rates of dissolution
of the feldspars. If the reacted feldspar with its protective surface layer was dried and reground, the rate
of dissolutionwas againrelatively fast initially because
of the regenerationof reactive surfaces.
The formation of such precipitates on dissolving
feldsparsleachingin closed systemscontainingsaturated solutions eventually became known as the
"amorphousprecipitationhypothesis"(seethe review
by Mogk [1990]). In closed systemsthe formation of
protectivecoatingswas believedto controlthe rate of
feldspardissolution.In a seriesof similarexperiments
usingclosedsystems,Busenbergand Clemency[1976]
and Busenberg [1978] reported the mineralogicalcompositionof such precipitatedphasesfor a range of
feldspar compositions.
upto 2000• werecalculated.
Thistheoryeventually
became known as the "leached layer hypothesis." At
the time there were no surface-sensitive analytical
instrumentsfor testing this theory. More recent solution chemical studies by Chou and Wollast [1984,
1985], Holdren and Speyer [1985], Wollast and Chou
[1985], and Schweda [1989] also predicted cation- and
Al-depletedleachedlayerson the surfaceof feldspars
dissolvingin opensystemsusingacidic solutions.Until recently, however, there was no direct evidenceto
confirm the formation of such layers.
The leached layer hypothesisof Correns did not in
any way contradictthe amorphousprecipitatehypothesis of Cushman: Cushman's studies were performed
in closed systemsin which the primary minerals dissolved, the productsof dissolutionwere allowed to
accumulate, and secondary phases precipitated. The
studiesby Correnswere performedin open systemsin
which aqueousreactionproductswere continuallyremoved from the reaction vessel, making it impossible
for secondaryphasesto precipitate.
Studies of naturally weathered feldspars using the
scanningelectron microscope(SEM) showed clearly
that feldspar dissolutionis nonuniform, resulting in
deep, extensivelydevelopedetch pits [Wilson, 1975;
Berner and Holdren, 1977, !979]. In order to calculate
the thicknessesof his proposed cation- and Al-depietedleachedlayers, however, Correnshad assumed
that the feldsparsdissolveduniformly to produce a
leached layer laterally constant in thickness around
the surfaceof eachgrain [Berner et al., 1985].Thus the
discovery of etch pits on the surface of naturally
weathered feldspars appeared to constrain the
"leached layer hypothesis"as proposedby Corrensto
explain the experimentalstudiesof feldspardissolution. Furthermore, studiesusing X ray photoelectron
2.2. The "Leached Layer Hypothesis"
(XPS) and Auger spectroscopy
failed to
The first systematic studies of feldspar dissolution spectroscopy
(greaterthan 100•) leached
in open systemswere publishedby Corrensand co- detectany significant
workers [Correns and yon Engelhardt, 1938a, b; von layerseither on naturallyor experimentallyweathered
EngeIhardt, 1939;Correns, 1940;Correns, 1962a, b]. feldspars[Petrovicet al., 1976;Berner and Holdren,
Using flow-throughreactors,they found that the rela- 1977, 1979; Holdren and Berner, 1979; Fleer, 1982;
tive proportionof A1 and Si which is releasedfrom Fung and Sanipelli, 1982; Hochella et ah, 1988].
dissolvingalkalifeldspardependeduponthep H of the Mainly on the basisof the lack of direct evidenceof
reacting solution. At p H 3, for example, K feldspar significantleachedlayerson the surfaceof dissolving
aluminosilicates, the leached layer hypothesis was
pH 6.6 the mineral lost slightly more Si than A1 [von largely discredited.
Engelhardt, 1939].
On the basisof the chemicalcompositionand molar 2.3. Effect of Dissolved Salts
lost more than 4 times as much A1 as Si. In contrast, at
volume
of the unreacted
mineral
and the chemical
compositionof the output solutions,Corrensand von
Engelhardt [1938b] concluded that cation- and A1depleted, residual, amorphous leached layers were
forming on the surface of the dissolvingfeldspar. They
suggestedthat the ratio of A1 to Si in the residual
surface layer depended upon the solution p H, with
Beyer [1871] found that salts of NH4 and Na increased the release of K from alkali feldspar by a
factor of about 3, in comparison to distilled water.
Cushman and Hubbard [1907] and Andrd [1913] re-
portedsimilarresults.AnderssonandLindqvist[1956]
found that the dissolution rate of microcline in 0.1 M
salt solutionsat p H 6 decreasedin the order LiC1 >
32, 2 / REVIEWS
OF GEOPHYSICS
Shotykand Metson:SECONDARY
ION MASSSPECTROMETRY
ß 203
water > NaC1 > KC1 > CsC1, with the rates of disso-
lution in a given salt solution decreasingwith increasing salt concentration.The resultsof the experiments
were interpretedin termsof the selectiveadsorptionof
cations on the dissolving mineral surface. Nash and
Marshall [1956a, b] performed similar experimental
Bruckert, 1982], but there have been few measurements. In a forest soil solution, Hue et ah [1986]
measureda total of 480 i•mol/L of organic acids, including succinic, malic, lactic, malonic, citric, and
oxalic(282, 117,52, 20, 5, and4 i•mol/L, respectively).
In experimental lab studies of mineral dissolution,
studies.
however,!igandconcentrationsas highas 0.1 M have
More recently, Sj6berg [1989] examined the effects been used [Schalscha et al., 1967]. Chin and Mills
of salt solutions on the rate of dissolution of labra-
[1991] showed that millimolar
dorite over a wide p H range at temperaturesup to
70øC.In HC1 solutionscontainingno significantconcentrationsof cationsother than H +, A1 was released
preferentiallyover Si, consistentwith the formationof
residualleachedlayers.Althoughno directanalysesof
oxalic, malonic, and salicylicacids promotedthe rate
of kaolinite dissolutionin comparisonto solutionsat
the samep H without organicligands. Welch and Ull-
the reacted surfaces were made, the calculated thick-
dissolution. However, even millimolar concentrations
concentrations of
man [!992, 1993] also used millimolar concentrations
and reported similar results for plagioclasefeldspar
nessesof the leachedlayers predictedto have formed may be 10-100 times higher than the organic acid
were100• atpH 4, 1000• atpH 3, and4000• atpH concentrationspresent in soil solutionsand other nat2. Adding KC1 (0.!, 0.01, or 0.001 M) significantly ural waters.
reduced the dissolution rates.
The effectof simpleorganicacidsversusmore complex organic ligands on the rates and mechanismsof
2.4. Effectof Complex-Forming
OrganicAcids
silicate dissolution has not been extensively investiThe first observations of siliceous layers on the gated. In contrast,the effect of organicligandson the
surfacesof silicatesdissolvingin solutionscontaining rates of dissolutionof oxide and hydroxide mineralsis
organic acids were published by Monier [1877]. He now reasonablywell documented [Furrer and $tumm,
leached 500 cm• of sodium silicate in a 1-L solution 1986; $tumm et al., 1985;Schindler and $tumm, !987;
containing75 g of oxalic acid and describedthe growth $tumm and Furrer, 1987]. The mechanisticinterpretaof siliceous layers which achieved a thickness of 7-8 tion of aluminosilicatedissolutionis more complicated
mm after 48 hours.These and other pioneeringstudies than that of oxides and hydroxides, but, as a first
were summarized in an exhaustive review of early approximation, the conceptsdeveloped for oxide and
work published by Julien [1879]. With respect to hydroxide surfacesmay be extended and applied to
acidic, organic-richforest soils suchas podzols,soil the kinetics and mechanismsof weathering reactions
scientistshave long been in favor of an important role of silicates [Stumm et al., !987; $tumm and Wollast,
for organicacidsin mineralweatheringand soilprofile 1990].The rate constants
for the dissolutionof A120•,
development[McKeague et al., 1986;Robert and Ber- for example, in the presenceof aliphatic organicacids
thelin, 1986; Tan, 1986].This theory is indirectly sup- decrease in the order oxalic > ma!onic > su½cinic,
ported by studieswhich have shown that a large pro- which form five-, six-, and seven-member chelate
portion of the dissolved A1 in forest soil solutions is rings, respectively [Stumm et al., 1985; Furrer and
organically bound [Driscoll et al., 1985; Litaor, 1987; $tumm, 1986]. For aromatic ligands the rate constant
Manley et al., 1987] and can help explain the high for salicylic acid, which forms a six-member ring, was
concentrationsof total dissolvedA1 (up to 5 rag/L) in 4 times that of phthalate, which forms a seven-member
such solutions [Antweiler and Drever, 1983]. How- ring. These ligands form surface complexesby ligand
ever, a number of unansweredquestionsremain. First, exchangewith surfacehydroxyl groups,bringingnegdo organicacidspromoteor inhibit the rates of mineral ative charge into the coordination sphere of the surdissolution?It has been suggestedthat simple, low- face A1 species.This polarizes the critical Al-oxygen
molecularweight organic acids tend to promote min- bonds, thus facilitating the detachment of A1 from the
eral dissolution while complex, large-molecular surface. The surface-bindingstep (adsorption) is fast,
weight humic materials inhibit this process (see be- followed by a slow detachment step in which the
low), but there have been few systematic studies. complexed metal ion is transferred to the solution.
Sincevery few datahavebeenpublishedregardingthe Five-member chelate rings are the most detachable,
concentrationsof individual organic acidsin forest soil followed by six- and seven-member rings. For comsolutions,it is difficultto designrealisticexperiments. parison, using millimolar concentrations of organic
Second,the relative importanceof organicacidsas acids, the rate of dissolution of kao!inite decreases in
proton donors versus their role as ligands or reduc- the order oxalate > malonate > salicylate > phthalate
tants is unclear [Manley and Evans, 1986;McColl and [Chin and Mills, 1991];this is the same order as given
Pohlman, 1986;Pohlman and McColl, 1986, 1988].
abovefor A120•.
Unlike the low-molecular weight organic acids,
Individual, low-molecular weight organic acids in
forest soil solutionsare thoughtto be presentat con- higher-molecularweight degradationproductsof biota
centrations
of the orderof 10-5 to 10-4 M [Vedyand suchas humic and fu!vic acids are expectedto become
204 ß Shotyk and Metson: SECONDARYION MASS SPECTROMETRY
32, 2 / REVIEWSOF GEOPHYSICS
10•V prknary i•
o
oooooooo
o©ooo¸o©
Figure 6, The progress of the collision
cascadein a solid surface, resulting in
secondary ion emission.
!O'•'•md8
strongly adsorbed upon silicate surfacesand inhibit
mineral dissolution[Stumm et al., 1987].Experimental
studies of Fe oxide dissolutionwith aqueousextracts
of chestnut (Castanea satira) leaf litter have shown
that this chemicallycomplex, organic-rich"tea" forms
protective coatingson the mineral surfaceand inhibits
the release of Fe [Blaser, 1974;Blaser et al., 1981]. At
pH 4.5, pyrocatechol violet is effectively adsorbedby
a!bite and oligoclaseand decreasestheir rates of dissolution [Schenk et al., 1989]. Experimental studiesof
kaolinitc weathering have shown that humic acid
creasesdissolutionrates in comparisonto solutionsat
the samepH without humic acid [Chin and Mills, 1991].
The effectsof saltsand organicligandson the chemical composition of dissolvingglass and mineral surfaces may be studieddirectly using SIMS.
cascade" is initiated, in which the incident energy is
dissipatedthrougha number of collisionpartners in
the solid [Williams, !979]. During the course of this
"cascade" the primary ion is typically implantedinto
the surface, and secondaryparticles are dislodgedand
may subsequently escape the surface (Figure 6). A
smallproportion,perhapsonly 1%, of this "sputtering
yield" is ionized and can be extracted directly into a
mass spectrometerfor analysis. Thus in the simplest
form of SIMS the user observesa mass spectrumof
secondaryions ejectedfrom the sample surface.These
ions reflect the compositionof the surface, although it
is important to note they are not directly proportional
to relative
element
abundances
at the surface.
Ion
yields sufferfrom a significantchemical matrix effect.
Althoughthe efficiencyof secondaryion productionis
low, detectionlimits still extend to around 10 ppb for
favorableelements,particularlythe alkali and alkaline
3.
SURFACE ANALYSES USING SIMS
earth metals (Figure 7).
Because the generation of secondary ions is the
31.
Introduction
limiting analyticalprocessin SIMS, any specieswhich
To analyze the surfaceof a mineral and studyprocan be ionized (positiveor negative) can, in theory, be
cessessuchas leachingoccurringat suchsurfaces,it is
analyzed. Thus most of the periodic table can be
necessaryto utilize methodswhich can directly examine a zone a few atomic layers in depth. In additionto
surface sensitivity, such methods should have reason107
]
.... [
]
[
:
able absolute sensitivity for a wide range of elements
and allow the analysisof compositionas a function of
'rt•
IdoRu
depth normal to the mineral surface. Secondary ion
mass spectrometry (SIMS), in its simplest form, is
'1•
#b
Hf
basedon the erosionor "sputtering" of a surfaceby a
primary ion beam and the subsequentcollection and '•
I
.
Pd
analysisof the secondaryions produced. SIMS fulfils
most of the above criteria and is thus an extremely
powerful tool in weathering/surfaceexchangestudies.
The acronym SIMS embraces a family of related techniques and instrument configurations,of which only
some are relevant to this type of work. The attributes
of these methods and their limitations in geochemical
analysis are discussed below. A detailed review of
• 102
SIMS in generalcan be foundin the work by Benninghoven et al. [1987].
10•
0
3.2.
I
I
i
I
20
40
60
80
100
The SIMS Method and Selection of
Instrumentation
Atomic number (Z)
When an incidention beamof energybetween5 and •i•re 7. •on yieldsfrom oxygenbombardmentof elemental
30 kV impacts on the surfaceof a solid, a "collision samples. (•mm the dat• o• •torms et M. [1977].)
32, 2 / REVIEWSOF GEOPHYSICS
Shotykand Metson' SECONDARYION MASS SPECTROMETRY
ß 205
cessed, including hydrogen [e.g., Yurimoto et al.,
1989a]. The raw spectrumwill, however, contain any
readily ionizable species, which includes molecular
ions, most importantly in mineral analysis, oxides.
Thus in geochemicalapplications,someform of spectral filtering is usually applied to extract the elemental
ion signature [Reed, 1980].
The experimental conditionsused for geochemical
analysis generally involve "dynamic" conditions
[Benninghovenet al., 1987], where high primary ion
current densities rapidly erode the sample surface.
Thus the total ion yield is high, and massresolutionof
the analyzing spectrometeris generally important in
resolving isobaric interferences to identify the target
species. High-transmissionmagnetic spectrometers,
and predominantlythe Cameca IMS3f through 5f instruments,currently dominate this type of study. Althoughmagneticspectrometerslack the transmission
attributes of the time of flight (TOF) analyzers typically usedin "static" SIMS, the largeextractionfields
used offer considerable flexibility, both in achieving
reasonabletransmissionefficiency and particularly in
energy filtering to remove molecular ions from the
spectrum.
In the Cameca instruments the spatial distribution
of secondaryions is maintained through the spectrometer; thus the instrument is a real-time ion microscope,
and a mass-resolvedimage of a large area can be
viewed simultaneously.This gives significanttransmissionadvantagesover an instrumentwhich collects
signalwhile rasteringthe primary beam acrossthe
surface (ion microprobe), although the Cameca can
also be used in this way. Because the microscope
The most significantlimitations include the following: (1) There are difficultiesinvolved in convertingion
abundancesin the secondary ion spectrum into element abundancesat the surface. The processof ionization at these relatively low energies is matrix sensitive and modeled only with considerabledifficulty.
(2) Mineral surfacesare usually insulating;thus the use
of chargedparticle beamsin sputteringand extraction
means chargingmust be controlled, and this is sometimes difficult. Negatively charged oxygen is the most
convenientprimary beam for mineral analysis;positive cesium is the most difficult. (3) The geometry of
ion incidenceat the mineral surface means shadowing
effects from surface roughnesscan severely limit the
accuracyof the depth profile. (4) SIMS is an ultrahigh
vacuummethod, so samplesmust be examinedex situ.
Thus samplehandlingand preparation are especially
critical.
It is worth consideringthe impact of some of these
factors in more detail by considering how the SIMS
processworks.
lhe generationof secondaryions. The processof
generatinga secondaryion can be separatedinto several steps.The sputteringprocessis relatively well
understood.On the basisof the Sigmund[1969, 1987]
model the transfer of energy between the incident ion
and its collisionpartnerscan be reasonablywell modeled, and the total sputteringyield is obtainedfrom
calculation. Accuracy is good for sampleswhere preferred orientation and channeling effects do not interfere, such as most metal surfaces, but is limited for
crystallinesolidsand thus many stn-faces
of geochemical interest.Typically the "collision cascade"is com-
spatialresolutionis dictatedby the ion opticalresolution of the extractionoptics and secondaryion column, which is generallyinferior to that of the primary
column, ion microprobesoffer an advantagein ulti-
pletein 10-•3 s andyieldssputtered
particles
withan
mate resolution [Betz and Rudenauer, 1991]. This can
be usefulin imagingbut is only rarely usefulin analysis
because of the sacrifice in sensitivity.
The most significantadvantagesof SIMS over, for
example,X ray methodsare as follows: (1) The very
will be ions is difficult.
effect.
relevance to this review is the bond-breaking model
energydistributionpeakingstronglybelow 10 eV and
extendingmany hundredsof eV above this energy.
However, predictingwhich of thesesputteredspecies
Secondaryion yieldsfrom similarmatricesfollow
an inverseexponentialrelationshipto either the ionization potential(positiveions) or electronaffinity
(negativeions). However, a strong dependenceon
largesignal-to-noise
advantages
of massspectrometric chemicalstate or, rather, the positionof the Fermi
detection give a very large usable dynamic range, level, is also observed.This partly explainswhy the
typically better than 8 orders of magnitude,between use of an electronegative species, such as oxygen, as
the most and least abundant elements which can be the primaryion enhances
positiveion yieldsby up to
observed.(2) The very limited escapedepthof second- severalordersof magnitude,while electropositiveceary ions,typicallyseveralatomiclayers,makesSIMS sium enhancesnegativeion yields [Wittmaack, 1981;
extremely surface sensitive. (3) SIMS is a natural Yu, 1986a]. Figure 7 showsthe relative ion yieldsof
depth-profilingmethod,in that the analyzedzone pro- positiveions,mostlyfrom elementalsamples,under
In enhancing
ionyields,the
gressivelymovesinto the sample,as the surfaceis oxygenionbombardment.
removedby sputtering.Erosionratesvary fromfrac- sourceof oxygenis not important.Thus oxygenleaks
tions of a monolayer to more than a nanometer per have been used, and yields are enhancedfrom oxide
second,dependingon primaryion speciesandcurrent matrices.
Numerous modelshave been developedto predict
density.(4) Mass spectrometricanalysisallowsisotopes to be resolved. Thus tracer studies such as ion yields.It shouldbe notedthat noneof thesemodleachingin deuteratedsolutionscan be usedto great elsis universalto all sampletypes. The modelof most
206 ß Shotyk and Metson- SECONDARYION MASS SPECTROMETRY
I
I
I
I
I
!
on ion yields [Nesbitt et al., 1986]; however, the same
approach is not available in depth profiling.
lhe specimenisolation method. Although the ion
yield problem may not seriously interfere with depth
profiling, the discrimination of molecular ions from
elemental ions certainly does. A major advantage of
'
102
the sector instrument
%
10ø
is that mass resolution
can be
used to discriminate between many common isobars.
Also, when using a large extraction field, the secondary ion energy spectrum can be used to select energy
windows where the elemental ion spectrum dominates
over the molecular ion spectra. This is best explained
by reference to Figure 8 where the narrow energy
Co'
o
\
•,Co2'
-1
32, 2 / REVIEWSOF GEOPHYSICS
spectrum of Co• is effectively removed by analyzing
10-2
-50
•
I
0
50
•
1 O0
•
!
•
150
200
250
300
œner•y (eV)
Figure 8. The energy distribution of Co- and Co•- from a
metal specimen,bombardedwith a Cs+ primary beam.
ions with emission energies above 100 eV.
This approach is limited by the degree of energy
filtering which is needed to eliminate the molecular ion
spectra. Hydrides and the oxides of heavier elements
are difficult to remove this way. One technique which
avoids this problem is an extreme form of energy
filtering known as the specimen isolation method [Metson et al., 1983; Lau et al., 1985]. Although more
widely used for trace element analysis, this method
has been usedwith considerablesuccessin depth profiling feldspars[Nesbitt and Muir, 1988; Muir et al., 1989,
1990; Muir and Nesbitt, 1991, 1992; Nesbitt et al., 1991;
proposed by Slodzian [1975] to explain the large secondary ion yields from ionic solids. This is the model
which also most comfortably applies to oxides [Williams, 1979]. The ion ejection process is seen as a
single bond cleavage which in the case of cation emissionresultsin a vacancy site's trapping a free electron.
If this electron is trapped at the vacancy longer than
Shotyk and Nesbitt, 1992]. Extreme energy filtering is
achieved by allowing the sample surface to charge, thus
creating a large in situ energy filter or threshold, controlled only by the dimension of a conductive aperture
placed above the surface (Figure 9). The method is a
straightforwardway to achieve both control of charging
and extreme energyfilteringon insulatingsurfacesand is
thetimescale
of thesputtering
event(10-13S),thenthe thus well suited to geochemicalanalysis.
cation will escape the surface. The model is not easily
formulated into a usable relationship, because of the 3.3. Depth Profiling
difficulty in describing the properties of the vacancy
SIMS, in dynamic mode, is a natural depth-profiling
(electron affinity and amplitude of the wave function). method, because the primary ion beam progressively
However, where usable approximations have been
applied, reasonable correlation with experimentally
observed positive ion yields has been found [Yu,
1986b].
The problem of quantifying ion yields can, in general, be avoided when depth profiling because the
matrix is often not substantially modified with depth,
and thus relative ion yields do not vary as the depth
profile proceeds. The usual procedure is to take the
Extractio• glate
ratio of the element of interest to a matrix element, for
example Si or A1, which is either not anticipated to
vary or varies predictably in the regime being examined. This procedure also removes the effects of artifacts, such as' surface roughening, which alter total
secondary ion yields.
If quantitation is essential, the only universal, but
often impractical, approachis still to have standards
availablewhich match the matrix of the sampleto be
analyzed. In whole rock analysis, matrix dilution has
been very successfullyusedto mitigate matrix effects
sI plate
Teflon •
Sample
Teflon spacer
Sampleholder
Figure 9. The sample support geometry used to exploit the
specimen isolation (SI) method. Note the isolation of the
sample from the secondary ion extraction voltage applied to
the SI plate.
32, 2 / REVIEWSOF GEOPHYSICS
Shotykand Metson: SECONDARYION MASS SPECTROMETRY
ß 207
erodesthe sample surface.Thus relative ion signalsas
a function of time representa depth profile of element
concentrations.If the primary beam is raster scanned
acrossthe samplesurface,then a symmetricalcrater is
formed; the erosion front is flat and ideally moves
perpendicularto the surface.There are, hqwever,a
number of factors which complicate this picture and
limit the accuracy of such an element profile.
Limits are imposed by three principal factors: crater edge effects, surface roughness, and ion beam
mixing and selective sputtering.
Crater edgeeffects. If the signalfrom the edgesof
the sputteredcrater contributesto the profile, then ion
intensities are simultaneouslybeing collected from a
rangeof depths, rather than the crater bottom. This is
most readily eliminated by either raster gating the
secondaryion beam in a microprobe-typeinstrument,
or by constrainingsecondary ion optics in a microscope-type instrument, so then only signal from the
central zone of the sputteredcrater is collected.
Surface roughness. In typical instrumentgeometries the primary beam impactsthe surfaceat an angle
between 30ø and 45øto the surface normal. Angles can
increase dramatically in high-extraction field instruments, when low primary beam energies are used
[Meuris et al., 1989]. Thus surface roughnesswill act
to shadow the surface and thus, at least initially, cause
substantial loss of definition in the observed profile.
Publishedprofiles on mineral surfaces are rarely as
dramatic as those from, for example, semiconductor
surfaces, for both this reason and the natural inhomogeneitiesin processesoccurring at mineral surfaces.
Ion beammixingand selectivesputtering. The dynamics of the sputtering process are complex. Initially, primary ions are implanted into the surfaceand
thus alter its composition.Also some surfacespecies
may be selectivelysputtereddue to the natureof their
bonding.Both effectsare transitory,and despitethe
continuederosion of the surface a steady state condition is rapidly reached, usually within seconds.The
collision cascade will, however, continue to mix the
compositionin the zone ahead of the sputteringfront.
This mixing ultimately limits the resolution of sharp
interfacesand the usefultotal depth of a profile. Owing
to the nature of interfacesin most natural systems,this
is rarely a limitation in depth profilingminerals.
3.4. ComparisonWith Other Methods
of SurfaceAnalysis
The use of SIMS in these applicationsmust be
weighed against other possible methods of surface
analysis. Both X ray photoelectron spectroscopy
(XPS) and Auger electron spectroscopy.
(AES) have
been used with considerable success in the surface
analysisof minerals.The applicationof thesemethods
to the surfaceanalysisof geologicmaterialshas been
reviewedin the bookby Perry [1990],andMogk [1990]
reviewedthe applicationof AES to chemicalweather-
ing. AES is rathersamplerestrictedin that samplesare
required to be at least partially conductive [Hochella,
1988], and XPS has been more widely used. Both
methodsrely for their surfacesensitivityon the limited
escape depth of low-energy (10-2000 eV) electrons
and are thus similarly sensitive to the outer 5-20
atomic layers. They lack the absolute surface sensitivity of SIMS and requirethe sputteringof the surfaceas
a separatestep, in order to create a depth profile. The
reliance on an external (usually inert gas) ion gun
typically limits sputtering rates to an order of magni-
tudelowerthanSIMS usinga Cs+ primarybeam.This
restrictsthe accessibledepth in an XPS/AES profileto
fractions
of a micrometer.
A more significantlimitation is the relatively poor
sensitivity of these methods, particularly for light elements. Although considerablybetter than X ray methods, sensitivityfor elementsbelow boron is poor and,
importantly, hydrogen is inaccessible.At best, these
methods are capable of detecting elements at the 0.1
at. % level, far inferior to SIMS. On the other hand,
the cross sectionsfor photoelectron and Auger electron emission are well known, and thus both methods,
but especially XPS, are immediately quantitative.
Electron yields are not matrix sensitive, and thus accurate major element profiles can be obtained.
Other ion beam techniques, based on the interaction of energetic(MeV) ion beams with surfaces,have
been applied with some successto mineral surface
leachingproblems[Schottand Petit, 1987;Petit et al.,
1990b]. Although particle-induced X ray emission
(P!XE) has insufficientdepth resolution(> 1 Ixm)to be
useful in this respect, Rutherford backscattering
(RBS), elastic recoil detection analysis (ERDA), and
especially nuclear reaction analysis (NRA) all have
applicationin this area. NRA has been usedwith great
effect in hydrogenprofilingof glassesand other matehals where the modified zone of interest is up to a
micrometerin depth, and the depth resolution(at best
5-10 nm) can be usefully applied [Della Mea et al.,
1983; Schreiner et aI., 1988;Petit et al., 1987, 1989a,
b, 1990a, b]. Casey and coworkers have contributed
significantlyto our understandingof feldspardissolution using RBS and NRA to study the formation of
leachedlayers on the surfaceof dissolvinglabradorite
[Caseyet al., 1988, 1989a, b]. Despite successwith
hydrogen, boron, fluorine, sodium, and some other
elements, however, NRA is extremely element selective. There are a limited number of suitable interac-
tions, and becausedepth resolution is energy sensitive, they do not generally have sufficient depth
resolution for many mineral weathering problems.
Also, lateral resolution is currently rarely better than
!00 Ixmfor the energeticion beam methods.Microfocused beams of a few micrometers are being further
developed,but the problemis technicallychallenging
with MeV particles.
32, 2 / REVIEWS OF GEOPHYSICS
208 ß Shotykand Metson: SECONDARYION MASSSPECTROMETRY
TABLE 2. Comparison
of SurfaceAnalysisMethodsAppliedto GeologicMaterials
SIMS
RBS
XPS
AES
Z > Li
50 nm
1 nm
0.1%
yes
Elementsanalyzed
Spatialresolution
Depthresolution
Bestsensitivity
Quantitative
Insulatingsamples
H-U
<30 nm
<1 nm
0.01ppm
difficult
yes
Z > Li*
5 Ixm
30-80rim1'
100ppm
yes
Z > Li
100Ixm
1 nm
0.1%
yes
yes
Chemical information
some
no
yes
yes
yes
difficult
*The abilityto detectandresolveelements
is a sensitive
functionof thedetectorresolution
andthe
particularelement/substrate
combination.
?Depthresolutionis againvery sensitiveto detectorresolutionand particlestraggling
in the
particularsubstrate(thusit is itselfa functionof depth).
A broad comparisonof the attributesof SIMS,
surface enrichments noted above, the SIMS depth
AES, XPS, and RBS is shown in Table 2.
profilesrevealedsignificant
enrichments
of Mg andSi.
4.
Observationsusing SEM/energy-dispersiveX ray indicatedthat a hydratedMg silicatehad precipitatedon
the reacted surface [Hayward et al., 1984]. Both of
APPLICATION
OF SIMS TO CHEMICAL
WEATHERING
these results were consistent with the calculated sat-
uration state which indicated the solution was super4.1.
Dissolution of Glasses
saturatedwith respectto sepiolite,a chainlattice clay
mineral.
Schreiner [1989] performed leaching experiments
lab studies. Dissolving glasses have been successfullydepthprofiledusingSIMS by a numberof with two glass compositions:15 K20' 25 CaO-60
authors[e.g., Gossinket al., 1979;Smetsand Lorn- SiO2 (wt %) and 16 K20' 15 CaO. 39 SiO2' 15
men, 1982; Lacharme and Lehuede, 1985;Richter et PbO - 10P20$ß3 MgO ß2 A1203(wt %); the latter is a
al., !985; Zoitos et al., 1989].Gossinket al. [1979],for dark green glasssimilar in compositionto the glass
example, used SIMS to depth profile a glass (20 usedin medievalstainedglasswindows. The glasses
Na20' !0 CaO-70 SiO2mol %) leachedin water at were leachedin HC! and H2SO• rangingfrom 0.1 M to
75øC for 15, 30, and 40 min. Using a 100-hA O- 0.1 mM at room temperaturefor up to 5 hours. The
primarybeam and a Ta diaphragmto reducesurface glass sampleswere gold coated to reduce surface
charging,a plot of measuredcrater depthversussput- chargingand probedwith a 100-nAO- primarybeam.
ter time showed an excellent linear correlation, giving Eachacidproducedan alteredlayer severelydepleted
anaverage
sputter
rateofapproximately
0.5•/s. The
in K and Ca but enriched in H and Si; however, the
samplewhich was reactedfor 30 min showedthat Na depthof alterationin 0.1-N H2SO4wasapproximately
was significantlydepletedrelativeto Si for more than one half that producedby 0.I-N HC1 (Figure 10).
2000 s; this correspondsto a leached layer approxi- Using high mass resolution (M/AM from 2600 to
mately1000• thick.Calciumwasalsodepleted
in 3100), $chreiner[1989]was able to clearly distinguish
from3202anddepthprofile
both32Sand35Ciinthe
relation
to Si,butonlytoa depthof500•, revealing
a 32S
dark greenglass:The glassleachedin HC1 showeda
preferentialleachingof Na.
enrichmentof C1,whilethe glassleachedin
Titanire (CaTiSiOs)is being consideredas the ce- significant
in S (Figure10).Thedifference
in
ramiccomponentof a glass-ceramic
hostfor immobi- H2SO4wasenriched
lizing radioactivewastes. To confirm the expected leachedlayer thicknessfor a given glasscomposition
thermodynamicstability of this mineral in groundwa- in the two acids suggeststhat the anions play an
ter environments,glass ceramicswere leachedfor up importantrole in glassdissolutionkinetics.
Fieldstudies. A numberof reportshavebeen pubto 1 year in synthetic saline groundwaterincluding
15,500mg/L Ca, 5047mg/L Na, and33,410mg/L C1at lished describingthe surfaces of medieval glasses
pH 6-8 [Hayward et al., 1984].SIMS analysesof the weathered in the ambient air for 600-700 years
glass-ceramicleachingin aleionizedwater showedde- [Schreineret al., 1984, !988; Schreiner, 1988, 1989].
pletionsof Ca at depthsup to 500nm andNa, Cs, and These glasseswere producednorth of the Alps beSr to depthsgreaterthan900nm; Si, Ti, Ce, La, andU tween the years 1100and 1500and are characteristiwere enriched at the surface. These results suggested callylow in SiO2(40-50 wt %) andhighin K20 (10-20
that the glassphasedissolvedpreferentially,with the wt %), with 10-15% CaO. The dark green stained
subsequentprecipitationof Si, Ti, Ce, La, and U. In glassesproducedduringthe sameperiodcontain,in
contrast, the synthetic groundwater effectively sup- additionto the above constituents,5-10% P205 and
pressed leaching of the titanire; in addition to the 10-30% PbO. Whereas the potash-limeglassshowsa
32, 2 / REVIEWSOF GEOPHYSICS
Shotykand Metson' SECONDARYION MASS SPECTROMETRY
ß 209
low chemical durability, the dark green versions are
much more stable [Schreiner et al., 1984].
Using an O- primary beam, Au-coated naturally
Si
weathered
glasses
weredepthprofiledfor •H+, 23Na+,
27A1
+, 29Si+,4•K+, 42Ca+,65Cu+,•38Ba+,and2øSpb+.
To convert the plot of secondary ion intensities as a
function of sputter time to concentrationdepth profiles, the "relative sensitivity factor" method was used
in conjunction with measured sputtering rates O
[Schreineret al., 1984].The sensitivityfactor (Sx/•a)
relatesthe measured
ion intensity(I•a) to the atomic
concentration(c•) as follows:
SX/M= [(Ix/cx) fx]/[(Im/cm) frei
where I is the secondary ion signal, c the atomic
concentration, and f the isotopic abundance of the
analytical (X) or reference element (M); in this case, Si
was selected as the reference
element.
The relative
sensitivity factor method has been discussed elsewhere [Metson, 1990]. The concentration-depthprofiles revealed a surfacelayer apparently enrichedin H,
Si, and A1 (the "gel zone") and depleted in K and Ca.
Schreiner et al. [1984] assumed that the dominant
processwas the following ion exchangereaction:
-=Si-O-M + + H + *->--Si-OH
and further
that the atomic
+ M+
concentration
of Si re-
mained constant during dissolution. The relative enrichment of Si at the surface was interpreted in terms
of an increasedsputteringrate of Si in this zone. The
concentration-depthprofiles were then correctedfor
this and are shownin Figure ! 1a. The leachedlayer is
1
2
3
Depth (pm)
Figure 11. (a) Concentration-depthprofiles of K, Fe, and
Ba, relative to Si, in medieval glass !9 naturally weathered
since the fifteenth century. An altered layer has developed
which is stronglydepletedin K and Ba, as well as Na and Ca
(not shown), and slightly depleted in Mg, A1, and Mn (not
shown) in relation to Si. Iron is neither enriched nor de-
pleted. (b) Concentration-depth
profilesof K, P, Fe, Ba, and
Pb, relative to Si, in dark green medieval glass 22 naturally
weathered since the fourteenth century. An altered layer has
developedwhich is stronglydepleted in K and Ba, as well as
Ca (not shown), and slightly depleted in Pb, as well as Mg,
A1, and Mn (not shown), in relation to Si. Again, Fe in the
leached layer is unchangedin comparisonto the bulk com-
position. NoQce the enrichment of P in the leached layer.
(From $chreiner [1988]. Reprinted with permission.)
approximately1.5 Ixmthick and significantlydepleted
in K, Ca, and Pb. The oqtermost layer, however,
revealed an enrichment of many elements, including
Pb, which may help to explain the durabilityof this
glass.Th{soutermost
layerwasalsoc!eadyenriched
in C1 and S in relation to the bulk glass.
The constantSi assumptionwas confirmedin subsequent studies of these glasses [Schreiner, 1988;
$chreiner et al., 1988]. Shorter analysis times and
subsequentcrater depthmeasurementswith a profilometer showed that the increased intensities of Si near
the surfacewere caused'partlyby a highererosionrate
of that layer by the primaryion beam. Using highmass
resolution, Schreiner [1988] showed that the intensity
of 3•p+wasmorethan 10timesthat of 3øSi•H+, allowing P to be depthprofiledquantitatively(Figure1lb).
The P depthprofilesshowedthat the leachedlayer
SputterTime (s)
formed on the dark green, phosphorus-containing
Figure10.(a) RawSIMS depthprofileof •H, 4•K, 298i,and glasswas significantlyenrichedin P [Schreiner,1988].
35C1of glasscontaining
15 wt 5½K20, 25%CaO,and605½ The H depthprofilesobtainedby depthprofilingwith
2000
4000
2000
4000
SiO2leachedin 0. ! N HC1 for 1 hour. An alteredlayer has SIMS [Schreiner et al., 1984] were confirmed using
developedwhichis depletedin K but enrichedin H, Si, and NRA [Schreiner, 1988; Schreiner et al., 1988].
C1,aswell as Ca (not shown).(b) Raw SIMS depthprofileof
A radwasteglassconsistingof 85 wt % nepheline
•H, 4•K, 29Si,and 32Sin the sameglasscomposition,
but
leachedin 0.1 N H2SO4 for 1 hour.The alteredlayerwhich
hasdevelopedis similarlydepletedin K and enrichedin H
andSi, as well as Ca (not shown),but the depthof alteration
is approximately
onehalfthat producedbY HC1.Noticethe
enrichmentof S. (From Schreiner[1989]. Reprintedby permission of the American Ceramic Society.)
syeniteand 15% CaO was analyzedusingSIMS after
leachingin groundwaterfor 28 years[Tait et al., 1986].
The chemicalcompositionof the freshglasswas51.2%
SiO2,20.4%A1203,15.0%CaO, 8.5%Na20, and4.3%
K20. SIMS analysesof the leachedglassrevealedthe
following:A1, Ca, and Na were depletedto depthsof
210 ß Shotykand Metson' SECONDARYION MASSSPECTROMETRY
approximately 20, 35, and 90 nm, respectively, and Si
enriched to 60 nm. Iron and Mg were slightly enriched
to depths of 45 nm, possiblybecauseof accumulation
from groundwater. Uranium was found to be slightly
enrichedand Th highly enrichedto depthsof about 100
nm; the Th profile was interpreted in terms of a residual enrichment caused by the preferential release of
the major glass constituents A1, Ca, and Na [Tait et
32, 2 / REVIEWSOF GEOPHYSICS
for 60 days in HC! at p H 3.5 was significantlydepleted
in A1 and Ca in relation to Si, for approximately 600 s.
Giventhemeasured
sputterratesof 1-2 •/s, theresultsindicated a cation- and A!-depleted leached layer
approximately
600-1200
• thick.In contrast,
leaching
the feldspar in distilledwater at p H 5.7 producedonly
very thin leached layers: A1 was depleted for 60-120
•, andCa for 100-200•. UsingXPS, Muir et al.
[1989]providedindependentconfirmationthat the surface of dissolvinglabradorite was also depleted in Na.
4.2.
Dissolution of Silicates
Cation- and Al-depleted leached layers on the surface of plagioclasefeldspar dissolvingat room temperLabstudies. SIMS depthprofilesof a titanitewhich ature have sincebeen reported in several studies(see
hadbeenleachedat pH 4 for 28 daysat 90øCshowedthat Casey and Bunker [!991] for a review). The steps
Ca and Fe were stronglydepletedin relationto Si and Ti involved in the formation of these residual layers are
[Bancroft et al., 1987]. On the same mineral leachingat thoughtto be [Casey et al., 1989a, b] (1) penetration
pH 8 in solutionscontaining400 mg/L Ca and saturated of water molecules(hydration), (2) protonation of catwith respect to dissolved Si, no leached layers were ion-oxygenbondsnear the surface of the mineral (the
observed;the lack of significantdissolutionin these depth of penetrationdependsupon p H, reaction time,
experimentswas confirmed by inductively coupled and temperature), (3) replacement of alkali and alkaplasma analysesof the leachingsolutions.In long-term line earth cations near the mineral surface by hydroleachingexperiments,SIMS depth profilesshowedsig- nium ions, (4) hydronium ion attack of A1 in bridging
nificant surfaceenrichmentsof Ca, Si, and A1, suggest- Si-O-AI sites, thereby depolymerizingthe surface, (5)
ing the regrowthof a phaseapproachingtitanitein com- diffusion of alkali, alkaline earth cations, and A1
position.SIMS depthprofilesof mtile (TiO2)leachingin through the reacting surface layer, (6) partial repolysolutionscontainingtitanite showedthat Ca, Si, A1, La, merization of remainingsilanolgroupsin the silica-rich
and U were actually enriched by up to 2 orders of surfaceto produce Si-O-Si or Si-O-A1 bonds, and (7)
magnitudein comparisonto bulk values;theseprofiles hydrolysis of the Si-O-Si bonds in the remaining siliconfirmedthat under appropriate conditions,not only ceoussurfacelayer. This last stepis thoughtto control
does titanite not dissolve, it may actually grow. The the overall rate of feldspar dissolutionin acidic ch!otitaniteswerealsoleachedin D20, andD + penetration ride solutions [Chou and Wollast, 1984, 1985; Wollast
followedusingSIMS. The depthprofilesshowedthat Ca and Chou, 1985; Brady and Walther, 1989; Casey et
depletion
wasrestricted
to thefirst300,l,, whereas
D al., 1989a, b; Blum and Lasaga, 1991].
Effect of plagioclasecomposition: Albite, oligopenetrated at least !0 times as deep [Bancroft et al.,
!987].
c!ase, labradorite, and bytownite were leached in HC1
In experimental studies of diopside dissolutionin at p H 3.5 for 90 days and depth profiled using the
deionized water at 100øCand pH 5.5, Schott and Petit SIMS [Muir eta!., 1990]. In each case, significant
[1987] used both SIMS and NRA to follow the reac- leached layers were found in which the cations and A1
tion. The SIMS depth profiles showedthat Ca and Mg aredel•leted
inrelation
toA1overdistances
asgreatas
were depletedin relation to Fe and Si, and H enriched, 1600 A. The thicknesses of the leached layers infrom the surface of the mineral over distancesof up to creasedwith increasingA1 and decreasingSi content
2000•. Theprofiles
wereinterpreted
ashaving
docu- (i.e., increasingmole percent anorthite) as shown in
mentedthe formation of a hydrated silicatedepleted in Figure 12. The SIMS findingsare consistentwith kiMg and Ca, but enriched in Fe owing to the precipita- netic studies which have shown that the reactivities of
tion of ferric oxide on the surface.
p!agioclasefeldsparsin acidic solutionsincreasewith
Potassium feldspar leached at p H 1 from 25ø to increasingA1 content [Blum and Lasaga, 1988;Brady
200øCwas analyzed using SIMS and a variety of other and Walther, 1989].
The thickness of leached layers may vary signifisurfacetechniques[Goossenset al., 1989]. The SIMS
depth profiles revealed an amorphoussurfacelayer up cantly with small differencesin plagioclasecomposito 75 nm thick, depletedin Na, A1, and K (plusRb, Cs, tion [Shotyk and Nesbitt, 1992]. For example, comSr, Ba), and enriched in H. The thickness of this pare the leachedlayersformed on an An54labradorite
residual layer was found to increase with increasing versusan An60!abradoritecomposition,both leached
reaction time and increasing temperature. SEM and in HC1at pH 4.0 (Figure 13). The leachedlayer formed
transmissionelectron microscopy images supported on the more calcic labradorite compositionis about 5
the view that the leachedlayer was a residualproduct times thicker than that formed on the more sodic
composition.Once again, theseresultsare consistent
of dissolution
and not an amorphous
precipitate.
al., 1986].
Dissolving
plagioclase
feldspars: Muir et al. [1989] with recent kinetic studies which have shown that the
showedthat the surfaceof labradoritefeldsparleached rates of dissolutionof the calcic plagioclasesare very
32, 2 / REVIEWSOF GEOPHYSICS
Shotykand Metson' SECONDARYION MASS SPECTROMETRY
ß 211
-20
l--"• -20
......
_;60•1
...... a ;6000
300
_2o
600
-
surface layer.
The SIMS depth profiles support kinetic studies
which have demonstrated the inhibiting effect of dissolved salts on feldspar dissolution rates [SjOberg,
1989]. The rate of dissolution of oxides and silicates is
600
-60
300
600
100
d
300
600
As the relative importanceof H + in the bulk solution
decreased, so too did the thickness of the residual
_20
-60
100
300
plus Na, K, Ca, and I mg/L AI, with either 1 or 0.1
mg/L Si), the contributionof H + to the total cationic
charge was reduced to approximately 30% (Table 3).
. .
SputterTime (s)
proportionalto the surfaceconcentrationof H + [Blurn
and Lasaga, 1991; Casey et al., 1992]. The presenceof
cations other than H + in the bulk solution results in
additional adsorption equilibria which compete with
H + for surfacesites[Sturnmand Wollast, 1990].Add-
Figure 12. Percentage change in A1/Si of four plagioclase ing Si to the solutions (Table 3) had no significant
feldspar compositions leached for 90 days in HCI at pH 3.5:
(a) albite, (b) oligoclase, (c) labradorite, (d) bytownite.
(After Muir et al. [1990].)
sensitive to small differences in chemical composition.
For example, Casey et al. [1991] found that the rates of
effect on leached layer formation (Figure 14) because
aqueoussilicadoesnot competewith H + andthe other
cations for surface
sites.
Effectof complex-formingligands: SIMS analyses
showedthat labradorite (An54and An60)leachedat p H
4 for 72 daysdissolvedincongruently
bothin 10-4 M
dissolutionof labradorite (An6o)and bytownite (An76) oxalicacidandin 10-4 M HF, yieldingalteredlayers
were14and212x 10-•5 molSi cm-'-s-l, respectively.
Effect of dissolved
salts:
The
effect
of dissolved
salts (Na, K, Ca, AI, and Si) on the formation of
leached layers on the surface of labradorite feldspar
has been evaluated in a series of simple experiments
[Nesbitt et al., 1991]. The experimental conditions
selected for study are shown in Table 3. Leached
-20 •'••
-40 •
layersup to 15003, thickdeveloped
on labradorite
(An54)
dissolving
in 10-4 M HCI atpH 4 (Figure14).In
solutions at the same p H containing 1 mg/L Na, K,
and Ca, the leached layers formed after the same
period of dissolution were about one half as thick
-60.L•
•
-8O ...•
',
a
-1000 ' 200
4•0
6•0
800
(about7003,) as thoseformedin HCI alone.In solutions at pH 4 containing these cations and also 0.01
mg/L A1, a leached layer formed which was indistin-
guishable
fromthesecond
case(about7003,).In other
words, adding0.01 mg/L A1 had no significanteffect on
leached layer formation at p H 4 in comparisonto the
combinedeffect of 1 mg/L Na, K, and Ca. In solutions
containingthese cationsplus 1 mg/L A1, no significant
leachedlayer formed (Figure 14). In the solutionat pH
4 containing1 mg/L Na, K, Ca, and AI, therefore, the
labradorite dissolved congruently.
In theexperiment
consisting
onlyof 10-4 M HCI,
there were no significant concentrationsof cations
other than H +. The hydronium ion, therefore, accounted for 100% of the total cationic charge of the
solution (Table 3). In the next experiment (HC1 plus
Na, K, Ca), the addition of other cationsreducedthe
contribution of H + to total cationic charge to 45%
(Table 3). In the experimentcontainingHC1, Na, K,
Ca, plus 0.01 mg/L A1, the addition of A1 did not
significantly
reducethecontribution
of H + to thetotal
-2O
-4O
-6O
-1000
460'' '8(•0'' i2'0(•' i600
SputterTime (s)
Figure13. (a) Percentchangein A1/Siof labradorite(An54)
leachedfor 72 days in HCI. Four spotson the surfaceof the
samplewere depth profiled.Given that the sputteringrate
wasapproximately
2 3ds,an Al-depleted
leached
layerapproximately
800• thickhasformedonthe surface
of the
dissolvingfeldspar. (b) Percent changein A1/Si of labradorite(An6o)leachedfor 72 daysin HCI at pH 4. Two spots
on the mineral surface were depth profiled. Here, AI is
depletedin relation to Si for 1500 s, correspondingto a
layerapproximately
3000/•thickFromShotyk
and
cationicchargein comparisonto the solutioncontain- leached
ing Na, K, and Ca. In the last two experiments(HCI Nesbitt [1992]. (Reprinted with permission.)
212 ß Shotykand Metson- SECONDARYION MASSSPECTROMETRY
32, 2 / REVIEWSOF GEOPHYSICS
TABLE 3. DissolvedSolidsConcentrations
and Distributionof CationicChargein Labradorite(An$4)Leaching
Experimentsin HCI at pH 4 [Nesbittet al., 1991]
Experiment 1
Experiment 2
Percentof
Experiment 3
Percentof
Experiment 4
Percentof
Percent of
Concentration,
Cationic
Concentration,
Cationic
Concentration,
Cationic
Concentration,
Cationic
mg/L
Charge
mg/L
Charge
mg/L
Charge
mg/L
Charge
H+
Na +
K+
Ca2+
Air
ßßß
1.0
1.0
1.0
ßßß
45.7
19.9
11.7
22.8
0.0
45.4
19.8
11.6
22.7
0.5
ßßß
1.0
1.0
1.0
1.0
30.3
13.2
7.8
15.1
33.6
' ßß
1.0
1.0
1.0
1.0
30.3
13.2
7.7
15.1
33.6
Sir
ßßß
0.0
ßß'
1.0
1.0
1.0
0.01
1.0
0.0
1.0
0.0
0.1
0.0
•)9ecies
stronglydepletedin A1 and Ca and residuallyenrichedin depthprofilesobtainedfrom An60labradoriteleachedin
Si. The thickestleachedlayerswereproducedby HF (up oxalic acid and in HF were normalized to those proto 1700• onAn54
andupto5000• onAn60).
Becauseduced in responseto HC1. This comparisonclearly
the H + concentration was the same in each solution showsthat the ligandsoxalate and fluoride significantly
(10-4 m), it waspossible
to clearlyseparate
therelative increasedthe formationof leachedlayersin comparison
importanceof proton-promoteddissolution(in HCI) and to the effectof H + alone(Figure15).
ligand-promoteddissolution (in oxalic acid and in HF)
for a given feldsparcomposition.To do this, the SIMS
!
o
(D
-20
2O
-40
-60
_80
[
c
13_
-100
0
500
1000
1500
2000
•i'
100
0
a
40o
800
12'00'1600
_c
-40
'• 0 -60
•
13__
-100
0
500
1000
1500
2000
Depth
0
400 800 1200 1600 2000 2400
SputterTime (s)
Figure 15. (a) Percent changein A1/Si of labradorite (An6o)
Figure 14. Percentage change in (a) A1/Si and (b) Ca/Si of
labradorite (An54) leached in HC1 at pH 4 with varying
concentrationsof dissolved salts. The compositionof each
solution is shown in Table 3. The thickness of the leached
layer formed in the solution containing Na, K, and Ca
(experiment 1) is about one half that formed in HC1, even
though the p H is the same in each case. The addition of 1
mg/L Si and 0.01 mg/L AI to the solution containingthe
cationshas no effect (experiment2), in comparisonto the
cationsalone.In solutionscontainingthe cationsplus1 mg/L
A1,no significantleachedlayer is found(experiment3). With
1 mg/L A1pluseachof the cations,increasingSi from 0.1 to
1 mg/L had no effect on leached layer thickness.(After
Nesbitt et al. [1991].)
leached in oxalic acid, normalized to labradorite leached in
HCI. The pH and acid concentrationswere the same in each
solution, allowing the effects of proton-promoted and ligandpromoted dissolutionto be clearly separated. The leached
layer shown here has formed in response to the oxalate
ligand. The differences between the three depth profiles
probably reflect chemical zonation in the mineral (b) Percent changein A1/Si of labradorite (An6o) leached in HF,
normalizedto labradoriteleachedin HCI. Again, the p H and
acid concentrationswere the same in each solution, allowing
the effectsof proton-promotedand ligand-promoteddissolution to be separated.The extent of leaching shown here can
be attributed to the fluoride ligand. (From Shotyk and Nesbitt [1992]. Reprinted with permission.)
32, 2 / REVIEWSOF GEOPHYSICS
Shotyk and Metson: SECONDARYION MASS SPECTROMETRYß 213
The three spots which were depth profiled show
approximatelythe samedepth of leaching, but clearly
differ in respect to the extent of surface attack (Figure
15). These differences in extent of leaching within a
sample are attributed to zoning, which is commonly
observed within feldspar crystals [Barth, 1969]. The
less extensive dissolution probably corresponds to
more sodic zones within the crystal, and the more
extensively dissolved areas to more calcic zones. Inskeep et al. [1991] have observed similar variation in
leached layer thicknesseson the surface of labradorite
leached in acidic solutions. They reported leached
-20
.
-40
-
'20'0 '460 '660 '860 'l oho
c-'"
• -10f**
½'**•ti•11•
il•"•'
fro-o. b
layersup to 700/ttin the sodiczonesof labradorite,
compared
to 1400/ttin theleached
layersof thecalcic
-•
oxalic
-30" "2dO '460 '660 '860 'i000
zones.
In contrast to the two labradorite compositions,
anorthite
leached in HC1 showed much thinner leached
layers, even though it is a more reactive composition.
Also, no significant leached layers were observed on
anorthite leached in oxalic acid and in HF (Figure 16).
Significantly higher concentrations of Si and A1 were
found in the oxalic acid and HF leaching solutions in
comparison to anorthite in HCI, indicating a more
extensive dissolution of the solid phase. However, in
each case the molar ratio of Si to A1 in the solutions
reflected the molar proportion of Si to A1 in the fresh
solid. In other words, in each of these experiments,
dissolution of the anorthite was essentially congruent
[Shotyk and Nesbitt, 1992].
These findings may help to explain why Mast and
Drever [1987] found that 1-mM oxalic acid had no
effect on the overall rate of dissolution of oligoclase
-20 *
30.........200
400
600
860 '1000
SputterTime (s)
Figure 16. (a) Percent changein AI/Si of anorthite (An,oo)
leached in HCI at pH 4 for 72 days. This leached layer is
significantlythinner than that of the An6olabradorite composition shown in Figure 12. However, the measuredconcentrationsof AI and Si in the output solutionsconfirmed
that the anorthite was dissolving more rapidly than the labradorite. (b) Percent changein AI/Si of anorthite leachedin
oxalic acid at the samepH. Although the percentchangein
AI relative to Si is less than that produced in HCI, the
concentrations of AI and Si in the oxalic acid output solutions are significantlyhigherthan in the HCI, confirmingthat
oxalic acid has enhanced the dissolution of anorthite, in
feldspar(An•3) while Amrhein and Suarez [1988]reported an acceleratedrate of anorthite (An93)dissolu- comparisonto HC1 at the samepH. (c) Percentchangein
tion in the presenceof similar concentrationsof oxalic AI/Si of anorthite leached in HF. Again, even though the
acid. The effect of complex-formingorganicligandson
the overall rate of plagioclasefeldspar dissolution,as
measuredby the rate of releaseof Si, dependsstrongly
upon the molar proportionsof A1 and Si in the solid.
Promotion of rates of feldspar dissolutionby organic
ligandsis favored by a relatively large ratio of A1 to Si
(e.g., 1:1 as in the case of the anorthite endmember).
Field studies. Samplesof naturally weatheredoligoclasefrom a soil profile weatheringfor 10,000years
have been depth profiledusingthe SIMS [Nesbitt and
Muir, 1988].Naturally weatheredfacesof the feldspar
were depth profiled and compared with freshly cut
surfaces.The freshly cut oligoclasesurfacerevealed
leachedlayer producedby HF is smallerthan the slightlayer
producedin responseto HCI, the concentrations
of A1andSi
in the HF output solutionsare significantlygreater than in
HCI at the same pH. (From Shotyk and Nesbitt [1992].
• (Figure17).At thesametime,however,
theSIMS
environment and all of the laboratory studiesis the p H
Reprinted with permission.)
dissolveincongruently,with Si being leached out of
the mineral preferentiallyover A1, they failed to ex-
plain how Si couldhave been removedwithoutalso
removing Ca.
In contrastto the naturallyweatheredoligoclase,all
of the feldsparsleachedin acidic solutionsin experimentallaboratorystudiesshowalteredlayersin which
no significantdifferencein Si, A1, and Ca ion intensi- A1 is depletedin relationto Si [Caseyet al., 1988,
ties with depth throughthe surfaceof the mineral.In 1989a, b; Goossenset al., 1989; Muir et al., 1989,
contrast,the depthprofilesof ninenaturallyweathered 1990;Hellmann et al., 1990;Nesbitt et al., 1991;Muir
plagioclasegrains all showed that Si was strongly and Nesbitt, 1991, 1992; Shotyk and Nesbitt, 1992].
depletedin relationto AI over a distanceof up to 1000 The main difference between the natural weathering
depthprofilesclearlyshowedthat Ca in the weathered and compositionof the solution in the till which
grainswas not depletedin relationto AI [Nesbittand leachedthe feldsparfor thousandsof years. Although
Muir, 1988; Muir, 1989]. Although the authors con- nopH or solutionchemicalinformationis givenin the
cludedthat naturally weatheredplagioclasefeldspars originalpaper[Nesbittand Muir, 1988],if the pH of
214 ß Shotykand Metson: SECONDARYION MASSSPECTROMETRY
(D
r-
20
SIMS depth profiles of dissolving glassesand silicates illustrate the complexity of the dissolutionprocess. The behavior of dissolvingplagioclasefeldspar,
for example, is very sensitiveto both the composition
of the mineral and the leaching solution. Small differences in mineral compositionmay give rise to large
0
r- •-20
o
-40
O
-6O
(D
-80
12_
differences
200
400
600
800
1000
Depth(,4,)
Figure 17. Percent change in the ratio Si/A1 versusdepth of
naturally weathered plagioclasegrains. The dashed curve
corresponds to a fresh (unweathered) otigoclase and the
other curves correspondto three weatheredgrains.The low
28Sisecondary
ion intensities
in relationto 27A1in the raw
SIMS depth profileswere interpretedas evidenceof preferentiaI removal of Si in comparisonto A1from the surfaceof
the fetdsparby natural weathering solutions.The calculated
Si/A! ratios emphasizethis change.(From Nesbitt and Muir
[1988]. Reprinted with permission.)
the soil solutionat the base of the profile was sufficiently closeto the neutralpH range,perhapsSi could
havebeenreleasedpreferentiallyover A1aslab experiments have shown [von Engelhardt, 1939; Schweda,
1989]; however, this could not account for the Ca
depthprofile. A secondpossibilityis that the presence
of dissolved cations and A1 in the leaching solution,
combined with a near-neutral p H, allowed Si to be
removed preferentially over A1 as feldsparsdissolving
in salt solutions [Nesbitt et aI., 1991]. Again, this
probablywould not explain the Ca depth profile.
Perhapsthe SIMS depth profiles simply indicated
an aluminousprecipitatecoveringthe feldsparsurface.
Aluminous coatingshave been found on the surfaceof
naturally weathered minerals from soils, including
hornblende [Mogk and Locke, 1988] and oligoclase
[Inskeep et al., 1993]. Given that the fundamental
processof podzolizationis the chemicalseparationof
Fe and A1 from Si in upper soil layers and the precipitation of amorphousFe and A1phasesfurther down in
the profile, this latter possibilityseemsthe mostlikely.
Unfortunately, there appear to be no other published
SIMS studies of naturally weathered minerals from
soilsfor comparison.
5.
SUMMARY
AND
32, 2 / REVIEWSOF GEOPHYSICS
CONCLUSIONS
in rates of dissolution
and thickness
of
leached layers. The presenceof relatively low concentrations of dissolved salts may greatly influence both
the rate and the mechanismof dissolution. Simple,
complex-formingorganic ligands may promote the release of A! in acidic solutions,but this may or may not
affect the overall rate of feldspar dissolution,depending on the mole percent anorthite.
Many questionsremain concerningrates and mechanisms of mineral weathering in soils and sediments.
The effect of organic acids on silicate dissolution, for
example, requires systematic study to clearly separate
the effects of simple and complex organic acids as
ligandsversusreducingagents.The combinedeffect of
complex-forming organic ligands and dissolved salts
on the formation of leached layers and dissolution
rates has not yet been investigated. There is a need for
more detailed experimental studies of the rates and
mechanisms of dissolution in systems more closely
resemblingnatural weathering solutions, using representativeconditionsof mineral composition,temperature, p H, and solutioncomposition.Such experiments
shouldinclude analysesof output leaching solutions,
as has been done traditionally, but should also include
surfaceanalysesof the dissolvingsolid usinga method
suchas the SIMS. Given the potential of the SIMS to
help us understand processes at the mineral-water
interface, there is also a great need to do much more
work with reacted surfaces of naturally leached minerals from soils.
ACKNOWLEDGMENTS. The SIMS depth profiles of
dissolvingfeldspars were performed at the University of
Western Ontario and supportedby the Natural Sciencesand
Engineering Research Council of Canada and the Ontario
Ministry of the Environment. The staff at Surface Science
Western (UWO) provided expert technical support. Subsequent weathering studies at the University of Berne are
funded by the Swiss National Science Foundation (grant
21-30207.90). We are grateful to Urs Eggenberger (Berne)
and Paul van der Heide (Auckland) for help in preparing the
figures.
Garrison Spositowas the Editor in charge of this paper.
He thanks Christopher Amhrein and William Casey for tech-
SIMS is extremely surface sensitive and allows
chemical and isotopic analysesof elementsranging nical reviews and Alan Chave for a cross-disciplinaryreview.
from H to U. Using the specimenisolationmethodto
reduce surface charging, SIMS is ideally suited to
REFERENCES
depthprofilingglassesand minerals.With sputtering
ratesof theorderof 1 •/s, SIMSdepthprofiles
have
been used successfullyto illustratechemicalchanges
that take place at the solid-solutioninterface during
the dissolutionof glassesand minerals.
Aagaard, P., and H. C. Helgeson, Thermodynamic and kinetic constraintson reaction rates among minerals and
aqueous solutions, I, Theoretical considerations,Am. J.
$ci., 282,237-285,
1982.
32, 2 / REVIEWSOF GEOPHYSICS
Shotykand Metsen: SECONDARYION MASSSPECTROMETRY
ß 215
Amrhein, C., and D. L. Suarez, The use of a surface complexation model to describethe kinetics of ligand-promoted dissolution of anorthite, Geochim. Cosmochim.
Acta, 52, 2785-2793, 1988.
Andersson,L.-H., and B. Lindqvist, Some experimentson
the interaction between feldspars and salt solutions,
Geol. Foeren. Stockholm Foerh., 78, 459-462, !956.
Andr6,,G., D6placement de la potassecontenuedans certainesrochesfeldspathiquespar quelquessubstances
employ6escommeengrais,C. R. Hebd. SeancesAcad. Sci.,
157, 856-858, 19!3.
Antweiler, R. C., and J. I. Drever, The weatheringof a late
Tertiary volcanic ash: Importance of organic solutes,
Geochim. Cosmochim.Acta, 47, 623-629, 1983.
April, R., R. Newton, and L. Truettner Coles, Chemical
weathering in two Adirondack watersheds: Past and
present-day rates, Geol. Soc. Am. Bull., 97, 1232-1238,
1986.
rare earth elementsin sixteen silicate reference samples
by secondaryion massspectrometryusing conventional
energyfilteringtechnique,Geostand.Newsl., 15, 51-57,
1991.
Bottazzi, P., L. Ottolini, and R. Vannucci, SIMS analyses
for rare earth elementsin natural mineralsand glasses:
An investigation
of structuralmatrix effectson ion yields,
Scanning, 14, 160-168, 1992.
Brady, P. V., and J. V. Walther, Controls on silicate disso-
lution rates in neutral and basic p H solutionsat 25 C,
Geochim. Cosmochim.Acta, 53, 2823-2830, 1989.
Brantley, S. L., Kineticsof dissolutionand precipitationm
Experimental and field results, in Water-Rock Interaction, editedby Y. K. Kharaka and A. S. Maest, pp. 3-6,
A. A. Balkema, Rotterdam, Netherlands, 1992.
Brongniart, A., De la cavit6 veineuse et de ses poches
accessoires:Sur !e nautile flamb6, Arch. Mus. Hist. Nat.
Paris, 2,285-314, 1841.
Arehart, G. B., C. S. Eldridge, S. L. Chryssoulis,and S. E. Brookins, D. G., Geochemical Aspects of Radioactive
Kesler, Ion microprobedeterminationof sulfur isotope
Waste Disposal, Springer-Verlag, New York, 1984.
variations in iron sulfides from the ?ost/Betze sedimentBrown, A.D., and L. J. Lund, Kinetics of weathering in
hosted disseminated gold deposit, Nevada, USA,
soils from a subalpinewatershed, Soil Sci. Soc. Am. J.,
Geochim. Cosmochim. Acta, 57, 1505-1519, 1993.
55, 1767-1773, 1991.
Bancroft, G. M., J. B. Metson, R. A. Kresovic, and H. W. Busenberg,E., The productsof the interaction of feldspars
Nesbitt, Leaching studiesof natural and synthetictitanwith aqueous solutions at 25 C, Geochim. Cosmochim.
Acta, 42, 1679-1686, 1978.
ites using secondaryion mass spectrometry, Geochim.
Cosmochim. Acta, 51, 9 ! 1-9 ! 8, 1987.
Busenberg, E., and C. V. Clemency, The dissolutionkinetBarth, T. F. W., Feldspars, Wi!ey-Interscience, New York,
ics of feldsparsat 25 C and I atm CO2 partial pressure,
1969.
Geochim. Cosmochim. Acta, 40, 41-49, 1976.
Bear, F. E., Chemistry of the Soil, 2nd ed., Reinhold, New Carroll, D., Rock Weathering, Plenum, New York, 1970.
York, 1964.
Casey, W. H., and B. Bunker, The leachingof mineral and
Benninghoven, A., F. G. Rudenauer, and H. W. Werner,
glasssurfacesduring dissolution, in Mineral-Water InterSecondary Ion Mass Spectrometry, John Wiley, New
face Geochemistry, Rev. Mineral., 23,397-426, 1991.
York, 1987.
Casey, W. H., H. R. Westrich, and G. W. Arnold, Surface
Berner, E. K., and R. A. Berner, The Global Water Cycle:
chemistry of labradorite feldspar reacted with aqueous
Geochemistry and Environment, Prentice-Hall, Englesolutionsat pH = 2, 3, and 12, Geochim. Cosmochim.
wood Cliffs, N.J., 1987.
Acta, 52, 2795-2807, 1988.
Berner, R. A., and G. R. Holdren, Jr., Mechanism of feld- Casey, W. H., H. R. Westrich, G. W. Arnold, and J. F.
Banfield, The surface chemistry of dissolvinglabradorite
spar weathering:Some observationalevidence, Geology,
5, 369-372, 1977.
feldspar, Geochim. Cosmochim. Acta, 53, 821-832,
1989a.
Berner, R. A., and G. R. Ho!dren, Jr., Mechanism of feldspar weathering, II, Observationsof feldsparsfrom soils, Casey, W. H., H. R. Westrich, T. Massis, J. F. Banfie!d, and
Geochim. Cosmochim. Acta, 43, 1173-1!86, 1979.
G. W. Arnold, The surface of labradorite feldspar after
Berner, R. A., G. R. Holdren, Jr., and J. Schott, Surface
acid hydrolysis, Chem. Geol., 78, 205-218, 1989b.
layers on dissolvingsilicates:Commentson "Study of the Casey, W. H., H. R. Westrich, and G. R. Holdren, Jr.,
Dissolution rates of plagioclase at p H = 2 and 3, Am.
weathering of albite at room temperature and pressure
Mineral., 76, 211-217, 1991.
with a fluidized bed reactor" by L. Chou and R. Wollast,
Geochim. Cosmochim. Acta, 49, 1657-1658, 1985.
Casey, W. H., C. Eggleston, P. A. Johnsson, H. R. WeBetz, G., and F. G. Rudenauer, Collision cascadelimit to
strich, and M. F. Hochella, Jr., Aqueous surface chemistry and corrosion of minerals, MRS Bull., I7(5), 23-29,
spatial resolution in focused ion beam processes,Appl.
1992.
Surf. Sci., 51, 103-!12, 1991.
Beyer, A., Ueber die ZersetzungdesFeldspathesunter dem Casey, W. H., J. F. Banfield, H. R. Westrich, and L.
McLaughlin, What do dissolution experiments tell us
Einflussvon Salz16sungenund einigenanderenAgentien,
about natural weathering?, Chem. Geol., 105, 1-15, 1993.
Chesworth, W., Weathering systems, in Weathering, Soils,
and Palaeosols, edited by I. P. Martini and W. Chesworth, pp. 19-40, Elsevier, New York, 1992.
1974.
Chin, P.-K. F., and G. L. Mills, Kinetics and mechanismsof
kaolinite dissolution: Effects of organic ligands, Chem.
B!aser,P., H. F1/iher, and T. Sidler, R/ickkoppelungsmechGeol., 90, 307-317, 1991.
anismen beim Stofftransport:Verlagerung yon Fe(III)
Chou, L., and R. Wollast, Study of the weathering of albite
durch wasser16sliche Streusubstanzen, Mitt. Dtsch.
at room temperature and pressure with a fluidized bed
Bodenk. Ges., 30, 111-122, 1981.
reactor, Geochim. Cosmochim. Acta, 48, 2205-2217,
Blum, A., and A. Lasaga,Role of surfacespeciationin the
1984.
low-temperature dissolution of minerals, Nature, 33!,
Chou, L., and R. Wollast, Steady-state kinetics and disso43 !-433, 1988.
Blum,A. E., andA. C. Lasaga,
•he roleof surface
specia- lution mechanism of albite, Am. J. Sci., 285, 963-993,
!985.
tion in the dissolution of a!bite, Geochim. Cosmochim.
Chryssoulis,S. L., Quantitativetrace preciousmetal analyActa, 55, 2193-2201, 199!.
sis of sulfideand sulfarsenideminerals by SIMS, in SecBottazzi,?., L. Ottolini, andR. Vannucci,Determinationof
Landwirtsch. Vers. Stn., 14, 314-322, 1871.
Blaser, ?., Mechanismen der Eisenaufnahme und-verlagerungdurch wasser16slicheStreusubstanzenin podsoligen Boden, Mitt. Dtsch. Bodenk. Ges., 20, 447-457,
216 ß Shotyk and Metson: SECONDARYION MASS SPECTROMETRY
ondary Ion Mass Spectrometry, SIMS 7, edited by A.
Benninghoven, pp. 405-408, John Wiley, New York,
1990.
Colman, S. M., and D. P. Dethier (Eds.), Rates of Chemical
Weathering of Rocks and Minerals, Academic, San Diego, Calif., 1986.
Correns, C. W., Die chemische Verwitterung der Silikate,
Naturwissenschaften, 28, 369-376, 1940.
Cottens, C. W., The experimental chemical weathering of
silicates, Clay Miner. Bull., 26(4), 249-265, 1962a.
Correns, C. W., Ueber die chemische Verwitterung von
Feldspaten, Nor. Geol. Tidsskr., 42,272-282, 1962b.
Correns, C. W., and W. von Engelhardt, Neue Untersuchungen fiber die Verwittemng des Kalifeldspates,
Naturwissenschaften,26, !37-138, 1938a.
Conens, C. W., and W. yon Engelhardt, Neue Untersuchungen fiber die Verwittemng des Kalifeldspates,
Chem. Erde, 12, 1-22, 1938b.
Crozaz, G., and E. Zinner, Quantitative ion microprobe
analysisfor the rare earth elementsin minerals,Scanning
Electron Mœcrosc., no. II, 369-376, 1986.
Cushman, A. S., A study of rock decompositionunder the
32, 2 / REVIEWSOF GEOPHYSICS
dissolutionusing surface replication combined with electron microscopic and spectroscopic techniques,
Geochim. Cosmochim. Acta, 46, 503-512, 1982.
Furrer, G., and W. Stumm, The coordination chemistry of
weathering,I, Dissolutionkinetics of 8-A1203and BeO,
Geochim. Cosmochim. Acta, 50, 1847-1860, 1986.
Gijbels, R., Determination of lanthanides and of uranium
and thorium in geological samples, Inorg. Chim. Acta,
140, 215-230, 1987.
Goldich, S.S., A study in rock-weathering, J. Geol., 46,
17-58, 1938.
Goossens,D. A., J. G. Philippaerts, R. Gijbels, A. P. Pijpers, S. Van Tendeloo, and E. Althaus, A SIMS, XPS,
SEM, TEM, and FTIR study of feldspar surfaces after
reacting with acid solutions, in Water-Rock Interaction,
edited by D. L. Miles, pp. 271-274, A. A. Balkema,
Rotterdam, Netherlands, 1989.
Gossink, G. G., Application of secondary ion mass spectrometry (SIMS) to glasssurfaceproblems, Glass Technol., 2!, !25-!33, 1980.
Gossink, R. G., H. A.M. de Grefte, and H. W. Werner,
SIMS analysisof aqueouscorrosionprofiles in soda-limeaction of water, Circ. 38, 10 pp., Off. of Public Roads,
silica glass, J. Am. Ceram. Soc., 62, 4-9, 1979.
U.S. Dep. of Agric., Washington, D.C., 1905a.
Grant, W. H., Weatheringof Stone Mountain granite, Clays
Cushman,A. S., The effect of water on rock powders,Bull.
Clay Miner. Proc. Conf., !1, 65-73, 1963.
92, 24 pp., Bur. of Chem., Washington,D.C., 1905b.
Harnois, L., The CIW index' A new chemical index of
Cushman,A. S., Rock decompositionand clay formation in
weathering, Sediment. Geol., 55, 319-322, !988.
the laboratory, Trans. Am. Ceram. Soc., 8, 180-197, Havette, A., Mineralogicalapplication of the ion microscope
!907.
elementaryanalysis,ScanningElectron Microsc., no. II,
Cushman, A. S., and P. Hubbard, The decompositionof
585-594, 1985.
feldspars, Bull. 28, 29 pp., Off. of Public Roads, U.S. Hayward, P. J., W. H. Hocking, S. L. Mitchell, and M. A.
Dep. of Agric., Washington, D.C., 1907.
Stanchell,Leachingstudiesof sphene-basedglass-ceramCushman,A. S., and P. Hubbard, The extractionof potash
ics, Nucl. Chem. Waste Manage., 5, 27-37, 1984.
from feldspathic rock, Am. Chem. Soc. J., 30, 779-797, He!geson, H. C., V. M. Murphy, and P. Aagaard, Thermo!908.
dynamic and kinetic constraintson reaction rates among
Daubr6e, A., Etudes Synthetiquesde G•ologie Experimenmineralsand aqueoussolutions,II, Rate constants,effectale, Dunod, Paris, !879.
tive surfacearea and the hydrolysisof feldspar, Geochim.
Dana, S. L., A Muck Manual for Farmers, 4th ed., A. O.
Cosmochim. Acta, 48, 2405-2432, 1984.
Moore, New York, 1858.
Hellmann, R., C. M. Egg!eston,M. F. Hochella, Jr., and D.
A. Crerar, The formation of leached layers on albite
ed., Eastburn, Kirk and Co., New York, and Ward and
surfaces during dissolution under hydrothermal condiLily, Boston, Mass., 1815.
tions, Geochim. Cosmochim. Acta, 54, 1267-1281, 1990.
Della Mea, G., J.-C. Dran, J.-C. Petit, G. Bezzon, and C.
Rossi-Alvarez, Use of ion beam techniquesfor studying Hervig, R. L., Oxygen isotope analysis using extreme energy filtering, Chem. Geol., 101, 185-186, 1992.
the leaching properties of lead-implanted silicates, Nucl.
Hervig, R. L., and P. Williams, SIMS microanalysis of
Instrum. Methods Phys. Res., 218, 493-499, !983.
minerals and glassesfor hydrogen and deuterium, in SecDiamond, L. W., D. D. Marshall, J. A. Jackman, and G. B.
ondary Ion Mass Spectrometry, SIMS 6, edited by A.
Skippen, Elemental analysis of individual fluid inclusions
Benninghoven,A.M. Huber, and H. W. Werner, pp.
in minerals by seconda• ion mass spectrometry,
961-964, John Wiley, New York, 1988.
Geochim. Cosmochim. Acta, 54, 545-555, 1990.
Drever, J. I. (Ed.), The chemistry of weathering, NATO AS! Hervig, R. L., P. Williams, R. M. Thomas, S. N. Schauer,
and I. M. Steele, Microanalysis for oxygen isotopes in
Ser., Set. C, Mathematical 149D, 1985.
insulators by secondary ion mass spectrometry, Int. J.
Drever, J. I., The Geochemistry of Natural Waters, 2nd ed.,
Mass Spectrom. Ion Processes, 120, 45-63, 1992.
Prentice-Hall, Englewood Cliffs, N.J., 1988.
Drever, J. I., and N. Swoboda-Co!berg, Mineral weathering Hochella, M. F., Jr., Auger electron and X-ray photoelectron spectroscopies,in SpectroscopicMethods in Minerrates in acid-sensitive catchments: Extrapolation of labalogy and Geology, Rev. Mineral., 18, 573-637, 1988.
oratory experiments to the field, in Water-Rock Interaction, edited by D. L. Miles, pp. 211-214, A. A. Balkema, Hoche!la, M. F., Jr., and A. F. White, (Eds.), Mineral-water
interface geochemistry,Rev. Mineral., 23, 1990.
Rotterdam, Netherlands, 1989.
Driscoll, C. T., 1'4.van Breemen, and J. Mulder, Aluminum Hochella, M. F., Jr., H. B. Ponader, A.M. Turner, and D.
W. Harris, The complexity of mineral dissolution as
chemistryin a forested Spodosol,Soil Sci. Soc. Am. J.,
49, 437-443, 1985.
viewed by high resolution scanningAuger microscopy:
Fleer, V. N., The dissolution kinetics of anorthite
Labradorite under hydrothermal conditions, Geochim.
Cosmochim. Acta, 52, 385-394, 1988.
(CaA12Si2Os)and synthetic strontium feldspar
(SrA12Si208)in aqueoussolutionsat temperaturesbelow Holdren, G. R., Jr., and R. A. Berner, Mechanism of feldspar weathering, I, Experimental studies, Geochim. Cos100 C: With applicationsto the geologicaldisposalof
radioactive nuclear wastes, Ph.D thesis, Penn. State
mochim. Acta, 43, 116!-1!71, 1979.
Univ., University Park, !982.
Holdren, G. R., Jr., and P.M. Speyer, Reaction rate-surface
area relationshipsduring the early stagesof weathering, !,
Fung,P. C., andG. G. Sanipe!li,Surfacestudiesof feldspar
Davy, H., Elements of Agricultural Chemistry, 1stAmerican
32, 2 / REVIEWSOF GEOPHYSICS
Shotykand Metson: SECONDARYION MASS SPECTROMETRY
ß 217
Initial observations, Geochim. Cosmochim. Acta, 49,
675-681, 1985.
Home, F., The Principlesof Agriculture and Vegetation,3rd
ed., A. Kincaid and J. Bell, Edinburgh, 1762.
Hue, N. V., G. R. Craddock, and F. Adams, Effect of
organic acids on aluminum toxicity in subsoils, Soil Sci.
Soc. Am. J., 50, 28-34, 1986.
Inskeep, W. P., E. A. Nater, P. R. Bloom, D. S. Vander-
surfacelayers in glasses,Mikrochim. Acta Suppl., 11,
145-161, 1985.
Lodding, A. R., P.M. Fischer, H. Odelius, J. G. Noren, L.
Sennerby, C. B. Johansson,J. M. Chabala, and R. Levi-
Setti, Secondaryion mass spectrometryin the study of
biomineralizations and biomateria!s, Anal. Chim. Acta,
241,299-314, 1990.
Lorin, J. C., Oxygen isotope analysis on the Cameca imsvoort, and M. S. Erich, Characterizationof laboratory
300, Chem. Geol., 10!, 193-195, 1992.
weatheredlabradoritc surfacesusingX-ray photoelectron Lorin, J. C., G. Slodzian, R. Dennebouy, and M. Chain-
spectroscopy and transmission electron microscopy,
Geochim. Cosmochim. Acta, 55, 787-800, 1991.
Inskeep, W. P., J. L. Clayton, and D. W. Mogk, Naturally
weathered plagioclasegrains from the Idaho Batholith:
Observationsusing scanningelectron microscopy, Soil
Sci. Soc. Am. J., 57, 851-860, 1993.
Jackson,M. L., and G. D. Sherman, Chemicalweatheringof
minerals in soils, Adv. Agron., 5, 22!-317, 1953.
Jackson, M. L., S. A. Tyler, A. L. Willis, G. A. Bourbeau,
and R. P. Pennington, Weathering sequenceof clay-size
minerals in soils and sediments, I, Fundamental generalizations, J. Phys. Colloid Chem., 52, 1237-1260, 1948.
Jenny, H., Factors of Soil Formation, McGraw-Hill, New
York, 1941.
Johnson, N.M., G. E. Lickens, F. H. Bormann, and R. S.
Pierce, Rate of chemicalweathering of silicate mineralsin
New Hampshire, Geochim. Cosmochim.Acta, 32, 531545, !968.
Johnston, J. F. W., Lectures on the Applications of Chemistry and Geology to Agriculture, Wiley and Putnam,
New York, 1847.
Julien, A. A., On the geological action of the humus acids,
Proc. Am. Assoc. Adv. Sci., 28, 31!-410., 1879.
Katz, B. G., O. P. Bricker, and M. M. Kennedy, Geochemical mass-balancerelationshipsfor selected ions in precipitation and stream water, Catoctin Mountains, Maryland, Am. J. Sci., 285, 931-962, 1985.
Kirkwood, D. E., and H. W. Nesbitt, Formation and evolution of soils from an acidified watershed: Plastic Lake,
Ontario, Canada, Geochim. Cosmochim. Acta, 55, 12951308, 1991.
Kittrick, J. A. (Ed.), Soil Mineral Weathering, Van Nostrand Reinhold, New York, 1986.
Lacharme, J.-P., and P. Lehuede, Interpretation of analyses
of silicateglassesand mineralsobtainedby secondary-ion
massspectrometry,J. Am. Ceram. Soc., 68, C!34-C!35,
1985.
Lau, W. M., N. S. Mcintyre, J. B. Metson, D. Cochrane,
and J. D. Brown, Stabilization of charge on electrically
insulatingsurfacesmExperimentaland theoretical studies
of the specimen isolation method, SIA Surf. Interface
Anal., 7, 275-281, 1985.
Liebig, J., Chemistry in Its Applications to Agriculture and
Physiology,3rd Englished., Taylor and Walton, London,
!843.
Lindsay, N. M., and J.P. F. Sellschop, Routine SIMS
microanalysis:Trace gold and platinumin sulfides,Nucl.
Instrum. MethodsPhys. Res., Sect. B, $5, 358-363, 1988.
treau, SIMS measurement of oxygen isotope-ratiosin
meteoritesand primitive solar system matter, in Secondary Ion Mass Spectrometry, SIMS 7, edited by A. Benninghoven, pp. 377-380, John Wiley, New York, !990.
Loughnan, F. C., Chemical Weathering of the Silicate Minerals, 154 pp., Elsevier, New York, 1969.
MacRae, N. D., Quantitative analysis for rare earth elements by SIMS, Am. Mineral., 72, 1263-1268, 1987.
MacRae, N. D., P. Bottazzi, L. Ottolini, and R. Vannucci,
Quantitative REE analysisof silicatesby SIMS: Conven-
tional energy filtering vs. specimen isolation mode,
Chem. Geol., !03, 45-54, 1993.
Manley, E. P., and L. J. Evans, Dissolution of feldsparsby
low-molecular-weight aliphatic and aromatic acids, Soil
Sci., 141,106-!12,
!986.
Manley, E. P., W. Chesworth, and L. J. Evans, The solution
chemistry of podzolic soils from the eastern Canadian
shield: A thermodynamic interpretation of the mineral
phasescontrolling
solubleA13+andH4SiO4,J. SoilSci.,
38, 39-51, 1987.
Mast, M. A., and J. I. Drever, The effect of oxalate on the
dissolution rates of oligoclase and tremolite, Geochim.
Cosmochim. Acta, 51, 2559-2568, 1987.
McColl, J. G., and A. A. Pohlman, Soluble organic acids and
their chelatinginfluenceon AI and other metal dissolution
from forest soils, Water Air Soil Pollut., 31, 917-927,
1986.
Mcfarlane, A. W., and N. Shimizu, SIMS measurementsof
• sulfur-34 in sulfide minerals from adjacent vein and
stratabound ores, Geochim. Cosmochim. Acta, 55, 52554!, 1991.
McKeague, J. A., M. V. Cheshire, F. Andreux, and J.
Berthelin, Organo-mineral complexes in relation to pedogenesis,in Interaction of Soil Minerals with Natural
Organicsand Microbes, SSSA Spec. Publ., 17, 549-592,
1986.
Merrill, G. P., Rocks, Rock Weathering, and Soils, Macmillan, New York, !921.
Metson, J. B., Secondaryion mass spectrometry,in Instru-
mentalSurfaceAnalysisof GeologicMaterials, editedby
D. L. Perry, pp. 311-352, VCH Publishers,New York,
1990.
Metson, J. B., G. M. Bancroft, N. S. Mcintyre, and W. J.
Chauvin,Suppressionof molecularions in the secondary
ion massspectraof minerals, SIA Surf. Interface Anal.,
5, 18!-185, 1983.
Metson, J. B., G. M. Bancroft, and H. W. Nesbitt, Analysis
of minerals using specimenisolated secondaryion mass
spectrometry,ScanningElectron Microsc., no. II, 595-
603, 1985.
ion massspectrometry(SIMS) in biologicalmicroanaly- Meuris, M., P. De Bisschop,J. F. Leclair, and W. Vandervorst, Determination of the angle of incidence in a Camsis: Technique comparisonsand prospects,Biol. Cell.,
eca IMS-4f instrument, S!A Surf. Interface Anal., 14,
74, 147-160, 1992.
739-743, 1989.
Litaor, M. I., Aluminum chemistry: Fractionation, separation, andmineralequilibriaof soilinterstitialwatersof an Mogk, D. W., Applicationof Auger electronspectroscopyto
studiesof chemical weathering, Rev. Geophys., 28, 337alpinewatershed,Front Range, Colorado,Geochim.Cos356, 1990.
mochim. Acta, 51, 1285-1295, !987.
Linton, R. W., and J. G. Goldsmith, The role of secondary
Lodding, A., H. Odelius, D. E. Clark, and L. O. Werme,
Elementprofilingby secondaryion massspectrometry
of
Mogk, D. W., and W. W. Locke, III, Application of Auger
electron spectroscopy (AES) to naturally weathered
218 ß Shotyk and Metson: SECONDARYION MASS SPECTROMETRY
hornblende, Geochim. Cosmochim. Acta, 52, 2537-2542,
1988.
Monier, E., Action de l'acide oxalique sur !e silicate de
soude, quartz hydrat6, C. R Hebd. Seances Acad. Sci.,
85, 1053-1054, 1877.
Muir, I. J., Secondaryion massspectrometryand its application to studiesin geochemistry, Ph.D. thesis, Univ. of
Western Ont., London, Canada, 1989.
Muir, I. J., and H. W. Nesbitt, Effects of aqueouscationson
the dissolutionof labradoritefeldspar, Geochim.Cosmochim. Acta, 55, 3181-3189, 1991.
Muir, I. J., and H. W. Nesbitt, Controls on differential
leaching of calcium and aluminum from labradorite in
dilute electrolyte solutions, Geochim. Cosmochim.Acta,
56, 3979-3985, 1992.
Muir, I. J., G. M. Bancroft, N. D. MacRae, and J. B.
Metson, Quantitative analysesof rare-earth elementsin
minerals by secondary ion mass spectrometry, Chem.
Geol., 64, 269-278, !987.
Muir, I. J., G. M. Bancroft, and H. W. Nesbitt, Characteristics of altered labradorite surfacesby SIMS and XPS,
Geochirn. Cosmochirn.Acta, 53, 1235-1241, 1989.
Muir, I. J., G. M. Bancroft, W. Shotyk, and H. W. Nesbitt,
A SIMS and XPS study of dissolving plagioclase,
32, 2 / REVIEWSOF GEOPHYSICS
A. Mando, and A. Paccagnella,Hydrated-layer formation
during dissolutionof complex silicate glassesand minerals, Geochim. Cosmochim. Acta, 54, 1941-1955, !990a.
Petit, J.-C., J.-C. Dran, and G. Della Mea, Energetic ion
beam analysis in the Earth sciences, Nature, 344, 621626, 1990b.
Petrovic, R., R. A. Berner, and M. B. Goldhaber, Rate
control in dissolutionof alkali feldspars, I, Study of residual feldspargrains by X-ray photoelectronspectroscopy, Geochim. Cosmochim. Acta, 40, 537-548, 1976.
Pohlman, A. A., and J. G. McColl, Kinetics of metal dissolution from forest soilsby solubleorganic acids, J. Environ. Qual., 15, 86-92, 1986.
Pohlman, A. A., and J. G. McColl, Soluble organicsfrom
forest litter and their role in metal dissolution, Soil Sci.
Soc. Am. J., 52,265-271, 1988.
Polynov, B. B., The Cycle of Weathering,translatedfrom
Russian by A. Muir, Macmillan, New York, 1937.
Reed, S. J. B., Trace element analysiswith the ion probe,
Scanning, 3, 119-127, 1980.
Reed, S. J. B., Geologicalapplicationsof SIMS, in Secondary Ion Mass Spectrometry, SIMS 4, Springer Ser.
Chem. Phys., vol. 36, pp. 451-455, Springer-Verlag,New
York, 1984.
Geochim. Cosrnochim.Acta, 54, 2247-2256, 1990.
Richter, T., G. H. Frischat, G. Borchardt, and S. Scherrer,
Nash, V. E., and C. E. Marshall, The surface reaction of
Shorttime leachingof a soda-limeglassin H20 andD20,
silicate minerals, I, The reactions of feldspar surfaces
Phys. Chem. Glasses, 26, 208-2!2, 1985.
with acidic solutions,Res. Bull. Mo. Agric. Exp. Stn., Rimstidt, J. D., and P.M. Dove, Mineral/solution reaction
613, 1956a.
rates in a mixed flow reactor: Wo!lastonite hydrolysis,
Nash, V. E., and C. E. Marshall, The surface reactions of
Geochirn. Cosmochim. Acta, 50, 2509-2516, 1986.
silicateminerals,II, Reactionsof feldspar surfaceswith Ripoli, C., A. Jauneau,F. Lefebvre, M. Demarty, and M.
salt solutions, Res. Bull. Mo. Agric. Exp. Stn., 614,
Thellier, SIMS determination of the distribution of the
1956b.
main mineral cations in the depth of the cuticle and
Nesbitt, H. W., and I. J. Muir, SIMS depth profiles of
pecto-ce!!ulosicwall of epidermal cells of flax stems:
weatheredplagioclase,and processesaffectingdissolved
Problemsencounteredwith SIMS depth profiling, Biol.
AI and Si in some acidic soil solutions, Nature, 334,
Cell., 74, 135-142, 1992.
336-338, 1988.
Robert, M., and J. Berthelin, Role of biological and bioNesbitt, H. W., and G. M. Young, Early Proterozoic clichemical factors in soil mineral weathering, in Interacmates and plate motions inferred from major element
tions of Soil Minerals with Natural Organics and Michemistry of lutites, Nature, 299, 715-717, 1982.
crobes, SSSA Spec. Publ., 17, 453-495, 1986.
Nesbitt, H. W., I. B. Metson, and G. M. Bancroft, QuantiRuxton, B. P., Measures of the degree of chemical weathtative major and trace element whole rock analysisusing
ering of rocks, J. Geol., 76, 518-527, 1968.
secondaryion mass spectrometry, Chem. Geol., 55,139Schalscha,
E. B., H. Appelt, and A. Schatz, Chelation as a
160, 1986.
weatheringmechanism,I, Effect of complexingagentson
Nesbitt, H. W., N. D. MacRae, and W. Shotyk, Congruent
the solubilizationof iron from minerals and granodiorite,
and incongruentdissolutionof labradorite in dilute, acidic
Geochim. Cosmochim. Acta, 31,587-596, 1967.
salt solutions, J. Geol., 99, 429-442, 1991.
Schenk,
D., A. Petersen, and G. Matthess, Acceleration and
Paces, T., Steady-state kinetics and equilibrium between
retardation of silicate weathering due to organic subground water and granitic rock, Geochim. Cosmochim.
stances, in Water-Rock Interaction, edited by D. L.
Acta, 37, 2641-2663, 1973.
Miles, pp. 605-607, A. A. Balkema, Rotterdam, NetherPaces, T., Rate constants of dissolution derived from the
lands, 1989.
measurementsof mass balance in hydrological catchments, Geochim. Cosmochim.Acta, 47, 1855-1863, 1983. Schindler, P. W., and W. Stumm, The surfacechemistry of
oxides, hydroxides, and oxide minerals, in Aquatic SurParker, A., An index of weatheringfor silicaterocks, Geol.
face Chemistry,edited by W. Stumm, pp. 83-110, John
Mag., 501-504, 1970.
Wiley, New York, 1987.
Perry, D. L., Instrumental Surface Analysis of Geologic
Schott, J., and J.-C. Petit, New evidence for the mechanisms
Materials, VCH Publishers, New York, 1990.
of dissolutionof silicate minerals, in Aquatic Surface
Petit, J.-C., G. Della Mea, J.-C. Dran, J.-C. Schott, and R.
Chemistry, edited by W. Stumm, pp. 292-315, John
A. Bemer, Mechanism of diopsidedissolutionfrom hyWiley, New York, 1987.
drogen depth profiling, Nature, 325, 705-707, 1987.
Petit, J.-C., J.-C. Dran, G. Della Mea, and A. Paccagnella, Schreiner, M., Deterioration of stained medieval glass by
Dissolution mechanismsof silicate minerals yielded by
atmosphericattack, 2, Secondaryion mass spectrometry
intercomparisonwith glassesand radiationdamagestudanalysis of naturally weathered glass surfaces, Glastech.
ies, Chem. Geol., 78, 219-227, 1989a.
Ber., 61,223-230, 1988.
Petit, J.-C., J.-C. Dran, A. Paccagnella,and G. Delia Mea, Schreiner,M., Seconda.ryion massspectrometeranalysisof
Structuraldependence
Of crystallinesilicatehydration potash-lime-silicaglassesleached in hydrochloric and sulduring aqueousdissolution,Earth Planet. Sci. Lett., 93,
furic acids, J. Am. Ceram. Soc., 72, 1713-1715, 1989.
292-298, 1989b.
Schreiner, M., G. Stingeder, and M. Grasserbauer, QuantiPetit,J.-C., G. Della Mea, J.-C. Dran, M.-C. Magonthier,P.
tative characterization of surface layers on corroded me-
32, 2 / REVIEWSOF GEOPHYSICS
Shotykand Metson: SECONDARYION MASS SPECTROMETRY
ß 219
dieval glass windows with SIMS, Fresenius Z. Anal.
Chem., 319, 600-605, 1984.
Schreiner, M., M. Grasserbauer,and P. March, Quantitative
NRA and SIMS depth profiling of hydrogenin naturally
weathered medieval glass, Fresenius Z. Anal. Chem.,
33!, 428-432, 1988.
Schweda, P., Kinetics of alkali feldspar dissolution at low
temperature, in Water-Rock Interaction, edited by D. L.
Miles, pp. 609-612, A. A. Balkema, Rotterdam, Netherlands, 1989.
to Chemical Weathering,246 pp., Lund University Press,
Lund, Sweden, 1990.
Swart, P. K., Calibration of the ion microprobe for the
quantitative determination of strontium, iron, manga-
nese, and magnesium in carbonate minerals, Anal.
Chem., 62,722-728, 1990.
Swoboda-Colberg,N. G., and J. I. Drever, Mineral dissolution rates: A comparisonof laboratory and field studies,
in Water-RockInteraction, edited by Y. K. Kharaka and
A. S. Maest, pp. 115-118, A. A. Balkema, Rotterdam,
Netherlands, 1992.
Schweda, P., Kinetics and mechanismsof alkali feldspar
dissolutionat low temperatures,Medd. StockholmsUniv. Swoboda-Colberg,N. G., and J. I. Drever, Mineral dissoluInst. Geol. Geokem., 281, 1990.
tion rates in plot-scalefield and laboratory experiments,
Chem. Geol., 105, 51-69, 1993.
Shimizu, N., and S. R. Hart, Applications of the ion microprobe to geochemistry and cosmochemistry,Annu. Rev. Tait, J. C., W. H. Hocking, J. S. Betteridge, and G. Bart,
Earth Planet. Sci., 10, 483-526, 1982.
Field burial results and SIMS analysis of the Chalk River
glassblocks, Adv. Ceram., 20, 559-565, 1986.
Shotyk, W., and H. W. Nesbitt, Incongruent and congruent
dissolution of plagioclasefeldspar: Effect of feldspar Tan, K. H., Degradationof soil mineralsby organicacids,in
compositionand ligand complexation, Geoderma, 55, 55Interactions of Soil Minerals With Natural Organicsand
78, !992.
Microbes, SSSA Spec. Publ., 17, 1-27, 1986.
Sigmund,P., Theory of sputter, I, Sputteringyield of amor- Thaer, A.D., The Principles of Practical Agriculture, translated by W. Shaw and C. W. Johnson, C. M. Saxton,
phous and polycrystalline targets, Phys. Rev., 184, 383Barker, and Co., New York, 1860.
416, 1969.
Sigmund, P., Mechanisms and theory of physical sputtering Tole, M.P., A. C. Lasaga, C. Pantano, and W. B. White,
The kinetics of dissolution of nepheline (NaA1Si04),
by particle impact, Nucl. Instrum. Methods Phys. Res.,
Geochim. Cosmochim. Acta, 50, 379-392, 1986.
Sect. B, 17, !, 1987.
Sj6berg, L., Kinetics and non-stoichiometry of labradorite Vandentop, G. J., M. A. Karolewski, and R. G. Cavell,
Secondary ion mass spectra of metal sulfides, int. J.
dissolution, in Water-Rock Interaction, edited by D. L.
Mass Spectrom. ion Processes, 89, 319-342, 1989.
Miles, pp. 639-642, A. A. Balkema, Rotterdam, Nethervan Grinsven, J. J. M., and W. H. van Riemsdijk, Evalualands, 1989.
tion of batch and column techniques to measure weathSlodzian, G., Some problemsencounteredin secondaryion
ering rates in soils, Geoderma, 52, 41-57, 1992.
emissionapplied to elementary analysis, Surf. Sci., 48,
Vannucci, R., R. Tribuzio, G. B. Piccardo, L. Ottolini, and
16!-186, 1975.
P. Bottazzi, SiMS analysis of REE in pyroxenes and
Smets, B. M. J., and T. P. A. Lommen, The leaching of
sodium aluminosilicateglassesstudied by secondaryion
mass spectrometry, Phys. Chem. Glasses, 23, 83-87,
1982.
Stauffer, R. E., and B. D. Wittchen, Effects of silicate
weatheringon water chemistry in forested upland, felsic
terrane of the USA, Geochim. Cosmochirn. Acta, 55,
3253-3271, ! 991.
Storms, H. A., K. F. Brown, and J. D. Stein, Evaluation of
a cesium positive ion source for secondary ion mass
spectrometry,Anal. Chem., 49, 2023-2030, 1977.
Streit, L. A., R. L. Hervig, and P. Williams, Quantitative
SIMS microanalysisof trace elements in geologicalsamples usingin situ ion-implantedstandards,in Microbeam
Analysis, edited by A.D. Romig and W. F. Chambers,
pp. 91-94, San Francisco Press, San Francisco, Calif.,
1986.
Stumm, W., and G. Furrer, The dissolution of oxides and
aluminum silicates; examples of surface-coordinationcontrolledkinetics, in Aquatic Surface Chemistry,edited
amphiboles from the Proterozoic Ikasaulak intrusive
complex(SE Greenland):Implicationsfor LREE enrichment processesduring postorogenicplutonism, Chem.
Geol., 92, 115-133, 1991.
Vedy, J. C., and S. Bruckert, Soil solution:Compositionand
pedogenicsignificance,
in Constituentsand Propertiesof
Soils, by M. Bonneauand B. Souchier, edited and translated by V. C. Farmer, pp. 184-213, Academic, San
Diego, Calif., 1982.
Velbel, M. A., Geochemicalmass balancesand weathering
rates in forested watershedsof the southern Blue Ridge,
Am. J. Sci., 285, 904-930, 1985.
Velbel, M. A., Influenceof surface area, surfacecharacteristics, and solutioncompositionon feldspar weathering
rates, in Geochemical Processes at Mineral Surfaces,
ACS Symp. Ser., 323, 615-634, 1986.
Velbel, M. A., Weathering of hornblende to ferruginous
productsby a dissolution-reprecipitation
mechanism:
Petrographyandstoichiometry,Clays ClayMiner., 37, 515524, 1989.
by W. Stumm,pp. 197-219,JohnWiley, New York, 1987. Velbel, M. A., Influenceof temperatureand mineral surface
Stumm, W., and J. J. Morgan, Aquatic Chemistry,2nd ed.,
characteristicson feldsparweatheringratesin naturaland
John Wiley, New York, 1981.
artificial systems:A first approximation, Water Resour.
Res., 26, 3049-3053, 1990.
Stumm, W., and R. Wollast, Coordination chemistry of
weathering:Kineticsof surface-controlled
dissolution
of Velbel, M. A., Constancyof silicate-mineralweathering-rate
ratiosbetweennatural and experimentalweathering:Imoxide minerals, Rev. Geophys., 28, 53-69, 1990.
plicationsfor hydrologiccontroldifferencesin absolute
Stumm, W., G. Furrer, E. Wieland, and B. Zinder, The
rates, Chem. Geol., 105, 89-99, 1993.
effectof complex-forming
ligandson the dissolutionof
oxides and aluminosilicates,in The Chemistry of Weath- Vickerman,J. C., Secondaryion massspectrometry,Chem.
Br., 23, 969-974, 1987.
ering,editedby J. I. Drever, pp. 55-74, D. Reidel,Norvon Engelhardt, W., Versuche fiber die Verwitterung des
well, Mass., 1985.
Feldspates,Fortschr.Mineral. Kristallogr.Petrogr., 23,
Stumm,W., B. Wehfii, and E. Wieland, Surfacecomplex120-123, 1939.
ation and its impact on geochemicalkinetics, Croat.
Welch, S. A., and W. J. Ullman, Dissolutionof feldsparsin
Chem. Acta, 60, 429-456, 1987.
Sverdrup,H. U., The Kineticsof BaseCationReleasedue
oxalic acid solutions, in Water-Rock Interaction, edited
220 ß Shotykand Metson: SECONDARYION MASSSPECTROMETRY
by Y. K. Kharaka and A. S. Maest, pp. 127-130, A. A.
Balkema, Rotterdam, Netherlands, 1992.
Welch, S. A., and W. J. Ullman, The effect of organic acids
on plagioclase dissolution rates and stoichiometry,
Geochirn. Cosmochim. Acta, 57, 2725-2736, 1993.
Williams, P., The sputteringprocessand sputteredion emission, Surf. Sci., 90, 588-634, 1979.
Williams, P., Secondary ion mass spectrometry, Annu. Rev.
Mater. Sci., 15, 517-548, 1985.
Wilson, M. J., Chemical weathering of some primary rockforming minerals, Soil Sci., 119, 349-355, 1975.
Wittmaack, K., Oxygen concentrationdependenceof secondary ion yield enhancement,Surf. $ci., I12, 168-180,
1981.
Wo!last, R., and L. Chou, Kinetic study of the dissolutionof
albite with a continuous flow-through fluidized bed reactor, in The Chemistry of Weathering, edited by J. I.
Drever, pp. 75-96, D. Reidel, Norwell, Mass., 1985.
Yu, M. L., Chemical enhancementeffects in SIMS analysis,
Nucl. Instrum. Methods Phys. Res., Sect. B, 15, 151-158,
32, 2 / REVIEWSOF GEOPHYSICS
Yurimoto, H., M. Kurosawa, and S. Sueno, Hydrogen anal-
ysis in quartz crystalsand quartz glassesby secondary
ion massspectrometry,Geochirn. Cosrnochirn.Acta, 53,
751-755, 1989a.
Yurimoto, H., A. Yamashita, N. Nishida, and S. Sueno,
Quantitative SIMS analysis of GSJ rock reference sample, Geochem.J., 23,215-236, 1989b.
Yurimoto, H., I. Sakaguchi, N. Nishida, and S. Sueno,
Determination of nickel in GSJ standard rock samples
using secondary ion mass spectrometry, Geostand.
Newsl., 15, 155-159, 1991.
Zoitos, B. K., D. E. Clark, A. R. Lodding, and G. G. Wicks,
Correlation of laboratory and Stripa field leaching studies, in Scientific Basis for Nuclear Waste Management
XII, edited by W. Lutze and R. C. Ewing, pp. 145-151,
Materials Research Society, Pittsburgh, Pa., 1989.
J. B. Metsonf Department of Chemistry, University of
Auckland, PrivateBag, Auckland, New Zealand.
W. Shotyk, Geological Institute, Universityof Berne,
Yu, M. L., Bond breaking and the ionization of sputtered
Balzerstrasse1, CH-3012 Berne, Switzerland.
atoms, Phys. Rev. Lett., 57, 1476-1479, 1986b.
1986a.