Chemical Geology 193 (2003) 273 – 293

www.elsevier.com/locate/chemgeo

Argillization processes at the El Berrocal analogue granitic system

(Spain): mineralogy, isotopic study and implications for the
performance assessment of radwaste geological disposal
L. Pérez del Villar a,1, E. Reyes b,*, A. Delgado b, R. Núñez b,
M. Pelayo a, J.S. Cózar a
a
Centro de Investigaciones Energéticas, Medio Ambientales y Tecnológicas (CIEMAT), IDAE,
C.H.E., Avda. Complutense 22, 28040 Madrid, Spain
b
Departamento Ciencias de la Tierra y Quı́mica Ambiental. Estación Experimental del Zaidı́n (CSIC),
Profesor Albareda 1, 18008 Granada, Spain
Received 17 April 2001; accepted 27 September 2002

Abstract

The El Berrocal granite/U-bearing quartz vein (UQV) system has been studied as a natural analogue of a high-level
radioactive waste repository. The main objective was to understand the geochemical behaviour of natural nuclides under
different physicochemical conditions. Within this framework, the argillization processes related to fracturing and formation of
the uranium – quartz vein were studied from a mineralogical and isotopic standpoint in order to establish their temperatures of
formation and thus complete the geothermal history of the system. For this purpose, d18O values were determined for pure
mineral from the unaltered granite and quartz from the uranium – quartz vein, as well as for mixture samples from the
hydrothermally altered granite (sericitised granite) and clayey samples from fracture fillings, including the clayey walls of the
uranium – quartz vein. The isotopic signature of quartz from the uranium – quartz vein and the monophasic nature of its fluid
inclusions led us to conclude that the isotopic signature of water in equilibrium with quartz was approximately in the range
from
8.3xto
5.7xV-SMOV, its temperature of formation being around 85 – 120 jC. The d18O values of pure sericite
from the hydrothermally altered granite, calculated by means of the oxygen fraction molar method, indicate that its
temperature of formation, in equilibrium with the aforementioned waters, is also in the range from 70 jC to approximately
120 jC. Clays from fracture fillings and clayey walls of the uranium – quartz vein are usually mixtures, in different
proportions, of illite, approximately formed between 70 and 125 jC; two generations of kaolinite formed at approximately
90 – 130 jC and at around 25 jC, respectively; smectite, formed at V 25 jC; and occasionally palygorskite, formed either
between 30 and 45 jC or 19 and 32 jC, depending on the fractionation equation used. These data suggest that sericite from
the hydrothermally altered granite, quartz from the uranium – quartz vein, illite and the first generation of kaolinite from the
fracture fillings resulted from the same hydrothermal process affecting the El Berrocal granite in relation to fracturing. Under
certain physicochemical conditions (T c 100 jC, pH c 8 and log [H4SiO4] between
4 and
3), illite and kaolinite can be
paragenetic. As a result of weathering processes, smectite was formed from hydrothermal illite and inherited albite under

* Corresponding author. Fax: +34-958-129600.

E-mail addresses: [email protected] (L. Pérez del Villar), [email protected] (E. Reyes).
1
Fax: +34-913-466542.

0009-2541/03/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved.
PII: S 0 0 0 9 - 2 5 4 1 ( 0 2 ) 0 0 3 2 3 - 6
274 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293

alkaline weathering, while the second generation of kaolinite was formed from smectite, under acid conditions and close to
the sulphide-rich uranium – quartz vein. Palygorskite is an occasional mineral formed probably either during the thermal tail
of the above-described hydrothermal process or during weathering processes. In both cases, palygorskite must have formed
from alkaline Si – Mg-rich solutions. Finally, these data and processes are discussed in terms of natural analogue processes,
drawing some implications for the performance assessment of a deep geological radwaste repository (DGRR).
D 2003 Elsevier Science B.V. All rights reserved.

Keywords: Clays; Quartz; Fracture fillings; Granite; Oxygen isotopes; Natural analogue

1. Introduction the clay minerals from its clayey walls and from the
fracture fillings of the system, in order to:
Over the last 20 years, the international scientific
community has paid particular attention to studying Establish the approximate temperature of the
natural analogues to gain information on the condi- hydrothermal processes that affected the El Berro-
tions of mobilisation and retention of natural nuclides cal granite located close to the UQV, and of this
and other analogue elements in geological formations, quartz vein itself.
especially igneous, clayey and salt formations (Chap- Differentiate the argillization processes in fracture
man and Smellie, 1986; Chapman, 1994; Choppin and fillings, according to the formation temperature of
Wong, 1996; Smellie et al., 1997; Hidaka and Hol- the different clay minerals present in these
liger, 1998; Miller et al., 2000). Granite formations fractures.
have been deemed highly favourable for geological
disposal of radioactive wastes (Chapman and Smellie, All these data have been interpreted taking into
1986; Parneix, 1992). In this context, the mineralog- account the geological and geochemical background
ical and geochemical analysis of the secondary min- of the system, emphasizing the relationships between
eral phases from granite formations, particularly of alteration and migration/retention processes affecting
clay minerals, is informative for charting the history natural radionuclides. Some implications for the per-
of the alteration processes of granite rocks and for formance assessment of a DGRR in granite forma-
determining the role of these minerals in the retention tions have been stated based on the geological
of natural radionuclides. analogies and analogue processes observed at the El
The El Berrocal granite/U-bearing quartz vein Berrocal natural system.
(UQV) system has been studied as a natural analogue
of a deep geological radwaste repository (DGRR), the
main objective being to understand the distant past, 2. Geological and geochemical background
recent past and present behaviour of natural nuclides
under real conditions, as well as to interpret it in terms The El Berrocal system is located some 90 km
of rock – water interaction (Rivas et al., 1997). In this southwest of Madrid, near the small town of Nom-
context, the low-temperature hydrothermal alteration bela, in the province of Toledo (Spain). The system
and weathering events in the El Berrocal system have takes its name from the El Berrocal granite pluton,
been comprehensively studied, since they caused the which is located in the central part of the Centro-
most important mobilisation, migration and retention Iberian Zone, within the Spanish Hercynian Belt
processes that affected the natural radionuclides in the (Julivert et al., 1972). The main granite facies of the
system (Pérez del Villar et al., 1994, 1995, 1996a,b, El Berrocal pluton, the so-called El Berrocal facies, is
1997; Reyes et al., 1998). the host-rock of a U sulphide-bearing quartz vein,
The aim of this work is to study the mineralogical mined in the 1960s and known as ‘‘El Berrocal U
and isotopic composition of sericite from the hydro- mine’’ (Arribas, 1965) (Fig. 1). Together, the host-
thermally altered granite, of quartz from the UQV and rock and the UQV, make up the so-called El Berrocal
 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293 275

Fig.1. Geographical location and geological scheme of the El Berrocal pluton: (a) biotitic granite of San Vicente type; (b) two mica porphyritic
granite (El Berrocal facies); (c) fine-grained leucogranitic facies; (d) Almorox-Navamorcuende aplitic dyke; (e) uranium quartz vein (UQV); (f)
undifferentiated Tertiary sediments (sands, clays and conglomerates).

system. The geological materials that form the El Notable among the accessory minerals are ilmenite,
Berrocal system are: (i) the El Berrocal facies or zircon, monazite, xenotime, apatite, uraninite, cassi-
reference granite, (ii) the hydrothermally altered gran- terite and primary sulphides. Muscovite II, fluorite,
ite, (iii) the weathered hydrothermally altered granite, sericite, chlorite, rutile, anatase, K-feldspar and albite
(iv) the weathered reference granite, (v) the UQV, II comprise the main secondary minerals, formed
including its clayey walls, and (vi) the fracture fillings probably during the deuteric and/or early postmag-
(Fig. 2). matic hydrothermal processes that affected the pluton.
The reference granite was classified mineralogi- The geochemical features indicate highly evolved
cally as a weakly altered alkaline-feldspar granite with hypocalcic granite, enriched in silica, phosphorus
two micas, muscovite being dominant over biotite. and alumina. This granite is fertile in U and belongs
276 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293

Fig. 2. N – S cross-section of the El Berrocal system showing the uranium mine gallery, and sampling points on boreholes.

to the ilmenite granite series or to the S-type granites 297 F 1 Ma ago, with an initial 87Sr/86Sr ratio of
(Pérez del Villar et al., 1994). 0.7175 F 0.0029 (Pérez del Villar et al., 1996a,b). This
The Rb – Sr isotopic data, as well as the structural strontium ratio is typical of melted granite derived
and textural relationships among the El Berrocal geo- from the continental crust (Faure and Powell, 1972).
logical materials and the mineral paragenesis indicate Furthermore, the El Berrocal facies was affected by an
that the El Berrocal pluton intruded approximately initial set of hydrothermal events at high temperature
 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293 277

(T>350 jC) defined as deuteric and/or early postmag- fracture fillings (see Fig. 2). Geothermometric studies
matic alteration processes, which transformed the of Mn-rich calcites, based on d18O values, indicate at
original facies into the so-called reference granite. least two generations of carbonates, the first formed
These processes are weak but pervasive since they under hydrothermal conditions, at a temperature range
affect the whole granite mass. The internal isochrone of between 25 and < 100 jC, and the second at
determined from apatite, biotite, albite, K-feldspar and ambient temperature ( < 25 jC) (Reyes et al., 1998).
bulk-rock sample of the El Berrocal facies suggests Almost all the samples analysed were mixtures of the
that the bulk rock remained as a closed system for the two types of carbonates. The isotopic signature of
Rb – Sr pair during late alteration processes, and that carbon in both generations of carbonates also indi-
the minerals were closed for Sr isotopes around cates the influence of edaphic carbon (Reyes et al.,
289 F 1 my ago (Pérez del Villar et al., 1996a,b). 1998).
Consequently, the difference between the intrusion As a function of the physicochemical conditions
age and the age estimated from the internal isochrone under which the alteration processes took place in the
(9 my) represents the time during which the first set of system, these processes produced mobilisation, migra-
hydrothermal alteration processes took place (Reyes at tion and retention, mainly by precipitation, of U (IV)
al., 1998). and Th (IV), as analogues for Pu (IV) and Np (IV); U
The second hydrothermal event recorded in the (VI), for Np (V and VI); light rare-earth elements
system is related to N80E and N110E fracture fami- (LREE), for Am, Cm and Np (III) (Chapman and
lies. This process transforms the reference granite Smellie, 1986) and Zr as a fission product (Pérez del
close to the fractures into the so-called hydrothermally Villar et al., 1994, 1995; Reyes et al., 1998).
altered granite. The alteration is evidenced by the During the first set of deuteric and/or hydrothermal
generalised sericitisation of feldspars, mainly albite, events, uraninite, monazite, xenotime and apatite in
the presence of secondary ankerite and calcite, which the granite were partially destabilised by F-rich fluids.
are scattered in the rock, and the yellow-green colour Thus, when uraninite crystals were surrounded by
of the rock (Pérez del Villar et al., 1994). Based only fluorite they were totally or partially pseudomorphised
on isotopic data, the minimum formation temperature by complex U silicophosphates. Under the same
of secondary carbonates is between 61 and 72 jC for circumstances, monazite was totally or partially de-
ankerite, and between 35 and 52 jC for calcite (Reyes pleted in REE, xenotime showed U and REE poor al-
et al., 1998). In both carbonates, the influence of tered zones and apatite was partially dissolved. The
edaphic carbon is evident, being stronger in the latter presence of U silicophosphates coating chloritized
than in the former (Reyes et al., 1998). biotite and filling cleavage planes of distorted musco-
In relation to this process, sulphide-bearing quartz vite and microfissures in quartz implies the migration
veins and their clayey walls were formed, filling the of U away from uraninite, as well as its retention in
aforementioned fractures. The U mineralisation took the rock matrix, respectively. On the contrary, Th from
place after the fracturing of these veins, mainly in monazite remains immobile, and no secondary REE-
those oriented N110E. The uraniferous paragenesis is bearing minerals have been found. These textural
formed by pitchblende, pyrite, barite and carbonates, features suggest that F is responsible for the alteration
the latter being detected by the presence of dissolution of U, Th and REE-bearing minerals, as well as for the
moulds filled with Fe (Mn) oxyhydroxides (Reyes et transport of U and REE, probably as fluoride com-
al., 1998). Fracture fillings, composed mainly of clay plexes. This hypothesis was confirmed by the pres-
minerals, Mn-rich calcite and minor secondary quartz, ence of secondary fluorite and U silicate compounds
result primarily from the interaction between fracture filling dissolution voids in albite (Pérez del Villar et
gouges and hydrothermal solutions. al., 1995).
The intensity and depth of the weathering effects The second hydrothermal process, responsible for
on the El Berrocal system are related to the degree of the hydrothermally altered granite and quartz veins,
fracturing and hydrothermal alteration of the granite, mobilised and also precipitated U, Th, REE and Zr.
as well as the mineralogical composition of the hydro- The mineralogical evidence for these processes are: (i)
thermally altered granite, UQV, its clayey walls, and the appearance of U silicophosphates lining and filling
278 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293

microfissures of secondary pyrite; (ii) precipitation of On the U ore body, the effects of weathering are
botryoidal mineraloids with a variable chemical com- represented by strong oxidation of pitchblende and
position from U>Th to Th>U silicophosphates; (iii) sulphides, dissolution of carbonates and neoforma-
presence of Th and Zr silicates filling microfissures in tion of secondary U minerals such as uranotyle,
quartz; (iv) neoformation of florencite and (v) the autunite, in secular equilibrium, torbernite and ura-
general increase of U, Th, Ce, Y, and Zr contents in nocircite. Furthermore, U is also adsorbed onto Fe
the hydrothermally altered granite. In this case, the oxyhydroxides.
chemical agent that mobilised U, Th, REE and Zr is On fracture-filling materials, similar weathering
not as clearly defined as in the former case. However, effects have been found. Uranium forms uranyl
the presence of secondary carbonates in this hydro- silicates and phosphates, such as uranotyle, autunite
thermally altered granite implies that carbonate sol- and phosphuranylite, and occasionally it is ad-
utions were at work in the system (Pérez del Villar et sorbed onto Fe oxyhydroxides. Currently, uranyl –
al., 1995). carbonate complexes are the predominant uranium
In relation to this second hydrothermal event, U species in oxidising groundwaters of the system
mineralised the main quartz vein. Its mineral para- (Pérez del Villar et al., 1995; Reyes et al., 1998).
genesis (pitchblende, carbonates, pyrite and minor
barite) suggests that U from uraninite scattered in
the granite was oxidised, mobilised and transported, 3. Sample collection and experimental methods
as uranyl –carbonate complexes, downwards through
the fractures. These solutions, reheated at depth by 3.1. Sampling
tectonic events, moved upwards, probably by con-
vective flow. After destabilisation of uranyl – carbo- Twelve samples from the reference granite (M)
nate complexes, caused by rock – water interaction were taken along the core of borehole 16, between
processes and a drop in temperature and pressure, 6.27 and 590.15 m in depth (Table 1, see Fig. 2) for
uranyl cations were probably reduced by H2S, pre- the separation of pure quartz, albite and muscovite, as
cipitating pitchblende, carbonates and pyrite. The well as mixtures of chlorite and biotite, since both
coexistence of pyrite and barite suggests that the phyllosilicates were closely associated in the samples.
precipitation occurred between
200 and
300 mV Ten samples from the hydrothermally altered granite
(Krivovichev, 1979). The same process also affected (DTG) were taken along borehole S-13, between
the hydrothermally altered granite close to the UQV 84.93 and 100.65 m in depth, on both sides of the
and, to a less extent, the intragranitic fractures. UQV (Table 1, see Fig. 2). For reference granite, the
The effects of weathering on the uranium minerals average sample size was cylinder-shaped, 20 cm long
of each part of the system may be summarised as and 8.6 cm in diameter, while for hydrothermally
follows: altered granite, the average size was cylinder-shaped,
1 cm long and 8.6 cm in diameter.
On the uraninite from the reference granite, the Nine samples from the weathered hydrothermally
effects vary depending on the nature and fracturing altered granite (LAR), six samples from the UQV
degree of the host minerals. When uraninite is (DQ) and four from its clayey walls (DAR) were
included in feldspars, the former is usually dissolved taken from the U mine gallery (Table 1, Fig. 3).
and only the outlines of the crystals remain, with Fracture fillings (RF) were sampled on the hole-cores
residual Th – Ca silicophosphates inside. Uranium of S-13, S-14, S-15, S-16 and S-17 (Table 1, see
from uraninite, forming U silicophosphates, usually Fig. 2).
fills adjacent microfissures or is adsorbed, togeth-
er with P, onto Fe oxyhydroxides. 3.2. Analytical methods
On the hydrothermally altered granite, weathering
mainly redistributed the U that forms large autunite For mineral separation from samples of the refer-
crystals and/or is adsorbed onto secondary sericite ence granite, 125 – 60 Am fractions were subjected to
and Fe oxyhydroxides. electromagnetic methods and dense liquids. The pow-
 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293 279

Table 1
Sampling from the El Berrocal site for mineralogical and isotopic purposes
Reference granite Hydrothermally altered granite U mine Fracture filling
(S-16) Depths (S-13) Depths gallery Samples Depths
Samples (m) Samples (m) (m)
M-2 6.27 – 6.45 DTG.10554 84.93 LAR-1 Hydrothermally RF.13-6 87.30
M-3 8.59 – 8.79 DTG.10555 85.40 LAR-2 altered granite, RF.13-7 87.86
M-4 13.53 – 13.73 DTG.10556 86.10 LAR-3 affected by RF.13-14 108.02
M-5 17.62 – 17.82 DTG.10557 86.24 LAR-4 weathering RF.13-17 113.07
M-6 21.60 – 21.90 DTG.10558 87.47 LAR-5 RF.14-10 106.85
M-10 48.70 – 40.90 DTG.10559 88.96 LAR-6 RF.15-17 129.50
M-16 115.00 – 115.19 DTG.10560 90.15 LAR-7 RF.15-18 132.29
M-26 185.09 – 185.20 DTG.10561 91.60 LAR-8 RF.15-29 143.72
M-39 300.16 – 300.36 DTG.10562 96.30 LAR-9 RF.16-54 220.31
M-49 399.90 – 400.10 DTG.10563 100.65 DQ-1 Uranium RF.16-55 221.31
M-59 500.34 – 500.54 DQ-2 quartz vein RF.16-8 238.16
M-69 589.95 – 590.15 DQ-3 RF.16-10 413.58
DQ-4 RF.16-21 471.04
DQ-5 RF.16-23 473.15
DQ-6 RF.16-40 569.15
DAR-1 Clayey Walls RF.16-45 596.65
DAR-2 RF.17-12 70.85
DAR-3 RF.17-15 76.83
DAR-4 RF.17-27 156.05
Types of samples and location.

der X-ray diffraction (XRD) method was used for fraction. An aliquot of the < 60 Am fraction was used
checking the purity of minerals. Mineralogical com- to determine the mineralogical composition, and ano-
position of the hydrothermally altered granite was ther to separate the < 2 Am fraction, according to the
determined by combining optical microscopy with Stoke’s law.
the powder XRD method, by using a plane-rotating The < 60 and < 2 Am fractions were characterised
sample holder, and by scanning as well as transmi- by XRD, combining the powder and oriented ag-
ssion electron microscopy (SEM and TEM) coupled gregate methods. The former was applied to both
to an energy dispersive X-ray analysis system (EDX). fractions, while the second, with ethylene glycol,
The mineralogical composition of < 2 Am fractions sulphoxide dimethyl and thermal treatments, was
from this type of granite was also determined. applied only to the < 2 Am fractions. For semi-
For preparation and study of clayey materials from quantification of the different clay minerals in the
the walls of the UQV and fracture fillings, the fol- samples, the Bradley and Grim (1961) method was
lowing method was used. After being dried at 30 jC, used. The percentage of apatite was calculated by
the samples were smoothly ground in a porcelain mor- stoichiometry, under the assumption that the P2O5/
tar to obtain a concentrate of the neoformed mine- CaO ratio is 0.73 in the apatite of the reference granite
rals and another of the inherited minerals, under the (Pérez del Villar et al., 1992). The clay minerals in the
assumption that the former were preferentially con- < 2 Am fractions were more accurately identified using
centrated in the soft fraction, while the second were TEM + EDX, thus facilitating the calculation of their
concentrated in the hard one. A 150-mesh (104 Am) respective structural formulae.
sieve was used to separate the two fractions. Only For isotopic measurements, 12 samples from the
soft fractions ( < 104 Am) were studied, after groun- reference granite were used for determining 43 d18O
ding and sieving to a grain size of < 60 Am. Then, values on pure quartz (12), albite (10), muscovite (12)
the carbonates were eliminated to clean the silicate and biotite – chlorite mixtures (9). Similarly, 10 ser-
280 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293

Fig. 3. Bird’s eye and side views of the inner part of uranium mine gallery (from 36 m to the end) showing fractures, quartz veins, including the
UQV and its clayey walls, and sampling points on weathered hydrothermally altered granite (LAR series), quartz veins (DQ series) and clayey
walls (DAR series).

icite-rich samples from the hydrothermally altered clayey mixtures from fracture fillings, more or less
granite, 9 samples from the weathered hydrothermally enriched in illite, smectite, kaolinite or palygorskite,
altered granite, 6 quartz samples from the UQV, 4 were also analysed. The d18O values of all the
clayey samples from its walls, and 19 samples of samples were determined using a Finnigan-MAT

Table 2
Oxygen isotope composition of pure minerals from the reference Table 3
granite (borehole S-16) (d values as xvs. V-SMOW) Mineralogical composition and d18OT of samples from the hydro-
Samples Q-d18O Ab-d18O Mus-d18O (B + Chl)-d18O thermally altered granite intersected by borehole S-13 (d values as
xvs. V-SMOW)
M.2 11.4 – 8.4 10.3
M.3 11.6 9.7 9.0 9.2* Samples Seric Q Ab K-Fd Chl Cc d18OT
M.4 11.7 8.9 9.9 9.0 DTG.10554 67 24 9 0 0 0 12.2
M.5 12.0 7.9 9.5 – DTG.10555 62 31 7 0 0 0 8.1
M.6 12.4 9.7 7.9 5.6 DTG.10556 81 16 3 0 0 0 8.5
M.10 11.5 11.1 8.7 – DTG.10557 65 31 4 0 0 0 7.5
M.16 11.5 9.1 10.1 – DTG.10558 34 31 23 10 2 0 10.4
M.26 11.3 10.8 8.0 4.6 DTG.10559 75 15 0 10 0 0 12.8
M.39 11.4 9.5 8.3 8.0 DTG.10560 39 26 0 8 0 27 12.4
M.49 11.6 – 9.2 5.2 DTG.10561 34 22 11 8 0 25 12.9
M.59 11.0 10.0 8.4 7.1* DTG.10562 40 9 20 12 7 12 13.5
M.68 10.8 8.1 8.0 7.1** DTG.10563 52 7 15 7 19 0 13.0
Q: quartz; Ab: albite; Mus: muscovite; B: biotite; Chl: chlorite; Seric: sericite; Q: quartz; Ab: albite; K-Fd: K-feldspar; Chl: chlorite;
* 100% biotite; ** 100% chlorite. Cc: calcite; d18OT: isotopic composition of the bulk sample (x ).
 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293 281

Table 4
Mineralogical composition of granite, quartz and clayey samples taken from the El Berrocal uranium mine gallery
Min Hydrothermally altered granite affected by weathering Quartz from UQV Clayey walls
(%) LAR- LAR- LAR- LAR- LAR- LAR- LAR- LAR- LAR- DQ- DQ- DQ- DQ- DQ- DQ- DAR- DAR- DAR- DAR-
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 1 2 3 4
Q 55 34 24 30 25 30 23 27 41 100 100 100 100 100 100 22 25 50 46
Ab 1 traces 7 4 3 – 3 4 4 – – – – – – 3 – 3 4
K-Fd – 7 15 8 9 8 7 8 – – – – – – – – – – –
Ap traces traces traces traces traces traces traces traces traces – – – – – – traces traces 5 traces
T.Ph. 43 59 54 58 63 62 66 60 54 – – – – – – 74 74 44 50
Q: quartz; Ab: albite; K-Fd: K-feldspar; Ap: apatite (determined by stoichiometry); T.Ph: total phyllosilicates; – : undetectable.

251 mass spectrometer, according to the ClF3 techni- for kaolinite (Land and Dutton, 1978); and
que (Borthwick and Harmon, 1982), later modified by
Vennemann and Smith (1990). These isotopic values 1000lna ¼ 2:6 ð106 T
2 Þ
4:28 ð4Þ
are reported using the d notation in units per mil (x)
for smectite (Savin and Lee, 1988). In addition, two
relative to V-SMOW, d18O values being reproducible
approaches were tested to calculate the temperature of
to F 0.2. Using this procedure, the d18O values of
palygorskite formation. The first, by using the equa-
NBS-30 (biotite) and NBS-28 (quartz) are + 5.1xand
tion for smectite, since experimental fractionation
+ 9.7x , respectively.
equation for palygorskite is not yet available in the
For geothermometric purposes, the following frac-
specialised literature, and smectite and palygorskite
tionation equations were used:
are chemically similar (Torres-Ruiz et al., 1994). The
second, by applying the equation proposed by Gauth-
1000lna ¼ 3:38 ð106 T
2 Þ
3:40 ð1Þ ier-Lafaye et al. (1993).
However, two main problems arise in determining
for quartz (Clayton et al., 1972); the temperature of the formation of these minerals.
Firstly, all the samples are mixtures either of different
1000lna ¼ 2:43 ð106 T
2 Þ
4:80 ð2Þ clay minerals or clay minerals and other granite-
forming silicates, mainly quartz and albite. Secondly,
for illite and sericite (Eslinger and Savin, 1973); no data could be gathered on isotopic signatures of
water in equilibrium with these minerals. However,
1000lna ¼ 2:5 ð106 T
2 Þ
2:9 ð3Þ fluid inclusion studies performed on quartz samples

Table 5
Mineralogical composition and isotopic signature of the < 2 Am fractions from samples taken in the El Berrocal uranium mine gallery
Min Hydrothermally altered granite affected by weathering Quartz from UQV Clayey walls
(%) LAR- LAR- LAR- LAR- LAR- LAR- LAR- LAR- LAR- DQ- DQ- DQ- DQ- DQ- DQ- DAR- DAR- DAR- DAR-
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 1 2 3 4
Q 19 18 17 18 17 14 18 14 12 100 100 100 100 100 100 9 16 13 11
Ab – – – – – – – – – – – – – – – – – – –
K-Fd – 14 23 10 11 12 16 5 17 – – – – – – – – – 18
Ap – – – – – – – – – – – – – – – – – 7 –
Ill 50 31 21 35 32 32 31 43 45 – – – – – 64 27 48 42
Sm 30 35 38 36 40 42 34 30 26 – – – – – – 26 48 32 28
K – 2 1 – – – – – – – – – – – – – 9 – –
d18O 11.2 11.9 13.6 13.7 13.6 14.7 14.4 12.9 8.6 14.2 14.4 15.3 12.7 14.2 12.6 9.7 14.3 10.2 10.1
Q: quartz; Ab: albite; K-Fd: K-feldspar; Ap: apatite (determined by stoichiometry); Ill: illite; Sm: smectite; K: kaolinite; – : undetectable; d18O:
in xV-SMOV.
282 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293

Table 6 and the isotopic signatures of bulk sericite-enriched

Chemical composition and structural formula of sericite – illite s. samples, pure quartz and albite from the reference
str., with < 5% of expanding layers
granite (Lawrence and Taylor, 1971; Ruiz-Cruz and
Samples DAR-3 DAR-4 DAR-3 LAR-1 DAR-3 DAR-3
Reyes, 1998). The theoretical d18O value for pure
Oxides (%) 1 2 3 4 5 6 illite was calculated from d18O signatures of several
SiO2 50.63 48.90 49.62 49.15 51.03 51.29 < 2 Am binary mixtures formed by illite + smectite
Al2O3 32.92 32.53 36.83 38.23 36.54 36.54 from fracture filling samples (RF series). For this,
Cr2O3 – – – – – –
the regression line between d18O values of the binary
Fe2O3 4.62 6.87 2.53 2.03 1.14 2.08
MgO – – – – – – mixtures and the illite/smectite ratios were calculated.
CaO – – – – 0.96 – The d18O value thus determined was also assigned to
K2O 11.83 11.71 11.02 10.59 10.32 10.09 illite from the clayey walls of UQV. The isotopic
Total 100 100 100 100 99.99 100 signature of pure smectite was measured on only one
sample from fracture fillings, formed by only smectite
Si4 + 6.467 6.308 6.265 6.180 6.395 6.413
Al3 + 1.533 1.692 1.735 1.820 1.605 1.587 (RF series). This value was also assigned to smectite
Total 8.000 8.000 8.000 8.000 8.000 8.000 from the clayey walls of the UQV. The theoretical

Al3 + 3.424 3.254 3.747 3.850 3.792 3.797

Fe3 + 0.444 0.667 0.240 0.190 0.110 0.200 Table 7
Cr3 + – – – – – – Chemical composition and structural formulae of beidellitic
Mg2 + – – – – – – smectite
Total 3.868 3.921 3.987 4.040 3.902 3.997 Samples DAR-2 DAR-2 DAR-2

K+ 1.927 1.927 1.775 1.700 1.651 1.610 Oxides (%) 39 40 41

Ca2 + – – – – – – SiO2 55.92 61.53 58.91
Mg2 + – – – – – – Al2O3 33.57 29.24 36.74
Total 1.927 1.927 1.775 1.700 1.781 1.610 Cr2O3 1.10 – –
Fe2O3 3.18 2.86 1.87
D
1.533
1.692
1.735
1.820
1.605
1.587 MgO 3.32 3.40 –
w
0.396
0.237
0.039 0.120
0.294
0.009 CaO 1.55 0.64 1.41
= 1.929 1.929 1.774 1.700 1.899 1.596 K2 O 1.38 1.25 1.07
Si/Al 1.304 1.275 1.142 1.089 1.184 1.191 Total 100.02 100.02 100.02
% 79.47 87.71 97.80 100 84.52 99.43
D: tetrahedral charge; w: octahedral charge; =: interlayer charge; Si4 + 6.700 7.290 6.928
%: of tetrahedral charge. Al3 + 1.300 0.710 1.072
Total 8.000 8.000 8.000

Al3 + 3.440 3.370 4.020

from the UQV have shown that primary fluid inclu- Fe3 + 0.290 0.260 0.165
sions are monophasic and that temperatures lower Cr3 + 0.100 – –
than 100 jC can therefore be inferred. Furthermore, Mg2 + 0.170? 0.370? –
the average d18O value (
7.6x) of the present day Total 4.000 4.000 4.185
meteoric waters at the El Berrocal site (Pérez del
K+ 0.210 0.190 0.161
Villar et al., 1997) was also considered for geother- Ca2 + 0.200 0.220 0.177
mometric purposes. Mg2 + 0.420? 0.23? –
Under these restraints, an approximate d18O value Total 0.830 0.64 0.338
for water in equilibrium with quartz from the UQV
was firstly established. For sericite from the hydro- D
1.300
0.710 1.072
w
0.170
0.37 0.555
thermally altered granite, an approximate theoretical = 1.450 ? 1.090 0.517
d18O value was calculated taking into account: the Si/Al 1.413 1.786 1.360
semi-quantitative mineralogical composition of the % 88.43 65.74 100
samples; the molar fraction of oxygen in each of the D: tetrahedral charge; w: octahedral charge; =: interlayer charge;
quantified minerals; the structural formula of sericite; %: of tetrahedral charge.
 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293 283

Table 8 values, while the d18O values for albite (An3) showed
Chemical composition and structural formula of kaolinite greater variability. These variations were even more
Samples DAR-2 DAR-2 DAR-2 DAR-2 DAR-2 DAR-2 accentuated in biotite – chlorite mixtures due to the
Oxides 42 43 44 45 46 47 nature of the samples. However, two values for almost
(%) pure biotite and one value for almost pure chlorite are
SiO2 54.08 52.83 52.22 53.69 54.93 52.98 also recorded.
Al2O3 43.40 46.29 45.40 46.31 44.56 45.79
Fe2O3 1.46 – 1.18 – 0.51 0.83
4.2. Mineralogy and d18O values of the hydro-
MgO – – – – – –
CaO – – 0.67 – – 0.40 thermally altered granite
K2O 1.07 0.89 0.52 – – –
Total The mineralogical composition and d18O values of
the bulk samples from the hydrothermally altered
Si4 + 4.045 4.060 4.000 4.040 4.060 4.013
granite intersected by borehole S-13 are listed in Table
Al3 + – – 0.09 – – –
Total 4.045 4.060 4.090 4.040 4.060 4.013 3. These data show that the main phyllosilicate in the
samples is sericite, while chlorite is also present in
Al3 + 3.82 3.82 3.910 3.980 3.890 3.940 three samples. The d18O values of the bulk samples
Fe3 + 0.08 – 0.070 – 0.030 0.47 ranged between 7.5xand 13.5xV-SMOV. For iso-
Mg2 + – – – – – –
topic purposes, it is noticeable that in the four first
Total 3.900 3.930 3.980 3.980 3.920 3.987
samples (DTG.10554 to DTG.10557) K-feldspar,
K+ 0.10 0.08 0.05 – – – chlorite and calcite were absent. Consequently, these
Ca2 + – – 0.053 – – – samples were used only for determining the theoret-
Mg2 + – – – – – – ical d18O value for pure sericite.
Total 0.10 0.08 0.156 – – 0.064

D 0.180 0.240 0.27 0.16 0.24 0.050

w
0.300
0.210
0.060
0.060
0.240
0.039 Table 9
= 0.120 0.030 0.15 0.10 0 0.013 Mineralogical composition and oxygen isotope composition of soft
Si/Al 1.06 1.06 1.00 1.01 1.04 1.018 fractions from fracture fillings intersected by boreholes S-13, S-14,
D: tetrahedral charge; w: octahedral charge; =: Interlayer charge. 42: S-15, S-16 and S-17
analysed point. Samples Q K-Fd Ab Chl Ill Sm K Pgk d18
OV-SMOW
d18O values of pure kaolinite and palygorskite were RF.13-6 2 – 3 – 61 33 1 – 9.5
calculated from samples enriched in these clay min- RF.13-7 – – – – 82 17 1 – 8.2
erals (DAR and RF series), by using the d18O values RF.13-14 5 6 4 4 11 70 – – 13.2
RF.13-17 – – – – 1 51 48 – 11.7
of their respective mixtures together with those meas- RF.14-10 – – – – 19 66 15 – 12.8
ured and calculated for pure minerals present in the RF.15-17 – – – – 46 54 – – 8.4
mixtures. In these cases, the same method as for RF.15-18 9 – 4 – 87 – – – 6.3
estimating of the isotopic signature for pure sericite RF.15-29 – – – – 79 21 – – 10.3
was also applied. RF.16-54 – – – – 13 30 57 – 10.4
RF.16-55 – 5 – – 9 40 46 – 11.6
RF.16-8 – – – – 73 27 – – 10.0
RF.16-10 8 – – 89 3 – – – 7.9
4. Results RF.16-21 – – – – – 100 – – 16.5
RF.16-23 3 3 3 – 3 88 – – 15.1
4.1. d18O values of pure quartz, albite, muscovite and RF.16-40 – – – – 10 90 – – 16.8
RF.16-45 – – – 68 27 5 – – 9.4
chlorite –biotite mixtures from the reference granite RF.17-12 1 – – – 7 21 – 71 15.1
RF.17-15 5 5 3 – 8 79 – – 15.4
The isotopic signatures (d18O values) of these RF.17-27 1 5 3 – 11 78 2 – 14.4
minerals (Table 2) were measured for correction Q: quartz; K-Fd: K-feldspar; Ab: albite; Chl: chlorite; Ill: illite; Sm:
purposes. Both quartz and muscovite gave constant smectite; K: kaolinite; Pgk: palygorskite; d18OV-SMOW: x.
284 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293
 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293 285

4.3. Mineralogy and d18O values of the weathered fibrous habit, but coated by Fe oxyhydroxides (Fig.
hydrothermally altered granite, quartz from the UQV 4e). The d18O value of pure smectite sample was
and its clayey walls 16.5xV-SMOW, while the d18O values of the re-
maining mixtures varied between 6.3xand 16.8x
The mineralogical composition of the bulk sam- V-SMOW.
ples from these materials (Table 4) shows that phyl-
losilicates are generally the main components in the
samples, except in those from the UQV. The minera- 5. Estimation of isotopic signatures for pure
logical composition of their corresponding < 2 Am minerals and geothermometry
fractions (Table 5) shows that the main clay minerals
in the samples, except for the quartz ones, are illite 5.1. Isotopic signature of water in equilibrium with
and smectite, the former predominating over the hydrothermal quartz
latter. Only three samples (LAR-2, LAR-3 and
DAR-2) contain minor kaolinite. The d18O values As mentioned in Section 3, primary fluid inclu-
for hydrothermal pure quartz ranged between 12.6x sions in quartz from the UQV were monophasic, and
and 15.3xV-SMOV, while for clay mixtures the temperatures lower than 100 jC can therefore be
range was between 8.6xand 14.7xV-SMOV (see inferred. On the basis of the highest d18O values of
Table 5). these samples (15.3x) and a temperature of approx-
The structural formulae of the main clay minerals imately 100 jC (or lower), the isotopic signature of
in the samples (Tables 6 –8) correspond to dioctahe- water in equilibrium with quartz could be more
dral clay micas (sericite/hydromuscovite and illite), negative than
5.7x. Similarly, given that the range
aluminian dioctahedral smectite, specifically beidel- of isotope values for quartz was between 12.6x
lite, and kaolinite, respectively. and 15.3x, the d18O range for water in equilibrium
was between
8.3xand
5.7x. However, though
4.4. Mineralogy and d18O values of fracture fillings this range can be considered quite likely, d18O va-
intersected by boreholes drilled at the site lues more negative than
8.3xare possible if the
temperature of formation of quartz was lower than
Mineralogy and oxygen isotopic signatures were 100 jC.
determined only in the < 2 Am fractions of these
materials. The results (Table 9) indicate that, in 5.2. Theoretical d18O values for sericite from the
general, the main clay minerals in the samples are hydrothermally altered granite
illite and smectite, while kaolinite and chlorite are
occasionally present. Only one sample (RF.17-12) Three of the ten samples analysed (DTG.10555 to
contains a significant amount of palygorskite and DTG.10557, see Table 3) were selected to estimate
another (RF.16-21) consists exclusively of smectite. the theoretical d18O value of pure sericite, since they
Chlorite is inherited from the granite; illite comes were ternary mixtures of mainly sericite, magmatic
from the alteration of albite and, to a less extent, of quartz and minor albite, with similar d18OT values.
K-feldspar and muscovite (Fig. 4a); and smectite, Sample DTG.10554, with a similar mineralogical
with a web-like pore-lining morphology (Fig. 4b), composition, had an extremely high d18OT value
mainly from the alteration of albite (Fig. 4c). Kao- and was therefore rejected. For calculation, the
linite is idiomorphic, with a face-to-face stack-type fraction molar method was used, taking into account
texture (Fig. 4d), and palygorskite shows its typical the structural formula of silica, albite (An3) and

Fig. 4. Secondary electron images showing: (a) the transformation of muscovite (1) into illite (2); (b) the web-like pore-lining texture of
smectite; (c) smectite from the alteration of albite. The chemical composition in points 1, 2, 3, 4, 5 and 6 corroborates the gradual albite –
smectite transformation; (d) the first generation of kaolinite (1), with its typical face-to-face stacked texture; (e) fibrous palygorskite coated by
Fe oxyhydroxides with some U and P adsorbed.
286 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293

sericite, and applying the following general equa- 5.3. Theoretical d18O values for smectite, illite,
tion: kaolinite and palygorskite from the clayey walls of
UQV, weathered hydrothermally altered granite and
fracture fillings
Xa d18 Oa þ Xb d18 Ob þ Xc d18 Oc þ . . . Xn d18 On

¼ d18 OT ð5Þ The wide range of isotopic values found in these

samples is attributable to the fact that all the < 2 Am
Abbreviated as: fractions were mixtures of different minerals that in
turn were impossible to separate physically (see
(Tables 4, 5 and 9)). This circumstance enabled the
X
c
Xn d18 On ¼ d18 OT ðn ¼ a . . . cÞ calculation of only an approximate d18O value for
n¼a each pure clay mineral. However, sample RF.16-21
consists exclusively of smectite with an d18O value of
16.5x, which agrees with the theoretical d18O value
where: a, b, c,. . .n = minerals in the sample or for a smectite in equilibrium with the current meteoric
mixture; d18On = isotopic signature of each pure waters at El Berrocal (d18O =
7.6x), at a near-
mineral in the sample or mixture, known from surface temperature.
pure minerals experimentally measured; Xn = num- For estimating the theoretical d18O value of illite
ber of mols of oxygen in each mineral, referred as from the clay mixtures, an initial approach was used
a fraction of the total mols of oxygen in the following a method similar to that of Whitney and
sample or mixture. Northrop (1988). Thus, from the correlation between
X
c the illite/smectite ratios and the experimental d18O
Xn ¼ 1 values determined from samples RF.13-7, RF.15-17,
n¼a RF.15-29, RF.16-8, RF.16-21 and RF.16-40, which
were binary mixture of illite and smectite (see Table
d18OT = d18O experimentally measured in the total 9), a theoretical d18O value ( + 6.7x) for illite was
sample or mixture. estimated (Fig. 6). A second approach was used to
For the particular case of the hydrothermally calculate the d18O value of illite, using the oxygen
altered granite, a, b and c are sericite, magmatic quartz fraction molar method (Eq. (5)) on samples formed
and albite (An3), respectively. For quartz and albite, mainly of smectite and illite, with minor quartz or
the average d18O values were used (Table 2). The quartz and albite (samples LAR-1; DAR-1, RF.13-6
mineralogy and d18OT of the samples used are shown and RF.15-18). For calculation, the d18O values used
in Table 3. were 16.5xfor smectite and 11.5xand 9.5xas
For sericite, the structural formula used was (Si6.387 average values for magmatic quartz and albite, respec-
Al1.613)(Al3.339Fe0.555)K1.927O20(OH)4, which repre- tively (see Table 2). With this second approach and
sents the average formula of the first two formulae data, the d18O value for illite ranged between 4.6x
(Eqs. (1) and (2)) listed in Table 6 since they are the and 8.0x.
most representative for this phyllosilicate. Thus, the The temperature of illite formation is difficult to
theoretical d18O value for pure sericite ranges between estimate since no data are available on the isotopic
5.2xand 7.8xV-SMOV. signature of water in equilibrium with illite. However,
For calculating an approximate temperature of from the structural relationships among UQV, its
formation of sericite, the range of the d18O values clayey walls and the hydrothermally altered granite,
(
8.3xand
5.7x) for waters in equilibrium with it can also be assumed that these materials are formed
hydrothermal quartz from the UQV was used since it as a result of the same hydrothermal process. Con-
was assumed that both UQV and hydrothermally sequently, the isotopic range for water in equilibrium
altered granite resulted from the same geological with quartz from the UQV (
8.3xand
5.7x)
process. According to these parameters, the minimum and the equation for illite –water fractionation (Eq. (2))
temperature varies from 70 to 117 jC (Fig. 5a). can be used to estimate the illite formation temper-
 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293 287

Fig. 5. (a) Plot of the Eslinger and Savin (1973) equation, from which the temperature range for sericite from the hydrothermally altered granite
has been estimated. (b) Idem for illite from fracture fillings and clayey walls of the UQV. (c) Plot of the Land and Dutton (1978) equation on
which the temperature of formation for the second generation of kaolinite has been calculated. (d) Idem for the first generation of kaolinite. (e)
Plot of the Savin and Lee (1988) equation for smectite – water fractionation, applied to palygorskite. (f) Plot of the Gauthier-Lafaye et al. (1993)
equation for palygorskite – water fractionation, also applied to the El Berrocal palygorskite.
288 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293

Fig. 6. Correlation between the illite/smectite ratios and the experimental d18O values obtained from the illite/smectite samples from which a
theoretical d18O value for pure illite has been estimated.

ature, this is ranging from 69 to approximately 124 jC kaolinite and hydrothermal quartz, has been consid-
(Fig. 5b). ered for shallow kaolinite, while samples RF.13-17,
Prior to the estimation of the isotopic signature and RF.14-10 and RF.16-54, which are mixtures of illite,
thermometry of the remaining clay minerals in the smectite and kaolinite, have been retained for deeper
samples, kaolinite and palygorskite, it is appropriate kaolinites.
to take into account the following assumption. Weath- From sample DAR-2, the d18O value estimated for
ering under acidic and intensive cation leaching con- pure shallow kaolinite is 19.1xV-SMOW by apply-
ditions or low-temperature hydrothermal processes ing Eq. (5) and using the d18O values for smectite
(Raymahashay, 1968) are usually the main geological (16.5x), the average value for hydrothermal quartz
mechanisms that originate kaolinite, if the country (13.9x) and the highest values for illite (8.0x). In
rocks are suitable. On the contrary, palygorskite is the same way, for pure deeper kaolinites the d18O
usually formed in very restricted geological environ- values vary between 2.6xand 8.3xV-SMOW. The
ments, under alkaline conditions, with high activities difference between the isotopic signatures of the shal-
of silica and magnesium. low and deeper kaolinites seems to corroborate the
In the El Berrocal system, kaolinite occurs either in different geological environments in which they were
the shallow and oxidised zone of the UQV-clayey formed, as mentioned above. Consequently, for esti-
walls, where pyrite is almost totally oxidised and mating the formation temperature of shallow kaolinite,
sulphate acid percolating waters (pH 3– 4) are present it seems reasonable to assume that it was formed in
today, or in deeper fractures, where reducing, neutral equilibrium with the present-day meteoric waters, with
or weakly alkaline conditions prevail. a weighted average d18O value of
7.6xV-SMOW.
On these bases, kaolinites from these two different Thus, the maximum formation temperature estimated
environments were considered separately to determine for shallow kaolinite was 26 jC (Fig. 5c).
their respective d18O values and formation temper- On the other hand, as kaolinite can also be formed
atures. Sample DAR-2, formed by smectite, illite, in hydrothermal systems under conditions similar to
 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293 289

those that form illite (Raymahashay, 1968), deeper signatures (d18O and d13C) of carbonates, suggested
kaolinites can also be equilibrated with waters in that the hydrothermal water at the El Berrocal site had
equilibrium with both hydrothermal quartz and illite, a meteoric origin, being reheated in depth and moving
provided that the three minerals resulted from the upwards by convective flow.
same hydrothermal process. Thus, the temperature of The d18O range estimated for sericite from the
formation of deeper kaolinites ranged from 88 to hydrothermally altered granite and the intensity of
133 jC (Fig. 5d). A similar temperature range re- the sericitisation process suggests that this mineral
sulted when present-day meteoric waters were used was formed at around 100 jC, with a high water/
for this purpose since both hydrothermal and mete- rock ratio, as corresponds to a highly fractured
oric waters at the site have very similar isotopic geological environment. Illite from fracture fillings
signatures. and clayey walls of the UQV was probably the result
The presence of different generations of kaolinites of the same hydrothermal process, since its forma-
(specifically dickites), including one of Quaternary tion temperature is also around 100 jC. Both pro-
age (dickite from sample DAR-2), has also been cesses—sericitisation of reference granite and illiti-
confirmed in the El Berrocal system, by using Fourier sation of fault gouges—were probably controlled by
transform infrared spectroscopy (FTIR) (Allard et al., the following two well-known general chemical
submitted for publication). reactions:
Regarding palygorskite, sample RF.17-12, formed
essentially by this mineral and smectite together with
minor illite and quartz, was considered only for K
feldspar þ bicarbonate
carbonate H2 O
isotopic and geothermometric purposes. For pure Zsericite or illite þ SiO2 þ Kþ þ Carbonates ð6Þ
palygorskite the d18O value varied from 15.5xto
15.8xV-SMOW, and the estimated temperature
ranged between 31 and 45 jC, when the assumption Albite þ bicarbonate
carbonate H2 O þ Kþ
proposed by Torres-Ruiz et al. (1994) was applied Zsericite or illite þ SiO2 þ Naþ þ Carbonates ð7Þ
(Fig. 5e). However, this temperature range was
approximately 10 jC lower (19 –32 jC) when the
equation proposed by Gauthier-Lafaye et al. (1993) In the hydrothermally altered granite, these reac-
was used (Fig. 5f). In any case, both temperature tions are supported by the presence of K-feldspar and
ranges agree with those estimated for low-temperature albite pseudomorphised by a mixture of sericite,
carbonates from the hydrothermally altered granite carbonates and microcrystalline quartz. Furthermore,
(52 – 35 jC) and for low-temperature carbonates from potassium released from reaction (6) was used in Eq.
fracture fillings, some of them formed at ambient (7) and sodium released from Eq. (7) probably
temperature ( V 25 jC) (Reyes et al., 1998). Finally, migrated outside the system. Neoformed sodium min-
palygorskite as well as carbonates need similar phys- erals have not been found in the samples. Similarly,
icochemical conditions for precipitation. silica released from both reactions was probably spent
in the formation of quartz veins, particularly the UQV.
This quartz was also formed at approximately 100 jC.
6. Discussion and conclusions The presence of kaolinite formed at the same
temperature as illite suggests that both clay minerals
6.1. Argillization processes were also formed during the same alteration pro-
cesses, but only in those fractures where the phys-
The approximate isotopic signature (d18O) esti- icochemical conditions were suitable for its formation.
mated for the hydrothermal waters at the El Berrocal In this sense, the pH – log [H4SiO4] stability diagram
site is very similar to that for the present-day meteoric (Raymahashay, 1968) shows that the boundary
waters, despite boiling and water – rock interactions between the muscovite and kaolinite stability fields
that could affect the hydrothermal ones. In this sense, at 95 jC is located at pH = 8 for a log [H4SiO4] range
Reyes et al. (1998), on the basis of the isotopic between
4 and
3 (Fig. 7). Consequently, illite
290 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293

(see Fig. 7) and with the presence of carbonates formed

at environmental temperature, which also have a sub-
stantial influence of edaphic carbon (Reyes et al.,
1998).
Kaolinite formed at approximately 25 jC can be
explained as having formed under the acid conditions
caused by the oxidation of sulphides from the UQV.
Under these physicochemical conditions, mainly
smectite is transformed into kaolinite, according to
the following general reaction:

Smectite þ acid H2 O
ZKaolinite þ Ca2þ þ Mg2þ þ Kþ þ Quartz ð10Þ
Fig. 7. Stability fields of gibbsite, kaolinite, muscovite, K-feldspar
and montmorillonite in alkaline environment after Raymahashay
The fact that only one sample from the oxidised
(1968).
zone of the clayey walls of the sulphide-bearing UQV
contains minor kaolinite (9%) is probably due to the
and this first generation of kaolinite are assumed to be
extremely low permeability of these clayey walls that
paragenetic in those fractures where the pH ap-
hindered the water – rock interactions between the
proached 8. Above this pH value, only illite would
sulphate acid percolating waters and smectite.
be formed.
For palygorskite, an occasional mineral in the
These parameters agree with the presence in the
fracture filling materials, only one datum concerns
hydrothermally altered granite and fracture fillings of
the temperature of its formation (31 – 45 or 19 –32 jC)
carbonates formed below 100 jC and with some
is available. This mineral was therefore probably
influence of edaphic carbon (Reyes et al., 1998),
formed either during the thermal tail of the hydro-
thereby corroborating the meteoric origin of the
thermal process described above and simultaneously
hydrothermal waters circulating in the system.
with the lower temperature carbonates or during
According to the textures observed by SEM, smec-
weathering processes. In both cases, palygorskite
tite comes mainly from the alteration of albite, prob-
must have formed from alkaline Si– Mg-rich solu-
ably as a result of the following general reaction:
tions. In any case, this fibrous clay mineral appears to
have formed specifically in those fractures of the
Albite þ bicarbonate
carbonate H2 O system where these physicochemical conditions
ZSiO2 þ Smectite þ Carbonates ð8Þ occurred.

6.2. Implications for the performance assessment

Illite can also be transformed into smectite by the
reaction: Some analogies between El Berrocal system and a
deep geological radwaste repository can be drawn:
Illite þ SiO2 þ bicarbonate
carbonate H2 O
ZSmectite þ Kþ þ Carbonates ð9Þ The UQV contains pitchblende (UO2 + x), which is
analogous to the main component of the spent
nuclear fuel (source term).
Furthermore, geothermometric studies indicate that The geometry of the UQV is very simple and it is
smectite was formed at ambient temperature (e.g. in bordered on two clayey walls analogous to the
relation to weathering processes) under alkaline con- engineered clayey barrier envisioned to isolate the
ditions. This agrees with the stability diagram at 25 jC spent nuclear fuel (near field).
 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293 291

The UQV is hosted in highly evolved S-type reference granite close to fractures, increasing their
granites (store rock or far field), which are well clay contents and therefore their retention capa-
represented in the European Hercynian Belt. bilities for natural nuclides and other analogous
elements (Pérez del Villar et al., 1995, 1997).
On the contrary, the UQV, highly fractured and
very permeable, is perturbed by mining activities and, If a similar process occurs in a nuclear repository,
particularly, contains considerable amounts of sul- two possibilities can be considered. The first would be
phides and carbonates, which in principle are two that the oxidising meteoric waters percolate down to
‘‘poisonous’’ minerals under oxidising conditions. the spent fuel, and the second that the meteoric waters
Their dissolution also facilitates the destabilisation lose their oxidising capability (by rock – water inter-
of pitchblende and encourages the migration of ura- action processes) before interaction with the spent
nium. fuel. In the first case, the source term would be
Finally, the granite country rock of the El Berrocal destabilised, and U and other nuclides would be
UQV contains uraninite and is therefore the second dispersed as a function of their solubility in these
source of U in the system. This feature hampers the reheated waters. However, clays from an engineered
delimitation of the dispersion haloes of U from the barrier, fracture fillings and country rock itself would
UQV and from the granite itself. In addition, the El work against the migration of radionuclides as in the
Berrocal granite shows a very high fracture density, El Berrocal natural system (Pérez del Villar et al.,
though fractures are generally filled by clay minerals, 1997). In the second case, the source term would
minor Fe oxyhydroxides and other minerals, mainly remain totally or partially unaltered. Only coffinitisa-
phosphates, that participate in the retardation of U tion of the UO2 + x from the spent fuel could be
(VI). expected (Pérez del Villar et al., 2002). Furthermore,
According to these geological and mineralogical the hydrothermal system artificially created in the
features, the El Berrocal system cannot be considered nuclear repository would alter the granitic storage
a completely natural analogue of a deep geological rock, mainly through the fracture zones, increasing
radwaste repository. However, in terms of the pro- its clay content and therefore its retention capability.
cesses affecting the system, both low-temperature In addition, past and recent past weathering pro-
hydrothermal and weathering processes can be com- cesses were more intensive and deeper on the UQV
pared to those expected in a nuclear repository after than in the hydrothermally altered and reference
burial, under extremely adverse circumstances. granites due to their different secondary permeability.
The low-temperature hydrothermal process could The pH of these waters in the upper part of the UQV is
be similar to that expected in a radwaste repository usually around 3 – 4, but close to 7 in the other
after burial, since a thermal peak related to radioactive hydraulically active fractures. In both cases, waters
decay of fission products from the spent fuel is become weakly alkaline in depth (Pérez del Villar et
envisaged (less than 100 jC in the ENRESA con- al., 1995). Accordingly, the following analogies can
ceptual model). be established:
In the natural system under investigation, the low-
temperature hydrothermal process caused: (a) In the UQV itself or source term, oxidation due
to weathering caused the total oxidation and
The accumulation in the UQV of U scattered in the dissolution of pitchblende, sulphides (mainly
granite, as a result of a first downward step of U pyrite), and carbonates, as well as the precip-
mobilisation, under oxidising and weakly acidic or itation of jarosite (pH 3 – 4) and Fe (Mn)
alkaline conditions, and environmental tempera- oxyhydroxides with minor U and P adsorbed
ture; and a second upward step of U (IV) preci- (Pérez del Villar et al., 1995, 1997). Conse-
pitation, under weakly alkaline and reducing con- quently, this intense downward oxic and acid
ditions at higher temperatures, but below 100 jC. alteration in the upper part of the UQV is of no
The argillization of the fault gouges, including the relevance in assessing repository behaviour. How-
UQV walls, and the hydrothermal alteration of the ever, once the conditions of neutrality – alkalinity
292 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293

and reducing potential are restored in depth, the References

original mineral paragenesis [pitchblende, sul-
phides (mainly pyrite) and carbonates] becomes Allard, Th., Ildefonse, Ph., Pérez del Villar, L., Sorieul, S., Pelayo,
stable, illustrating again the well-known stability M., Boizot, B., Balan, E., Calas, G. Radiation-induced defects in
dickites from the El Berrocal natural analogue granitic system.
of natural UO2 + x under reducing conditions. In the
Eur. J. Mineral. (submitted for publication).
transition between oxidised and reduced zones, in Arribas, A., 1965. Mineralogı́a y metalogenia de los yacimientos
situ transformation of pitchblende into uranocircite españoles de uranio. ‘‘El Berrocal’’, Escalona (Toledo). Notas
and autunite together with the precipitation of Comun. Inst. Geol. Min. España 77, 67 – 92.
torbernite occur, this zone being therefore relevant Borthwick, J., Harmon, R., 1982. A note regarding ClF3 as an
alternative to BrF5 for oxygen isotope analysis. Geochim. Cos-
as a geochemical discontinuity for natural nuclide
mochim. Acta 46, 1665 – 1668.
retardation. Bradley, W.F., Grim, R.E., 1961. Mica clay minerals. In: Brown, G.
(b) In the clayey walls or near field, with lower per- (Ed.), The X-ray Identification and Crystal Structures of Clay
meability than UQV itself, inherited albite and Minerals. Mineralogical Society, London, pp. 208 – 241.
particularly illite were transformed firstly into Chapman, N.A., 1994. The geologist’s dilemma—predicting the
future behaviour of buried radioactive wastes. Terra Nova 6,
smectite under weakly alkaline conditions increas-
5 – 19.
ing therefore the buffer capability of this natural Chapman, N.A., Smellie, J.A.T., 1986. Introduction and summary
clayey barrier. After this transformation, the of the workshop. In: Chapman, N.A., Smellie, J.A.T. (Eds.),
system became acid by the oxidation of sulphides Natural Analogues to the Conditions Around a Final Reposi-
and minor smectite was occasionally transformed tory for High-level Radioactive Waste. Chem. Geol., vol. 55,
pp. 167 – 173.
into kaolinite, a clay mineral without any adequate
Choppin, G.R., Wong, P.J., 1996. Current status of radioactive-
buffer and physicomechanical properties to work waste disposal. J. Radioanal. Nucl. Chem. 203, 575 – 590.
as barrier against migration of U and other trace Clayton, R.N., O’Neil, J.R., Mayeda, T.K., 1972. Oxygen isotope
analogue elements. Consequently, smectite – kao- exchange between quartz and water. J. Geophys. Res. 77,
linite transformation must be avoided regardless 3057 – 3067.
Eslinger, E.V., Savin, S.M., 1973. Mineralogy and oxygen isotope
of the geological site selected for radwaste
geochemistry of the hydrothermally altered rocks of the Ohaki-
disposal. broadlands, New Zealand Geothermal Area. Am. J. Sci. 273,
(c) In the shallow hydrothermally altered granite and 240 – 270.
fracture fillings (far field), weathering also trans- Faure, G., Powell, J.L., 1972. Strontium Isotope Geology. Minerals,
formed illite and residual albite into smectite, Rocks and Inorganic Materials. Springer-Verlag, Berlin.
Gauthier-Lafaye, F., Taieb, R., Paquet, H., Chahi, A., Prudencio, I.,
generally increasing their retention and buffer
1993. Composition isotopique de l’oxygène de palygorskites
capabilities against the mobility of natural radio- associées à des calcrètes: conditions de formation. C. R. Acad.
elements. Uranium is usually trapped by the Sci. Paris T. 316 (Serie II), 1239 – 1245.
clayey minerals in the rock, Fe oxyhydroxides Hidaka, H., Holliger, P., 1998. Geochemical and neutronic charac-
and sometimes precipitated as autunite, as a result teristics of the natural fossil fission reactors at Oklo and Ban-
gombé, Gabon. Geochim. Cosmochim. Acta 62, 89 – 108.
of the presence of significant amounts of phos-
Julivert, M., Fontboté, J.M., Ribeiro, D., Conde, L., 1972. Mapa
phorous in the system. This is a relevant anion- tectónico de la Penı́nsula Ibérica y Baleares. E. 1:106. Inst. Geol.
forming element for the retention, by precipita- Min. España, Madrid.
tion, of the complex uranyl cation and other Krivovichev, V.G., 1979. The paragenesis of minerals and analysis
analogous elements. of mineral equilibrium in barite and barite polymetallic ore de-
posits. In: Rundkvist, D.V. (Ed.), Minerals and Mineral Para-
genesis in Rocks and Ores. Leningrad, IzdatelVstvo Nauka,
Leningradskoe Otdelenie, pp. 45 – 60.
Acknowledgements Land, L.S., Dutton, S.P., 1978. Cementation of a Pennsylvanian del-
taic sandstone: isotopic data. J. Sediment. Petrol. 48, 1167 – 1176.
ENRESA, CIEMAT and EU provided the financial Lawrence, J.R., Taylor, H.P., 1971. Deuterium and oxygen-18
correlation: clay minerals and hydroxides in Quaternary soils
supports for this work. Dr. N. Clauer and another
compared to meteoric waters. Geochim. Cosmochim. Acta 35,
anonymous reviewer are also thanked for their useful 903 – 1003.
criticisms, comments and suggestions that improved Miller, W., Alexander, R., Chapman, N., McKinley, I., Smellie, J.,
the manuscript. [PD] 2000. Geological Disposal of Radioactive Wastes and Natural
 L. Pérez del Villar et al. / Chemical Geology 193 (2003) 273–293 293

Analogues. Waste Management Series, vol. 2. Pergamon, Am- fillings from the El Berrocal granitic site (Spain). J. Contam.
sterdam, 316 pp. Hydrol. 26, 45 – 60.
Parneix, J.C., 1992. Effects of hydrothermal alteration on radioele- Pérez del Villar, L., Bruno, J., Campos, R., Gómez, P., Cózar, J.S.,
ment migration from a hypothetical disposal site for High-Level Garralón, A., Buil, B., Arcos, D., Carretero, G., Ruiz Sánchez-
Radioactive-Waste—example from the Auriat granite, France. Porro, J., Hernán, P., 2002. The uranium ore from mina Fe
Appl. Geochem. 7, 253 – 268. (Salamanca, Spain) as a natural analogue of processes in a spent
Pérez del Villar, L., de la Cruz, B., Pardillo, J., Cozar, J.S., 1992. fuel repository. Chem. Geol. 190, 395 – 415.
Cálculo de la distribución mineralógica del uranio y torio en el Raymahashay, B., 1968. A geochemical study of rock alteration by
granito sano del plutón de El Berrocal (Sierra de Gredos, Pro- host springs in the Paint Pot Hill area, Yellowstone Park. Geo-
vincia de Toledo). Bol. Soc. Esp. Mineral. 15, 123 – 136. chim. Cosmochim. Acta 32, 499 – 522.
Pérez del Villar, L., de la Cruz, B., Pardillo, J., Cózar, J.S., Ménager, Reyes, E., Pérez del Villar, L., Delgado, A., Cortecci, G., Núñez,
M.T., 1994. Uranium distribution in relation to the mineralogical R., Pelayo, M., Cózar, J.S., 1998. Carbonation processes at the
and geochemical characteristics of the granite from the El Ber- El Berrocal natural analogue granitic system (Spain): inferences
rocal uranium mine (Spain). Proc. of the Fourth International from mineralogical and stable isotope studies. Chem. Geol. 150,
Conference on the Chemistry and Migration Behaviour of Acti- 93 – 315.
nides and Fission Products in the Geosphere (Charleston, USA). Rivas, P., Hernán, P., Bruno, J., Carrera, J., Gomez, P., Guimerá, J.,
Oldenbourg Verlag, München, pp. 363 – 375. Marı́n, C., Pérez del Villar, L., 1997. El Berrocal Project. Char-
Pérez del Villar, L., Cózar, J.S., de la Cruz, B., Pardillo, J., Fernán- acterisation and validation of natural radionuclide migration
dez, A.M., 1995. Influence of the alteration processes on the processes under real conditions on the fissured granitic environ-
U – Th – REE minerals in the El Berrocal site/U-bearing quartz ment. E.C. Nuclear Science and Technology. EUR 17478 EN,
vein system (Spain): an example of natural analogue processes. 522 pp.
In: Pasava, J., Kribek, B., Zak, K. (Eds.), Mineral Deposits: Ruiz-Cruz, M.D., Reyes, E., 1998. Kaolinite and dickite formation
From Their Origin to Their Environmental Impacts. AA Balke- during shale diagenesis: isotopic data. Appl. Geochem. 13,
ma, Rotterdam, pp. 689 – 692. 95 – 104.
Pérez del Villar, L., Crespo, M.T., Pardillo, J., Pelayo, M., Galán, Savin, S.M., Lee, M., 1988. Isotopic studies of phyllosilicates. Rev.
M.P., 1996a. U and Th series disequilibrium in the unaltered and Miner. 19, 189 – 223.
hydrothermally altered granites from the El Berrocal site. Appl. Smellie, J.A.T., Karlsson, F., Alexander, W.R., 1997. Natural ana-
Radiat. Isotopes 47, 1115 – 1119. logue studies: present status and performance assessment impli-
Pérez del Villar, L., Fernández-Dı́az, M., Quejido, A.J., Pardillo, J., cations. J. Contam. Hydrol. 26, 3 – 17.
1996b. Dating of the El Berrocal pluton and the Almorox-Nav- Torres-Ruiz, J., López-Galindo, A., González-López, M., Delgado,
amorcuende dyke by the rubidium – strontium method. El Ber- A., 1994. Geochemistry of Spanish sepiolite – palygorskite de-
rocal Project: Characterization and Validation of Natural Ra- posits: genetic considerations based on trace elements and iso-
dionuclide Migration Processes Under Real Conditions on the topes. Chem. Geol. 112, 221 – 247.
Fissured Granitic Environment. Geological Studies, vol. 1. Vennemann, T.W., Smith, H., 1990. The rate and temperature of
ENRESA, Madrid, Spain, pp. 235 – 253. reaction of ClF3 with silicate minerals, and their relevance to
Pérez del Villar, L., Pelayo, M., Cózar, J.S., de la Cruz, B., Pardillo, oxygen isotope analysis. Chem. Geol. 86, 83 – 88.
J., Reyes, E., Caballero, E., Delgado, A., Nuñez, R., Ivanovich, Whitney, G., Northrop, H.R., 1988. Experimental investigation of
M., Hasler, S.E., 1997. Mineralogical and geochemical evidence the smectite to illite reaction mechanisms and oxygen isotope
of the migration/retention processes of U and Th in fracture systematic. Am. Mineral. 73, 77 – 90.