Professional Documents
Culture Documents
Mobilization and Redistribution of Rees and Thorium in A Syenitic Lateritic Profile: A Mass Balance Study
Mobilization and Redistribution of Rees and Thorium in A Syenitic Lateritic Profile: A Mass Balance Study
4419-4434
Copyright B 1993 Pergamon Piss Ltd. Pnnted inU.S.A.
Abstract-REE-Th geochemistry and mineralogy have been studied in a lateritic profile derived from a
syenite at Akongo in SW Cameroon. REE and Th mass balance caiculations for the host-rock minerals
show that at least 70% of the LREEs and 50% of the HREEs are contained in allanite, apatite, titanite,
and epidote and at least 50% of the Th is controlled by the same accessory minerals which represent
about 2 wt% of the unaltered syenite. These accessory phases are destroyed during the first stages of
weathering causing most of the REEs and Th to be rapidly released into the soil.
Comparison of the variation in the Zr, Ti, and Th content as a function of the apparent density of the
different zones of the saprolite shows that Th is the least mobile element. The presence of secondary
thorianite (ThOz), the etched surface on zircon 8rains, and the presence of Ti in secondary cerianite
support this geochemical inte~~tation. The ~n~ntration of thorium was, therefore, chosen as invariant
relative to the concentration of the other elements, especially the REEs, in mass balance calculations.
Most of the REEs are leached in the iron-rich upper horizons (loose nodular horizon, iron crust, and
top of mottled clay horizon). Where the ~oundwater table moves (saprolite and bottom of the mottled
clay horizon), the REEs are fractionated and redistributed. There is a juxtaposition of leached and
accumulation zones with precipitation of LREE aluminous hydrated phosphates.
This study supports the existence of two different cycles for the redistribution of elements in the soil:
( 1) as dissolved ions in the saprolite horizon, and (2) as individual particles in the upper part of the
profile.
of secondary REE and Th-bearing minerals within the ac- mains. The red domains (RD) are soft, isolated ferrugi-
cumulation zones, created by either in situ formation after nous mottles in which the parent rock texture can still be
the primary phase breakdown or by direct deposition from recognized. Goethite, hematite, and kaolin&e are the main
soil solutions. Finally, the changes in elemental abundances constituents. White domains (WD) are soft and are
in the soil samples relative to those in the parent rock have mainly composed of kaolinite. The upper limit of this
been calculated. These mass balance calculations allow a horizon is undulating.
comparison of the movement of the REEs and Th on the 3) An iron crust, 230 cm thick, composed of a purple red,
local scale to that of the elements ( 1) which characterize the ferruginous, and indurated material crossed by a network
main lateritic features (Si, Al, Fe, . . .) and (2) which are of ochre-yellow and white, soft and porous spots. In this
generally considered as immobile during weathering (Zr, Ti). layer, the parent-rock structure and texture have com-
pletely disappeared. The upper limit is gradational.
DESCRIPTION OF THE PROFiLE 4) A loose nodular ferruginous horizon, 90 cm thick, in
which indurated hematitic nodules are surrounded by
The studied profile, partly described in BRAUN et al.
yellowish brown, mainly kaolinitic, materials. Near the
( 1990), is located at the top of a syenitic hill. It is five meters
surface, organic matter accumulates down to 5 cm.
deep and displays four different horizons (Fig. i ). From bot-
tom to top, the principal structural and mineralogical features
of the horizons are the following: The White Clayey Network
1) A saprolite, 100 cm thick, which is composed of two do- The seams represent pre-existing joints where clayey ma-
mains with distinct structures. The first domain corre- terials have been both deposited and precipitated. In the sap-
sponds to the sandy weathering products of the syenite rolite and the mottled clay horizons, the seams of massive
in which the preexisting rock texture and structure are structure are essentially composed of halloysite with residual
preserved. These zones are generally lens-shaped. They primary minerals such as microcline, magnetite, and zircon.
present more or less weathered syenitic boulders sur- At the bottom of the saprolite, the seams are extremely po-
rounded by a loose yellow sandy material which becomes rous. Open cracks up to 2 mm are present. In the middle
redder towards the edges of the lens and the top of the part of the saprolite, the seams are continuous, well-individ-
saprolite. The red coloration is due to the increasing in- ualized, and with low porosity. Their margins are marked by
dividualization of goethite and hematite after the break- thin iron oxide coatings. The fissures situated within the seams
down of ferromagnesian minerals. The lenses showing a are coated by cutanes which are related to the circulation of
gradual weathering are embedded in a second domain clays in suspension and their oriented deposition on the l&sure
composed of a network of anastomosing white clayey walls. At the boundary between the saprolite and the mottled
seams. A more detailed description of the seams along clay horizon, the seams are packed and cut by mottled dayey
the profile is given next. The bottom of the saprolite is material. in the mottled clay horizon, some packed relicts of
composed of a syenite flag which is already weathered. the seams delimit three zones called a, b, and c (Fig. I). The
The saprolite shows a gradual, undulating, and discon- seam residues disappear towards the top of the profile and
tinuous upper limit. their composition becomes more kaolinitic.
2) A mottled clay horizon, 75 cm thick, consisting of jux-
taposed red decimeter-size and white centimeter-size do- The Water Table
SAMPLING
I5 kg of the fresh and weathered rocks w&e sampled in been determined by ICPMS &er the d~om~sition of the grains
the bottom flag of the saprolite. with HF at 280°C in Teflon beakers.
2) Saprolite (including sandy domains and white clayey
seams). During the sampling of the sandy domains, dis- Mass Balance CaleaIationsin the Whole Profile
tinctions were made based on the color of the weathering
Density measurements
products which allow two weathering stages to be distin-
guished: a less weathered, yellow saprolite (YSAP) and a The bulk densities were determined with a Go8uel* densimeter
for the freshest and weathered syenite with boulders weighing about
more weathered red saprolite (RSAP). About 15 kg of
400 g, and by the cylinder method for the yellow and red saprolite.
each were sampled. Only three samples of the white clayey
seams ( WCSl, WCSZ, WCS3) (weighing about 200 g
Quantitative elemental analysis
each), situated in the middle part of the saprolite, were
taken for complete analysis. They represent the seams AAer app~p~te crushing, pulve~~tion, and hom~eni~tion
procedures, each soil and rock sample was analyzed for major (Si,
showing the largest positive Ce-anomalies ( BRAUNet al.,
Al, Fe, Mg, Ca, Na, K, Ti, P, Mn) and trace elements (REE, Zr) by
1990). inductively coupled plasma atomic emission spectrometry (ICPAES).
3) Mottled clay horizon (MCH ). A selective sampling of The detection limit was about one to three times chondrite for the
red [RD(a), RD(b), RD(c)] and white domains REE. The errors ( f 1u) are about +5% for Zr and each REE, except
[WD(a), WD(b), WD(c)] was made within the three Lu ( + 15%). Gadolinium concentrations, inducing the positive Gd-
anomalies mentioned by BRAUN et al. (1990), but not discussed,
parts of the mottled clay horizon namely a, b, and c. have been shown to be analytical artifacts. Because of their low con-
4) Iron crust (ICR) and loose nodular horizon (LNH). tents in the rocks, U ana Th were determined by ICPMS at IRSHP-
Representative (about 2 kg) samples of both horizons Nancy (VG@’ISOTOPE PLASMAQUAD, C. Rosin, analyst) and at
were collected. In the loose nodular horizon, the nodules CRPG-Nancy (PERKIN-ELMER SCIEX, G. Mevelle and G. Gov-
indaraju, analysts). The BR (basalt) and DRN (diorite) international
(NOD) and the yellow clayey matrix (MAT) were sep-
rock standards were used. The samples were decomposed by fusion
arated by sieving. with LiBq and dissolved in HNOl 1 N ( GOVINDARAJIJand MEV-
ELLE, 1987). Analytical uncertainties are less than 3%for U and 5%
LABORATORY TECHNIQUES for Th. The detection limit under the analytical conditions used was
0.5 ppm and 0.25 ppm for Th and U, respectively.
Study of the Syenite Minerals and Their Weathering Products
Weathering petrography RESULTS
The identification of the primary minerals and their weathering
products was done on ( I ) thin sections of rock and soil, (2) mineral The Parent Rock: The Akongo Syenite
separates, derived m~neticaily and by the use of heavy liquids, and
(3) undisturbed soil volumes with optical microscopy and scanning The syenite sample used as a geochemical reference in this
electron microscopy (STEREOSCAN 250 CAMBRIDGE coupled study was collected from the bottom flag. The hand specimens
with a PGT SYSTEM III Energy Dispersive Spectrometer). Back- of the Akongo pit syenite have a granular structure with a
scattered electron imaging was also used. slight mineral orientation defined by pyroxene crystals. The
texture is porphyroblastic with pyroxene and perthite blasts.
Quantitative eIementa~analysis
The major minerals are perthite (microcline and albite), ae-
Separated grains mounted in resin and polished were analyzed girine-augite, actinolite, biotite, and magnetite. The accessory
with a CAMECA SX 50 for the accessory minerals (epidote, apatite, minerals include &on, titanite, alla&e, epidote, and apatite
sphene, atlanite) and selected secondary REE-Th-bearing minerals.
whose occurrence and aIteration features are described by
The experimental conditions are indicated in BRAUNet al. ( 1990).
The standards used were synthetic monominerals of inter-metallic BRAUN et al. ( 1990). Based on color and magnetic suscep-
com~unds (Ge, Ru) for REEs (La, Ce, Nd, Sm, Gd) and oxide for tibility, two populations of titanite crystals can be distin-
Th. The detection limit was approximately 500 *g/g in the best con- guished in grain size fractions. The most abundant is orange
ditions. For each mineral, point analyses and profiles were done. The yellow while the second, much more rare, is light yellow.
numerous analyses were statistically checked by principal component
analysis (PCA) in order ( 1) to remove the spurious analyses, and Fu~e~ore, EDIMO( 1985), who studied the parageneses
(2) to determine inter-elemental correlations especially between the of the Akongo-Lolodorf syenitic axis, reported the presence
elements constituting the &essory minerals and the secondary REE- of rare thorite ( ThSi04) and monazite (LREEPO,+) crystals
Th b-earing phases. in some samples of the Akongo syenite collected near the
studied pit. The occurrence of accessory quartz crystals is
REE and Th anal.vsisin mineral separates also mentioned by BILONG( 1988) in the pit syenite.
REE and Th contents of separates of alkaline feldspars (albite and
microcline), aegirine-augite, magnetite, apatite, epidote, allanite, ti- Modal Analysis
tanite, and zircon were analyzed by isotope dilution mass spectrometry
(IDTIMS) (CAMECA 206TSN) following MICHARDet al. ( 1986)
Modal analysis of the syenite was based on point counts
and/or inductively coupled plasma mass spectrometry (ICPMS) fol-
lowing JARVIS( 1988). Particular attention was paid to the purity of of four thin sections totalizing about 40 cm2. A large number
mineral separates. The mineral fractions were weighed with an ab- of points {2~) were counted with an interactive video-
solute precision of +-1 pg, and dissolved in teflon beakers with HNOr graphic analyzer ( LAPIQUEet al., 1988) COUQ~~~ with an op-
HF-HCI. The precision for each REE is in the range of 0.1 -I% for tical microscope to reduced uncertainties in the abundance
IDTIMS and about &5% for ICPMS. REE contents of zircon grains
have been determined by secondary ion mass spectrometry (SIMS)
of the accessory minerals. Selected mineral density data
(CAMECA IMS3F) at CRPG (Nancy). The precision is about (DEER et al., 1985) were used to convert volume fractions
2 10% for each REE. The Th contents of two zircon separates have obtained from point counting into weight fractions taking
4422 J. J. Braun et al.
TABLE 1.Whole syenite ICPAES concentrations of major elements (in oxide wtl)
and ten REE, U, Th and Zr (in ppm) and modal analyses of the syenite minerals from
point counting (in volume and weight percent). Errors are calculated both from counting
statistics and from density uncertainties.
whole rock analysis (FSY) modal abundance
% ppm volume % weight %
* Apatite and zircon weight percent assume that all of the whole rock P is included
in apatite and Zr in zircon.
relative density errors of 5%. The volume percent and the give coherent values. This was due to the fact that the whole
calculated weight percent modal abundances of the syenite syenite analysis reflects a slightly weathered rock in which
minerals are reported in Table 1. The whole syenite analysis Na and Ca are already depleted. Moreover, apatite and zircon
is also given. Although accessory mineral abundances ob- abundances have also been calculated assuming all of the
tained from point counting are not entirely reliable ( GROMET whole rock P is included in apatite and Zr is in zircon. These
et al., 1983), this method was applied here because the cal- calculations may overestimate the amount of apatite and zir-
culation of mineral weight fractions from the whole rock con in the rock. The results are reported in Table 1. For both
quantitative analysis and the mineral compositions did not minerals, the data are similar indicating that the apatite and
ws 42.05 f0.67 SiOz 29.16 f0.25 30.44 + 0.33 sioz 37.36 rt 0.67 33.43 f 2.17 30.65 * 0.37
CaO 57.16 f0.40 AlsO, 1.39 f0.13 1.12 zto.10 AW’a 20.73 f 0.37 16.65 f 2.34 12.66 f 0.46
Na,O 0.32 f0.04 FeO, 1.67 f0.14 1.22 1tO.16 FeOt i3.88 i 0.48 14.55 * 0.71 15.46 f 0.46
Ce#s 0.34 f0.11 CaO 26.06 f0.25 28.17 f 0.31 CaO 20.75 f 0.44 16.11 i 2.71 11.59 f 0.51
La205 0.10 f0.06 TIOz 35.29 f 0.34 37.23 f0.44 M&IO cdl 0.34 *0.21 0.96 * 0.21
F 2.14 f0.04 Las03 0.25 f0.12 < dl MnO 0.28 fo.lo 0.20 f0.12 0.41 f 0.18
Cl 0.05 f0.03 CeaOs 1.90 f0.14 0.45 f0.12 Ti02 c dl 0.24 to.14 0.66 +0.17
102.17 Nd203 1.02 f0.21 < dl La203 < dl 2.75 k 1.22 6.10 + 1.15
H,O’ [l.ll] [l.lS] C-w3 < dl 7.18 f3.40 12.19 + 0.30
-0.F 0.90 NdzO, < dl 2.73 -t 1.04 2.40 to.52
-O&I 0.01 HzO’ Il.751 [1.60] rl.481
0.91
zircon percentages are about 1 wt% and 0.02-0.04 wt%, re- for apatite and titanite are twenty and one hundred times
spectively. those of the syenite, respectively.
TABLE 3. REE and Th concentrations (in ppm) of mineral separates of the Akongo syenite as determined by different analytical
techniques. Average microprobe analyses are in oxide wtl. The syenite REE concentrations (FSY) are indicated.
LO 1659 1471 0.09 f0.07 3144 2590 0.22 +0.11 <dl 528 26265 2.3 f 1.0 5.2 il.0
Ce 4060 4158 0.30 i 0.10 11650 13337 1.62 f0.12 0.39 fO.ll 1080 37743 6.1 f 2.9 10.4 f 0.2
Nd 2525 2079 nd 7850 5736 0.87 f0.19 cdl 495 10506 2.3 f0.9 2.1 f 0.4
Sm 395 367 nd 1395 1109 nd nd 76 584 < dl <dl
EU 66 64 nd 330 225 nd nd 26 107 nd nd
Gd 230 295 nd 800 772 <dl <dl 47 584 <dl <dl
J'Y 86 89 nd 362 275 nd nd 19 58 nd nd
Er 28 nd nd 117 nd nd nd 6.3 39 nd nd
Yb 19 nd nd 89 nd nd nd 5.3 <lO nd nd
ZREE 9088 8523 0.39 25737 24044 2.71 0.39 2263 75886 10.7 17.7
(La/Yb),,, 58.3 23.8 - 67.5 >moo -
EulEu* 0.9 0.7 - 1.0 0.8 1.4 0.6
ce/ce* 1.0 1.2 - 1.3 1.9 1.0 0.9 1.3 1.3
Th 35 226 24 661
Zircon crystals (Fig. 3a) which came directly in contact phosphates of LREEs, Ca, Sr, and Ba. However, LREEs are
with corrosive soil solutions have coJ-rosion features as shown not strongly correlated (r = 0.5) with P and Al, suggesting
by grains found in the white clayey seams of the saprolite that they could also be incorporated in other phases. Thorium
(Fig. 3b). However, some crystals preserve their original seems to be correlated with Fe, Ti, and Si, but this correlation
morphology throughout the whole profile when they are is not well defined because Th concentrations are often below
embedded in clayey ferruginous concretions. the detection limit of the electron microprobe. Cations that
Florencite [ LREEAls( P04)20H6] has been clearly iden- are able to substitute for LREEs in the florencite lattice, such
tified by X-ray diffraction in an aggregate separate obtained as Sr, Ca, and most importantly Ba, are always abundant
by heavy liquid and magnetic separation from the weathered (Fig. 4). The florencite fraction of the donuts ranges from
syenite. Zircon and titanite peaks have also been identified 50 to 70%, calculated assuming the idealized formula (LREE,
in the mixture. The clay material observed by SEM and an- Ba, Sr, Ca)AlJ( PO,h. OH6 and assuming that all P is in the
alyzed with EDS (Fig. 3d) does not show X-ray signature. aluminous phosphate.
The florencite donuts (from 1 to 5 pm in diameter) grew on The fractionation of Ce in aluminous phosphates is com-
the residual walls and in the dissolution cavities of weathered plex. Florencite- ( La) with negative Ce-anomaly ( Ce /Ce *
apatite (Fig. 3c) and allanite crystals. Florencite was also = 0.4), florencite-( Ce) with slight negative Ce-anomaly (Ce/
found admixed with anatase (Ti02) in the weathering prod- Ce* x 0.7) and florencite-(Ce) without Ce-anomaly can
ucts of titanite. Quantitative point analyses and step profiles coexist in the weathered syenite. The most abundant is the
have been done within the most homogeneous zones and the florencite-( Ce) with Ce/Ce* ofabout 0.7. The ICPMS chon-
REEs richest donuts of thirteen selected aggregates similar d&e-normalized REE pattern of an aggregate separate is
to that presented in Fig. 3d. The analyzed donuts were pre- compared to those of the REE-bearing accessory minerals in
viously identified with the SEM using backscattered electron Fig. 5. Compared to allanite, the aggregate has a flatter REE
mode. The average analyses and the standard deviation of pattern and a slightly negative Ce-anomaly. The weathering
three representative donut aggregates are reported in Table products are strongly enriched in LREEs compared to apatite
4. An analysis from the Si-Al-Fe matrix admixed with the and titanite. The ICPMS Th concentration of the aggregate
donuts and from a 2 pm rhabdophane crystal found in a separate is 2000 ppm. It is enriched in Th compared to the
dissolution cavity of an apatite grain are also given. Samarium Th-richest accessory minerals from which it developed. The
and sulfur have been analyzed but they are generally present analyzed rhabdophane crystal located in an apatite dissolution
in quantities too small to be detected with the electron mi- cavity presents both an enrichment in LREEs [(La/Nd),,,
croprobe, except for Sm in rhabdophane. The relative errors, = 3.31 and a slight Ce-anomaly (Ce/Ce* = 0.7). The Si,
calculated for each analyzed element, are indicated in Table Al, and Fe contents are probably due to nearby clays rather
5. All the analyses of donuts have low totals (73-86 wt%), than the rhabdophane itself.
indicating porous and/or hydrated materials. Significant In the other parts of the saprolite the donuts appear either
variations in element concentrations of donuts are found both isolated in the sandy matrix or as chains at the boundary
between different aggregates and within single aggregates. The between porous (red sandy saprolite) and compact domains
one hundred ninety one electron probe analyses conducted (white clayey seams). Cerianite ( CeOz), showing a possible
on the donuts and the associated matrices suggest that the substitution of Ce by Ti (Fig. 6), has been identified in sig-
donuts are complex assemblages of weathering products nificant amounts within the white clayey seams of the middle
composed of a mixture of Fe, Ti, Si oxides and aluminous part of the saprolite ( BRAUN et al., 1990).
Geochemistry of REEs and Th in a lateritic soil profile 4425
Table 3. (Continued)
rim core
1 att8lYsir 5 atta1vses
Mass Balance Calculations Using Thorium as the Invariant Assuming that their distribution is homogeneous in the
in the Whole Profile parent syenite, Ti, Zr, and Th can be used as immobile ele-
ments for mass balance calculations in the supergene envi-
The concentrations of major (Si, Al, Fe, Mn, Mg, Ca, Na,
ronment. The behavior ofthese elements has been compared
K, Mn, Ti, P) and selected trace elements (REE, U, Th, Zr)
to select the best among them with isovolume mass balance
from the different zones of the profile used for the calculations
are reported in Table 6. calculations ( MILLOT and BONIFAS, 1955) in the weathered
syenite and the yellow and red saprolite where the original
structure and texture are preserved. This is mainly supported
by petrographical observations showing that the well-pre-
IO’
served K-feldspar crystals constitute the framework of the
saprolite zones ( BILONG, 1988). The isovolume mass balance
equation can be written as follows:
(1)
lo5
e.!Jr
- 1
c es* PS
% change = A- I -100,
[C Ppr
10’
(G,s/Ci,s)
Er Vb calculated according to NESBITT ( 1979
FIG. 2. Chondrite normalized REE patterns for eight mineral phases
in the Akongo syenite. The “aegyrine-a&e” pattern represents a
global analysis including the inclusions. Zircon rim and core (average
=[fCe,pr,Ci,pr)
%change - 11-1W (2)
of five analyses) patterns are shown. The whole rock REE pattern is
also shown. See Table 3 for data. Data measured by ICPMS, IDTIMS, where C’,, and Ce.prare the same parameters as in Eq. 1 and
and SIMS. Ci., and Ci,p, are the concentrations of the immobile element
4426 J. J. Braun et al.
FIG. 3. (a) Unweathered zircon grain from the freshest syenite (SEM-b&scattered electron mode). (b) Partially
dissolved zircon grain from a white cIayey seams of the saprolite (SEM-backscattered electron mode), (c) Detail view
of residual walls of an apatite crystal on which tlorencite donuts (Fl) have grown. (d) Polish section of an aggregate
cf florencite donuts (ff ). The donuts are mixed with an amorphous AI-Si phase (AS) (SEM-scatty electron
mode).
( I ) in the soil sample and the parent rock, respectively. Dia- 1) Zr shows only small losses in the saprolite and iron-rich
grams in which the percent change of the REE are plotted horizons and Ti shows significant redistribution within
vs. atomic number are presented in Fig. 8. The mass balance the whole profile, especially in the white seams where it
has not been calculated for the white clayey seams due to is most strongly concentrated.
their origin. It is important to note that the Th content of 2) Except for Fe, the major elements are depleted in the
these zones is less than that of the parent material. All ele- upper horizons (ICR, LNH, and the top of the MCH).
ments which are present in these zones may be considered In the zones which are periodically exposed to the water
to be accumulated. From Tables 6 and 7, the following may table ( WSY, YSAP, RSAP, red and white domains of the
be concluded. bottom of the MCH), alkalines, alkaline earths, and Si
TABLE4. Avenge microprobe analyses and standard deviation (+ I (I) for the donuts
of 13 selected aggregates. The number of analyses for each aggregate is indicated. An
analysis of the Si-Al-Fe-Ti mixing phase and an analysis of a rhabdophane crystal are
also indicated. Analytical erron are indicated. (i..a/Nd)ti and Ce/Ce* have been calculated.
Florenclte donut aggrogetes matrix rhsbdophane
Number of 1401 [ill r151 [II 111
p205 16.6 f 2.0 18.1 f 2.1 13.6 k 1.0 0.3 ?:0.1 29.0 TO.6
A1203 28.5 f 4.9 29.3 f 2.4 21.0 f 1.1 18.7 t 0.4 r.2 +0.1
FeOt 10.1 f 8.1 7.5 f 4.1 20.4 t 3.5 21.1 + 1.0 0.8 i0.2
TlO2 1.9 _+1.6 1.2 + t.2 2.2 f 0.5 0.5 t 0.3 <dl
ThO2 0.3 f 0.2 0.1 + 0.1 0.3 + 0.2 <dl <dl
sioz 2.2 f2.0 3.3 k3.7 7.3 + 3.4 45.3 i:0.5 0.2 4 0.1
cao 0.9 f 0.2 1.1 c 0.1 0.7 * 0.1 cdl 3.0 f0.2
SaO 3.1 kO.9 3.9 f 1.4 2.3 f 1.1 cdl 0.4 k 0.1
SrO 1.5 io.5 2.0 f 0.3 0.8 f 0.3 cdl <dl
La203 4.9 f 1.6 3.8 f 1.1 5.5 * 1.0 cdl 19.0 f 0.8
Ce203 5.3 k1.2 6.5 + 1.2 3.9 f 1.0 cdl 22.2 il.3
Nd203 2.5 IO.4 2.6 f 0.6 3.4 i 1.4 edi 11.1 +f.$
Sm203 cdl cdl sdl <dl 1,o +0.5
__ _
Ca 10 - -
Sr 17 -
Ba - 10 -
1, . 10 3
._
2 - lb : La Ce Nd SmEuGd Dy Er Yb
_ _
Tl :; _ _
FIG. 5. Chondrite normalized REE patterns for the main primary
bearers (allanite, 0, apatite, 0, titanite, A, and the donut aggregate
separate (0) of which the major part is composed of florencite-(Ce).
are removed. Iron is redistributed over short distances,
leading to the presence of Fe-poor and Fe-rich domains
side by side within the mottled clay horizon. Iron is leached cumulated whereas the HREEs are partially leached away.
in the saprolite. Aluminium is relatively stable and is ac- Ce tends to be enriched compared to the other REEs in
cumulated at the bottom of the profile. In the white seams, the yellow and red saprolite.
Si and Al are concentrated.
DISCUSSION
3) Throughout the profile, U is strongly depleted, behaving
much like the most mobile alkalines and alkaline earths Distribution of Thorium and REEs in the Syenite:
(Ca, Mg; Na). The REE have variable behaviors within The influence of the Accessory Minerals
the different zones of the profile. Most REEs are leached
in the Fe-rich upper horizons (loose nodular horizon, iron The fractional contribution of each mineral to the whole
crust, and the top of the mottled clay horizon, Fig. 8a). rock REE and Th contents of the syenite is shown in Fig. 9.
Where the water table moves, the REEs are concentrated Because of the uncertainties in the REE concentrations of
in the clayey zones in which the soil solutions have a the aegirine-augite, they have not been taken into account
longer residence time during water table discharge. These in these calculations. The SIMS average REE contents of
zones are the white domains of the mottled clay horizon core zircon were used for these calculations. The ICPMS REE
(Fig. 8b) and the white clayey seams of the middle part and Th contents of the allanite separate are considered to be
of the saprolite in which Ce is strongly concentrated. The representative despite the heterogeneous distribution of ele-
REE behavior in the isovolume weathering zones is char- ments in this mineral.
acterized by the juxtaposition of leached and accumulation It is apparent that accessory minerals control the REE and
zones. The leached zones are the yellow and red saprolite Th concentrations in the Akongo syenite. These minerals
(Fig. 8c) and also the red domains of the mottled clay are, in decreasing order of importance: apatite, titanite, al-
horizon bottom (Fig. 8b). The accumulation zone is the lanite, and epidote. This accessory phase control on the lan-
weathered syenite (Fig. 8c) in which the LREEs are ac- thanide and actinide distribution is common within most of
the granitoids ( FOURCADEand ALLEGRE, 1981; PAGEL, 1982;
GROMET and SILVER, 1983). In the Akongo syenite, apatite
Ba
and titanite are of dominant importance in the middle and
heavy REEs, whereas allanite is important for the lightest
0
s o”0 O
z ,_ 0
E
l-
0 O
I
. ’ o . 8 .
10 20 30 40
Ce02 (wt%)
La+Ce+Nd sr+ca
FIG. 6. Correlation between TiOz and Ce02 in the cerianite ac-
RG. 4. Electron microprobe analyses for Sr, Ra, Ca, Ce, La and cumulation zone of the white clayey seams of the middle part of the
Nd from aggregated donuts of the weathered syenite. saprolite (microprobe analyses).
TABLE 6. ICPAES concentrations of major elements (in oxide wt8) ten REE and Zr fin ppm) and ICPMS concentrations of U and Th (in ppm). Three replicate
analyses (2 1u). The characteristic ratios are calculated as follows for the REEs:
LOOSE
SANDY SAPROLITE WHITE CLAYEY MOTTLED CLAY HORIZON IRON NODULAR
SEAMS CRUST HORIZON
FSY WSY YSAP RSAP WCS1 WCS2 WCS3 RDa WDa RDL WDb RDc WDc ICR MAT NOD
SiO2 59.37 58.95 54.92 50.34 46.52 45.95 45.21 44.67 44.18 45.82 44.32 43.92 46.79 36.35 45.82 17.23
16.57 17.95 21.10 22.73 32.oo 33.28 33.59 22.23 33.65 22.33 34.96 24.32 32.29 24.35 22.91 19.03
A’2O3
5.09 5.34 5.92 a.89 1.62 1.92 1.45 14.75 3.47 13.35 2.24 13.89 2.95 22.17 13.85 46.45
Fe203
MnO 0.06 0.02 < dl <dl < dl < dl IC:
dl < dl < dl 0.01 < dl 0.01 cdl < dl 0.02 0.03
MgO 1.14 0.48 0.12 0.10 0.07 0.10 o.o4 0.07 o.io 0.04 0.05 0.04 0.08 0.11 0.14 0.07
CaO 2.54 0.60 0.03 0.05 0.04 o.o6 0.06 0.03 0.05 0.01 0.01 < dl < dl 0.02 0.02 n.d.
Na20 3.47 2.50 0.44 0.29 0.10 0.06 0.07 0.22 0.03 0.24 0.01 0.17 0.05 0.04 0.08 n.d.
9.19 10.92 11.41 8.99 3.07 2.37 2.W 7.21 1.11 7.53 0.76 5.82 2.00 1.68 2.49 0.11
K2*
TiO2 0.24 0.26 0.29 0.45 0.20 0.34 0.17 0.58 0.53 0.64 0.34 0.56 0.71 0.76 1.06 0.48
P2Os 0.40 0.26 0.26 0.26 0.22 0.2O 0.28 0.30 0.32 0.30 0.32 0.30 0.20 0.28 0.25 0.40
L.O.I. 0.92 1.61 4.57 6.96 16.02 15.39 17.15 9.04 16.23 a.78 16.70 10.06 14.54 13.93 12.95 14.62
Totsi 98.99 96.99 99.06 99.06 99.92 99.67 100.02 99.10 99.67 99.05 99.71 99.11 99.61 99.69 99.59 98.42
La 69.48 115.15 e4.74 100.24 47.16 49.83 69.36 521.56 154.45 55.85 177.27 64.90 70.70 83.14 67.86 52.26
ce 139.97 234.81 190.14 239.50 715.64 1395.34 2355.79 157.00 318.84 123.20 345.43 127.21 132.97 145.89 123.19 95.46
Nd 65.&Q 96.91 67.93 al.32 33.15 41.14 70.94 48.&l 156.54 50.71 162.36 64.71 51.54 51.80 44.07 31.25
Sm 10.71 14.01 10.24 12.12 4.51 6.07 11.01 7.95 25.30 a.21 25.99 9.07 8.41 8.50 6.95 5.51
EU 2.93 3.39 2.34 2.67 1.02 1.54 3.013 1.91 8.53 2.01 6.38 2.24 2.02 1.99 1.63 1.26
Gd 6.54 7.34 a.55 5.90 (5.37) (9.aa) (19.88) 4.44 16.58 4.62 16.08 5.29 5.27 5.41 4.43 3.87
DY 2.66 2.48 1.93 2.02 0.82 1.42 2.72 1.94 8.07 1.77 7.56 2.09 2.43 2.45 2.48 I.92
Er 1.13 0.94 0.74 0.77 0.39 0.56 0.97 0.81 2.90 0.75 2.70 0.87 1.06 1.22 1.19 0.89
Yb 0.90 0.63 0.42 0.51 0.27 0.43 0.61 0.62 1.69 0.52 1.49 0.57 0.76 0.88 1.11 0.70
LU 0.13 0.11 0.07 0.11 0.07 0.13 0.27 0.13 0.22 0.10 0.17 0.11 0.15 0.18 0.17 0.17
XREE 300.29 475.77 373.10 445.16 808.40 1506.34 3034.61 276.20 691.22 247.74 745.43 267.05 275.31 301.46 253.08 193.29
U 5.5O?0.09 2.35t 0.06' 1.08f0.05' 0.92+I0.02 0.3 0.5 0.3 0.7 1.0 0.4 1.1 0.9 1.0 21 2.7 3.1
Th 3.11f 0.19" 3.65c 0.03' 5.05i 0.11' 5.76t 0.32 21 29 2.6 6.7 72 4.7 5.5 5.8 5.2 13.3 15.2 20.9
UiTh 1.8 0.6 0.2 0.2 0.1 0.2 0.1 0.1 0.1 0.1 02 0.1 02 0.2 02 0.1
Zr 169 211 178 259 79 110 125 344204 357 118 309 270 305 685 315
Geochemistry of REEs and Th in a lateritic soil profile 4429
100 and 50% of the Th. The REE and Th contribution of zircon
is extremely low ( <O. 1%) even for the heaviest REE which
Ill 50 is due to both the low volume fraction and low REE and Th
(3 contents. Concerning the major phases, perthitic feldspars
z 0 provide about 5% of the LREEs, except for Eu (35%), and
3 3% of the Th. As for magnetite, its, contribution is less than
g-50 1%. The REE and Th contents of the reconstructed rock are
lower than those of the whole rock. This could be due to: ( 1)
-1001 ’ I the lack of reliable data for aegirine-augite, actinolite, and
FSY WSY YSAP RSAP biotite, (2) nonrepresentative mineraL separates, (3) the
““;,~;;“’ (2,75iO.O5)(2.62~0.05)(1.4~0.1) (1.6kO.l) complex zoning in allanite, and (4) either HREE or LOPE
and Th-bearing minerals (e.g., xenotime, monazite, thorite)
INCREASING DEGREE *
OF WEATHERING which were not analyzed because of their low abundance
( ~0.0 1%) . Considering the distribution coeficients of REE
FIG. 7. Percentage change of Ti, Zr, and Th with relation to the
for the clinopyroxene (HENDERSON, 1984), the aegirine-au-
freshest syenite. The calculations are based on equal volumes of the
saprolite zones. Bulk densities of syenite, weathered syenite, and yel- gite ( 14 wt% of the whole rock) may account for about IO-
low and red saproiite are indicated. The shaded zone (&2O%) rep t 5% of the lightest REE, 30-40% of the intermediate REEs,
resents the range in which an element may be considered as immobile and 30% of the heaviest REE. On the other hand, considering
taking into account the analytical errors. its ICPMS Th content, the aegirine-augite provides a maxi-
mum of 14% of this element. The clinopyroxene may account
REEs and Th. Similar results for a granodiorite have previ- for the lack of REEs but not for Th for which about 40%
ousIy been reported ( GROMET and SILVER, 1983 ). These remains unexplained. According to MARIANO ( 1989), only
accessory phases account for about 70% of the lightest REEs, 4 - 10-3 wt% of xenotime would he sufficient to fill the de-
60% of the intermediate REEs, 40% of the heaviest REEs, pletion of both heaviest REE and Th. On the other hand,
TABLE7. Comparative mobility of the major elements, REEs, U, and Zr in the profile zones. Calculations have been made using Th as an
“invariant” element,
LOSSES
low low I
average high
-40% f
0% I > 100%
0%
+lOO%
LOOSE Fe
NODULAR
HORIZON Ti,Zr
Ti
IRON CRUST F8
REE,U
WSY I Ca
I f&l
U
Si,P,Fe,Na
AI,Ti,Zr.HREE
K
LREE
4430 J. J. Braun et al.
+ reconetn~eted rock
X alkallna
feldspars
* magnstfte
9 apatfts
f2 tltanlte
A epkfote
0 alianfte
,,*
La Ce Pr NdPmSmEuQdTb DyHo ErTmYbLuTh
FIG. 9. Fmc nal contribution of each mineral to the whole rock REE and Th contents of the syenite. A composite
reconstructed ck, calculated from the sum of the mineral contributions, is also plotted.
compared to those of LREEs. SMEDLEY ( 199 1), who studied The mineralogical investigation of the saprolite suggests
shallow groundwaters from the granitic Cammenellis area that phosphates play a major role in the REE trapping. The
( SW England), found higher REE contents in groundwaters aggregates of florencite donuts and rhabdophane crystals were
of pH < 6. Assuming that the pH of the water table is about selected from the basal weathered syenite. C&mite was found
6, the soluble REE should be present mainly as free ions or in the clayey Seams of the middle saprolite. Both domains
sulfate complexes after the models from TURNER et al. ( 198 1) are accumulation zones. In the weathered syenite (WSY),
and WOOD ( 1990). However, a part of the HREE may exist all the LREEs were concentrated, whereas the white seams
as carbonate complexes. show a large increase in Ce contents. Therefore, REEs must
The rise and fall of the water table induce ( 1) contrasting have been removed from the other parts of the profile, trans-
dynamics of the soil solutions in the sandy saprolite and in ported in solution and deposited as REE-bearing secondary
the white clayey seams and (2) the return of oxidizing con- minerals.
ditions. A part of the REEs are leached away from this sap- In the weathered syenite, there is simultaneously an in situ
rolite located at the hill top. precipitation of REEs from the primary minerals and a con-
4432 J. J. Braun et al.
tribution to the growth of secondary phases of REEs and Loose nodular horizon
some cations with similar geochemical behavior (Ba, Sr, Ca)
which were brought in the soil solutions. This is demonstmted The REEs which are retained in this horizon may be con-
by the presence of phosphates with or without Ce-anomalies. tained either in resistant primary or secondary minerals or
Barium and Strontium would be provided by the dissolution adsorbed onto clays (PRICE et al., 199 1). In this acidic media
of alkaline feldspars. An ICPAES analysis of a perthite sep- ( pH c 6)) the probable REE-bearing secondary minerals are
arate gave 5000 and 1500 ppm of Ba and Sr, respectively. generally phosphates ( STUMM and MORGAN, 198 1; MARI-
When the water table falls, the soil solutions are more and ANO, 1989). REE carbonates should only precipitate at higher
more confined and may be saturated in REE-Ba-Sr-Ca phos- pH values, rarely encountered in the top horizons of lateritic
phates which ( 1) either precipitate in the dissolution cavities proliles ( HERBILL~N and NAHON, 1988). The relative ac-
of the weathered accessory minerals, or (2) grow at the ex- cumulation of resistant zircon grains can provide a slight
pense of apatite crystals. The in situ replacement of apatite enrichment in the heaviest REEs, especially in the sandy
by LREE phosphates has been proposed by JONASSON et al. fraction.
When ligands such as PO:-, CO’,- or organic acids are
( 1985) and ELDERFIELDet al. ( 1981). SAWKA et al. ( 1986)
absent from the solution, the affinity of surface minerals in-
and MARIANO ( 1989) show that the secondary REE phos-
creases from La to Lu ( ROALDSET, 1974: AAGAARQ 1974).
phates (florencite and rha~op~ne) are often depleted in
ELPRINCE et al. ( 1980) demonstrated that the selectivity of
Ce. BANF~ELDand EGGLETON ( 1989) studied in detail the
La is stronger than Na and Ba on montmorillonite. SHOL-
weathering of apatite crystals from the Southern North Wales
KOVITZ( 1989) has shown the strong retention of 15’Eu, used
Granite (Australia). In the weathered rock, the REE-bearing
as a tracer in his experiments, on clayey sediments from Buz-
assemblage is similar to that observed in this study. It consists
zard Bay, USA. When inorganic and organic ligands are
of donuts with complex chemical compositions including
present, the LREEs are preferentially adsorbed on solid sur-
florencite-( La), florencite-( Ce), and rha~ophane in which
faces by comparison to HREE (BYRNE and KIM, 1990). In
Ba, Sr, and Ca are present. These phases are mixed with Fe-
Ti oxides. Aluminium is often in excess in the phosphate the loose nodular horizon, REEs may be leached by acidic
donuts. Rhabdophane could precipitate in Al-free micro- rainwater and are perhaps more commonly transported in
domains, On the other hand, DUDDY ( 1980), who studied suspended form than in soluble form. If an analogy is made
a weathering profile developed from a volcanogenic sandstone between the REE behavior in rivers and in soil solutions,
(Southern Victo~a-Austria), invoked adsorption processes some assumptions may be reached on the REE state in the
to explain the enrichment in REEs, Ba, Sr, and Rb at the Akongo profile. GOLDSTEINand JACOBSEN( 1988) suggested
weathering front. He concluded that REEs and the other ele- that REEs are strongly controlled by pH, with highest levels
ments would be adsorbed on the weathering products of bio- in acidic rivers. EL.DERFIELDet al. ( 1990) showed that 0.45
tite, mainly swelling clays such as vermiculite in which up pm-filtered water samples from streams of North Yorkshire
to 10% of (La + Ce + Nd + Y) were present. However, no have a strong negative correlation between Nd and pH. They
particular attention was paid to study the presence of clay suggest that there are two major pools for the REE in rivers:
sized REE-bearing minerals in the weathered sandstone. a colloidal reservoir of REE-rich particles having a shale-like
pattern and a dissolved reservoir characterized by HREE en-
richment. Recently, SHOLKOVITZ ( 1992) has clearly dem-
The Upper Profile (Loose Nodular Horizon, Iron Crust, onstrated this fact, showing that both the solution and col-
and Mottled Clay Horizon) loidal phases are HREE and LREE enriched, respectively,
reiative to the continen~l crust. Moreover, river colloids are
In the upper horizons (loose nodular horizon, iron crust, enriched in Ce, inducing negative Ce-anomalies in the filtered
and mottled clay horizon) coupled mineralogical and geo- river water. In the loose nodular horizon, soil solutions rich
chemical investigations of REEs are difficult because of the in COZ and organic acids have a pH between 4 and 5. These
strong weathering of this zone and the opacity caused by the conditions would favor both the dispersion and the leaching
presence of iron oxides and clay minerals. of clay particles within the profile. Considering the REE be-
havior in rivers, an assumption of LREE migration associated
with clays may be made in this part of the profile. Neverthe-
Iron crust and mottled clay horizon
less, the loose nodular horizon of laterites which is the result
In contrast to the mottled clay horizon, a distinction be- of poorly understood polyphased weathering processes, is a
tween white and red domains has not been made within the place where the major and trace element behavior is difhcult
iron crust. Calculations based on Th immobility show REEs to define.
leaching in the iron crust to be similar to that observed in
CONCLUSIONS
the loose nodular horizon. The REEs are leached at the top
of the mottled clay horizon (zone c), whereas there is a jux- 1) The REE and Th mass balance calculations for the host-
taposition of accumulation and leaching zones at the bottom rock minerals show that at least 70% of the LREEs and
(zones a and b). The white domains, characterized by both 50% of the HREEs are found in allanite, apatite, titanite,
illuviation and iron loss, are enriched in REEs while the red and epidote and at least 50% of the Th is controlled by
domains, characterized by acidification generated by iron these same accessory minerals. These accessory phases
precipitation (AMBROSI et al., 1986), are depleted in REEs. represent only about 2 wt% of the whole rock and are
REEs may be transported by clay scavenging. destroyed during the early stages of weathering. Most of
Geochemistry of REEs and Th in a lateritic soil profile 4433
the REEs are released in the soil and are directly influenced and GUILLET B. ( 1990) C&urn anomalies in lateritic profiles.
by weathering processes. Geochim. Cosmochim. Acta S4,781-795.
BRIMHALL G. H., ~.EWISC. J., FORD C.,BRA’I-IJ., TAYLORG., and
2) Comparison of the compositional change of Zr, Ti, and WARIN 0. ( 199 1) ~anti~tive geochemical approach to pedo-
Th as a fun~ion of the apparent density of the different genesis: importance of parent material,volum~~c expansion, and
zones of the saprolite shows that Th is the least mobile eolian influx in later&&on. Geoderma 51,5 l-9 1.
element and may be used as a tracer to calculate mass BYRNER. H. and KIM K. H. ( 1990) Rare earth element scavenging
in seawater.Geochim. Cosmochim. Acta 54,2645-2656.
balance.
CRAMER J. J. and NESBITT H. W. ( 1983) Mass-balance relations
3) The mobility of Zr and Ti is shown by the weathering and trace-element mobility during continental weathering of var-
features on zircon grains and the presence of Ti with Ce ious igneous rocks. Sci. G.?ol., Mt?m. 73,63-73.
in the clayey seams. Thorium is precipitated in situ as an CUNEY M. and FRIEDRICHM. ( 1987 1 Phvsicochemical and crvstal-
oxide after the breakdown of the primary minerals. Sec- chemical controh on accessory mineral pamgenesis in graniioids:
Implications for uranium metallogenesis. Bull. Mi&al. 110,235-
ondary thorianite was identified. 247.
4) The REEs are fractionated and redis~but~ in the sap- DEER W. A., HOWIE R. A., and ZUSSMANJ. ( 1985) An Intr~~ction
rolite. The REEs are leached in the most weathered zones to the Rock Forming ~ineraIs. Longman.
and accumulated in the weathered syenite and in the white DUDDY 1. R. ( 19801 Redistribution and fractionation of rare earth
and other elements in a weathering profile. Chem. Geol. 30,363-
clayey seams. The weathering products, which are com-
381.
plex assemblages of aluminous REE phosphate and poorly EDIMOA. ( 1985) Le massif syenitique d’Akongo-Lolodorf(Sud-Ca-
crystallized Ti-Fe-Al-Si phases, are strongly enriched in meroun ) Interp&ation des anomalies radiomttriques. Compa-
LREEs. There is a transport of the REE in solution within raison avec l’arc sytnitique Mont des El6phants-Rocher du Loup.
the saprolite which leads to ( 1) the deposition of cerianite Ph. D. dissertation, Univ. Orleans.
ELDERFIELDH., HAWKESWORTHG. H., GREAVESM. J., and CAL-
(CeOz) produced by the pH and Eh changes when the VERT S. E. ( 198 I ) Rare earth element geochemistry of oceanic
water table moves and (2) the precipitation of the other fe~oman~n~ nodules and associated sediments. Geochim. Cos-
RED, especially LREEs, as aluminous phosphates de- moch~m. Acta 45,5 13-528.
pleted in Ce at the bottom of the profile. Aluminous ELDERFIELDH., UPSTILL-GODDARDR., and SHOLKOVITZE. R.
( 1990) The rare earth elements in rivers, estuaries and coastal seas
phosphates without Ce-anomalies are also formed during
and their significance to the composition of ocean waters. Geochim.
the incipient weathering of the REE-bearing accessory Cosmochim. Acta !%I,97 l-99 1.
minerals. ELPRINCEA. M., VANSELOWA. P., and SFOSITOG. ( 1980) Heter-
5) In the bottom of the mottled clay horizon, there is a jux- ovalent, ternary cation exchange equilibria: NH4+-Ba’+-La3+ ex-
taposition of leaching zones (red domains) characterized change on montmorillonite. Soil Sci. Sot. Amer. J. 44,964-969.
FOURCADES. and ALLEGREC. J. ( 198 I ) Trace elements behavior
by the dissotution of kaolinite andfor halloysite and ac- in granite genesis: a case study. The &c-alkaline plutonic asso-
cumulation zones (white domains) characterized by the ciation from the Querigut Complex (Pyre&es, France). Cant&.
illuviation of the same clay minerals. mineral. Petrol. 76, f 77- i 95.
6) The majority of the REEs are leached from the Fe-rich GARDNERL. R. ( 1980) Mobilization of Al and Ti during weathering-
upper horizons (loose nodular horizon, iron crust, and isovolumetric geochemical evidence. Chem. Geol. 30, 151-165.
GOLDSTEINS. J. and JACOBSENS. B. f 1988) Rare earth elements in
the top of the mottled clay horizon). Transport of the river water. Earth Planet. Sci. Lett. 89, 35-47.
REE associated with the clayey phase may be possible. GOVINDARAJUK. and MEVELLEG. ( 1987) Fully automated disso-
lution and separation methods for inductively coupled plasma
Acknowledgments-We are most grateful to the Yaoundt University atomic emission spectrometry rock analysis. Application to the
Soil Science staff and ORSTOM for facilitating the sampling at the determination of rare earth elements. Plenary lecture. J. Anafy.
Akongo site. Annie Michard, Paul All&,Kupusami Govindaraju, Guy Atomic. Spectromeiry 2, 6 15-62 1.
Mevelle, Jean-Marie Claude, Alain Kohler, Jean-Michel Bertrand, GROMET L. P. and SILVER L. T. ( 1983) Rare earth element distri-
Jacques Yvon, and Frederic Viellieras are thanked for their technical butions among minerals in a granodiorite and their petmgenetic
assistance with IDTiMS, SIMS, ICPMS, electron microprobe analyses, impIi~tions. Geochim. ~osm~h~rn. Acta 47,925-939.
scanning electron microscopy, and data processing. M. Wiedendeck HENDERSONP. ( 1984) General geochemical properties and abun-
and L. Reisbek are thanked for their review of the manuscript and dances of the rare earth elements. In Rare Earth Element Geo-
the two GCA reviewers, P. Jakes and 1. R. Duddy, for their comments. chemistry (ed. P. HENDERSON), pp. l-32. Elsevier.
This project has been partly supported by a DBT-INSU grant ( AP HERBILLONA. J. and NAHON D. ( 1988) Laterites and laterization
89/3828). processes. In Iron in Soils and Clay Minerals. (ed. J. W. STUCKI
et al.); NATO AS1 Series, 779-796. D. Reidel Pub]. Co.
Editorial handling: T. Paces HINTON R. W. and UPTON B. G. J. ( 199 1) The chemistry of zircon:
Variations within and between large crystals from a syenite and
REFERENCES
alkali basalt xenoliths. Geochim. Cosmochim. Acta 55,3287-3301.
AAGAARDP. ( 1974) Rare earth element adsorption on clay minerals. JARVISK. E. ( 1988 ) Inductively coupled plasma mass spectrometry:
Burl/. Group. Franc. Argiles 26, 193- 199. A new technique for the rapid or ultratrace level dete~ination of
AMBROSIJ. P., NAHON D., and HERB~LLONA. J. (1986) The epi- the rare earth elements in geological material. Chem. Geol. 68,
genetic replacement of kaolinite by hematite: laterite-petrographic 31-39.
evidence and the mechanisms involved. Geoderma 37, 283-294. JONASSONR. G., BANCROFTG. M., and NESBITTH. W. ( 1985 )
BANFIELDJ. F. and EGGLETON R. A. (1989) Apatite replacement Solubilities of some hydrous REE phosphates with implications
and rare earth mobilization, fractionation, and fixation during for diagenesis and sea water concentrations. Geochim. Cosmochim.
weathering. Clays Clay Miner. 37, I I3- 127. Acta 49.2 133-2 139.
BILONG P. ( 1988) Genese et developpement des sols ferrallitiques KAUFMANN A. (1969) The 232Th concentration of surface ocean
sur syenite alcaline potassique en milieu forestier du centre-sud water. Geochim. Cosmochim. Acta 33,7 17-724.
Cameroun. Comparaison avec les sols ferrallitiques d&elopp&s sur LANGMUIRD. and HERMANNJ. S. ( 1980) The mobility of thorium
roches basiques. Ph.D. dissertation, Univ. Yaoundt. in natural waters at low temperatures. Geochim. Cosmochim. Acta
BRAUN J. J., PAGEL M., MULLER J. P., BILONG P.,MKHARD A., 44, 1753-1766.
4434 J. J. 3raun et al.
LAPIQUEF., CHAMPENOIS M., and CHEILLETZA. ( 1988) Un ana- thorium, and rare earth elements in two radioactive granites of
lyseur videographique interactiE description et applications. BUN. the Vosges, France. Mineral. Mug. 46, 149-16 1.
Mineral 1il, 679-687. PLILERR. and ADAMSJ. A. S. ( 1962) The distribution of thorium
MARlAN A. M. ( 1989) E~~~Kuz&geology of rare earth minerals. and uranium in a Pennsylvanian weathering pro&e. Ge~him.
In Geochemistry and ~i~alo~ of Rare Earth Elements (ed. Cosmochim. Acta 26, 1t37- I 146.
B. R. LIPIN and G. A. MCKAY). Rev. Mirteral. 21.309-334. PRICER. C., GRAYC. M., WKSONR. E., FREYF. A., and TAYLOR
MCLENNAN S. M. ( 1989) Rare ear& elements in sedimentary rocks: S. R. ( I99 1) The effects of weathering on rare earth element, Y
influence of provenauce and sedimentary processes. In Geochem- and Ba abundances in Tertiary basalts from southeastern Australia.
istry and Mineralogy of Rare Earth Elements (ed. B. R. LIPIN and Chem. Geol. 93,245-265.
G. A. MCKAY ); Rev. Mineral. 21, 169-196. ROALDSET E. ( 1974) Lanthanide distributions in clays. Bull. Group.
MICHARDA., MONTIGNY R., and SCHLICH R. ( 1986) G~hemist~ Franc. Argiles 26,201-209.
of the mantle beneath the Rodriguez Triple Junction and the South- SAWKAW. N., BANFIELDJ. F., and CHAPPELLB. W. (1986) A
East lndiin Ridge. Earth Planet. Sci. Lett. 78, 104-l 14. weathering-relatedorigin of widespreadmonazite in S-type granites.
MID~ELBURGJ. J., VAN DER WEIJDENC. H., and WOITTIEZ Geochim. Cosmochim. Acta !W, 17I- 175.
J. R. W. ( 1988) Chemical processesaffectingthe mobility of major, SHOLKOVITZ E. R. ( 1989) Artefact associated with the chemical
minor and trace elements during weathering of granitic rocks. leaching of sediments for the rare earth elements. Chem. Geol. 77,
Chem. Geol. 68,253-273. 47-51.
MILLOTG. and BONIFASM. ( 1955) Transformation isovolumetrique SHOLKOVITZ E. R. ( 1992) Chemical evolution of rare earth elements
darts Ies phenom&es de fat&i&ion et de ~uxiti~tion. Bull. Serv. fractionation between colloidal and sotution phases of filtered river
Carte G&of.Alsace-Lorraine 10, 17-2 1. water. Earth Planet. Sci. Lett. 114, 77-84.
MILNE~A. R. and FITZPATRICKR. W. ( 1989) Titanium and zir- SMEDLEYP. L. ( 1991) The geochemistry of rare earth elements in
conium minerals. In Minerals in Soil Environments, 2nd ed. (ed. groundwater from the Cammenellis area, southwest England.
J. B. DIXON and S. B. WEED),pp. 103l-1206. Soil Sci. Sot. Amer. Geochim. Cosmochim. Ada s&2767-2779.
MOINEB., RAKOTONDRATSIMBA C., and CUNEYM. ( 1985) Les py-
STUMMW. and MORGANJ. J. ( 1981) Aquatic Chemistry. An intro-
roxenites Buranothorianite du Sud-Est de Madagascar: conditions
duction emphasizing chemical equilibria in natural wafers, 2” edi-
physi~himiqu~ de la m~mato~. BuK ~~in~ra~.108, 32%
tion. Wiley Interscience.
340.
MUMP~ONF. A. and ROYR. ( 196t ) Hydrotbermal stability studies TAYLORS. R. and MCLENNANS. M. ( 1985) The Continental Crust:
of the zircon-thorite group. Geochim. Cosmochim. Acta 21,2 17- Its Composition and Evolution, An Examination ojthe Geochem-
238. ical Record Preserved in Sedimentary Rocks. Blackwell.
MURALSA. V., PARTHA~ARATHY T. M., and SANKAR DAS M. ( 1983) TURNERD. R., WHITFIELDM., and DICKSONA. Cl. ( 1981) The
Trace element characteristics, REE patterns and partition coeffi- equilibrium speciation ofdissolved components in fresh water and
cients of zircons from different geological environments. A case seawater at 25°C and I atm pressure. Geochim. Cosmochim. Ada
study on Indian zircons. G~him. Cosm~him. Acta 47, 2047- 45,855-88 1.
2052. WOOD S. A. ( 1990) The aqueous geochemistry of the rare earth
NESBITTH. W. ( 1979) Mobility and fractionation of rare earth ele- elements and yttrium. I. Review of available low-temperature data
ments during weathering of a granodiorite. Nature 279,2062 10. for inorganic complexes and the inorganic REE speciation of nat-
PAGELM. ( 1982) The mineralogy and geochemistry of uranium, ural waters. Chem. Geol. 82, 159-186.