GeolCarp_Vol60_No6_535

www.geologicacarpathica.sk

GEOLOGICA CARPATHICA, DECEMBER 2009, 60, 6, 535—543 doi: 10.2478/v10096-009-0039-7

Introduction

It  is  envisaged  that  high-level  nuclear  waste  (HLW)  will  be
disposed  of  in  underground  repositories.  Several  current  de-
signs  for  the  future  geological  disposal  of  HLW  suggest  the
use of a multibarrier system with two basic components – a
host  rock  and  an  engineered  barrier.  The  engineered  barrier
comprises the respective metallic containers filled with radio-
active waste and barrier made up of bentonite blocks. The me-
tallic  containers  could  be  made  up  of  iron  or  copper  (JNC
2000;  Push  2001,  2003;  Thorsager  &  Lindgren  2004;  SKI
2005; Arthur et al. 2005). To predict the long-term properties
of the components in the designed barriers of HLW repository
it is essential to study their interactions. This paper focuses on
the interactions between bentonitic clays and iron.

This rather complex topic has been the subject of several re-

search  projects  in  the  last  years.  Börgesson  et  al.  (2002)  as-
sume that the container will corrode anaerobically and that the
principal  corrosion  product  will  be  magnetite  (Smart  et  al.
2002).  Nevertheless,  oxidizing  conditions  are  likely  to  occur
before  closure  of  the  repository  and  other  iron  compounds
than magnetite may also be formed at this stage.

Several experimental studies of iron-clay interactions

showed the systematic destabilization of the initial clay mate-
rial and the subsequent crystallization of reaction products
(Guillaume et al. 2003, 2004; Lantenois et al. 2005; Wilson et

al. 2006; Perronnet et al. 2007). The nature of these reaction
products depends on experimental conditions such as temper-
ature and the nature of the initial clay material. Fe-rich chlorite
species  are  newly-formed  after  interaction  of  smectite  with
iron  at  higher  temperature  (300 °C)  (Guillaume  et  al.  2003),
whereas  Fe-rich  serpentine-like  species  are  synthesized  at
lower  temperatures  (80 °C)  (Haber  2000;  Lantenois  2003;
Perronnet 2004). Lantenois et al. (2005) observed that diocta-
hedral smectites are destabilized while trioctahedral smectites
are  essentially  unaffected  under  similar  experimental  condi-
tions during iron-clay interactions. They also found that smec-
tite destabilization by the interaction with iron is enhanced by
structural  Fe

in the smectite, by the interlayer Na

cations,

and by the alkaline pH, respectively. However taking into ac-
count the results of Lantenois et al. (2005), Perronnet (2004)
and Perronnet et al. (2007) it is clear that some additional pa-
rameters may also have an impact on the reaction rate of the
iron-smectite interactions. Apparently the smectite layer
charge and the textural and energetic surface heterogeneities
of smectite crystals may play a role.

The clay-water ratio should be taken into considerations

when  the  experimental  results  are  interpreted.  For  instance
Müller-Vonmoos  et  al.  (1991)  reviewed  by  Madsen  (1998)
found no significant changes in the bentonite samples after in-
teraction with iron and magnetite at 80 °C for 29 weeks, when
no  additional  water  was  present  in  the  reaction  system.  The

Experimental interactions of Slovak bentonites with

metallic iron

MAREK OSACKÝ

, MIROSLAV HONTY

, JANA MADEJOVÁ

, THOMAS BAKAS

and VLADIMÍR ŠUCHA

Department of Geology of Mineral Deposits, Comenius University, Mlynská dolina G, 842 15 Bratislava, Slovak Republic;

osacky@fns.uniba.sk; honty@fns.uniba.sk; sucha@fns.uniba.sk

Institute of Inorganic Chemistry, Slovak Academy of Sciences, Dúbravská cesta 9, 845 36 Bratislava, Slovak Republic;

jana.madejova@savba.sk

Department of Physics, University of Ioannina, 45110 Ioannina, Greece; tbakas@cc.uoi.gr

(Manuscript received December 18, 2008; accepted in revised form June 25, 2009)

Abstract: The experimental stability of four bentonites and one K-bentonite from Slovak deposits in the presence of
iron was studied to simulate the possible reactions of clays (bentonite barrier) in the contact with Fe containers in a
nuclear waste repository. The batch experiments were performed at 60 °C for 30 and 120 days in aerobic conditions.
The reaction products were examined by XRD, FTIR, and Mössbauer spectroscopies and CEC (cation exchange capaci-
ties) were determined. Reaction solutions were analysed for selected elements using AAS (atomic absorption spectro-
metry). The results show that bentonites do not interact equally with metallic iron. Bentonites from the Jelšový Potok,
Kopernica and Lieskovec deposits reacted similarly whereas the interaction between the bentonite from Lastovce and
the iron was less intensive. The lower reactivity of the bentonite from Lastovce can be explained by its low content of
smectite. During iron-clay interactions the iron was consumed and Fe oxides (magnetite, lepidocrocite) were formed.
Decrease of the smectite diffraction peaks intensity and CEC values during the experiments show rather the rearrange-
ment of the original smectite crystals than dissolution of smectite. In the K-bentonite from the Dolná Ves deposit where
the mixed-layer illite-smectite is present instead of smectite, the dissolution of illite-smectite was observed along with
the neoformation of smectite. The structure of illite-smectite deteriorated more than the structure of smectites which
suggests that this mixed-layer illite-smectite is much less stable in the presence of iron than smectites.

Key words: nuclear waste repository, clay stability, bentonite, illite-smectite, smectite, magnetite, iron.

536

OSACKÝ, HONTY, MADEJOVÁ, BAKAS and ŠUCHA

aim  of  the  present  work  was  to  investigate  the  stability  of
smectite and illite-smectite in the presence of iron to simulate
the possible reactions of bentonites in contact with a Fe con-
tainer in a nuclear waste repository. Most of the experimental
studies  of  iron-clay  interactions  were  performed  under  oxy-
gen-free  conditions.  There  is  a  lack  of  experimental  data  re-
garding iron-clay interactions in aerobic conditions, however
oxidizing conditions are likely to occur before closure of the
nuclear  waste  repository.  For  these  reasons  our  experiments
were  carried  out  in  the  aerobic  environment.  In  addition  we
bring here the first results of the stability of mixed-layer illite-
smectite in the presence of iron.

Materials, experimental setting and analytical

methods

Four bentonites from Slovak deposits at Jelšový Potok

(J45),  Kopernica  (K45),  Lieskovec  (L45),  Lastovce  (La45)
and one K-bentonite from the Dolná Ves (DV45) deposit were
used for experiments (Fig. 1). All the bentonites are commer-
cially  available  products  technologically  treated  by  drying,
grinding and size separated to  < 45 µm. The iron represents a
99% pure metallic Fe

powder with particle size of 10 µm.

The mixtures of iron with bentonite in redistilled water

(2 g/10 g/106 ml) were prepared in polyethylene (PE) bottles
in fully aerobic conditions. After mixing the starting materials,
the PE bottles were closed and the experiments were conduct-
ed at 60 °C for 30 and 120 days. It means that the treatment of
the samples before and after the experiments was conducted in
fully  aerobic  conditions.  However,  the  experiments  them-
selves  were  conducted  in  semi-aerobic  conditions.  Once  the
plastic containers were closed no additional oxygen was avail-
able. The initial volume of free air in the bottles was approxi-
mately  195 ml,  the  estimated  molar  ratio  of  atmospheric
oxygen to iron was about 0.047.

At the end of the reactions the samples were cooled to room

temperature  and  opened  in  air.  After  centrifugation  the  solu-
tion was filtered (0.45 µm) and pH, Eh and conductivity were
subsequently  measured.  The  filtered  solutions  were  analysed
for selected elements (Si, Al, Ca, Na, Mg, K, Fe) using AAS
(atomic  absorption  spectrometry).  The  solid  fraction  was
dried at 60 °C overnight, ground and analysed. The fine frac-

tion ( < 2 µm) of samples before and after experiments was
separated by sedimentation and used for analyses as well.

X-ray diffraction (XRD) analyses of oriented (air-dried and

ethylene glycolated) and random specimens were carried out
using a Phillips PW 1710 diffractometer (35 kV, 20 mA) with
CuK

α radiation and a graphite monochromator. All samples

were scanned with a step 0.02° 2 theta and count time 0.80 s
per step over a measuring range of 2 to 50° 2 theta.

Quantitative XRD analysis was performed by RockJock

modelling. RockJock is a computer program (Eberl 2003) that
determines quantitative mineralogy in powdered samples by
comparing the integrated X-ray diffraction intensities of indi-
vidual minerals in complex mixtures to the intensities of an in-
ternal standard (ZnO). Samples (3.000 g) were mixed with
internal standard ZnO (0.333 g) and wet ground in a McCrone
Micronizing Mill for 5 minutes (Środoń et al. 2001). Random
specimens were scanned with a step of 0.02° 2 theta and count
time 2 s per step over measuring range of 4 to 65° 2 theta. The
X-ray data were entered into the RockJock computer program
and the mineral compositions of the samples were calculated.

The quantitative XRD analysis (RockJock) of samples after

the experiment was not performed because not all the minerals
present in bentonites are listed in the RockJock mineral data-
base. Only minerals listed in the RockJock mineral database
can be quantified. The quantitative amount of SiO

forms

(cristobalite, volcanic glass) was determined by RockJock
from XRD patterns as well.

The relative amount of the residual iron after the experi-

ment was estimated from the intensity of diffraction peaks of
iron in XRD patterns. We assumed that more intensive iron
diffraction peaks indicated higher amounts of residual iron
(i.e. lower consumption) and less intensive peaks indicated
lower amounts of the iron (i.e. higher consumption) after the
experiment.

The Fourier transform infrared (FTIR) spectra in the middle

region (4000—400 cm

—1

) were obtained using a Nicolet Magna

750 spectrometer with a DTGS detector and a KBr beam split-
ter. The KBr pressed-disc technique (1 mg of sample and
200 mg of KBr) was used for the transmission measurements.
Discs were heated in a furnace overnight at 150 °C to mini-
mize the water adsorbed on KBr and the clay sample.

Mössbauer spectra were obtained at liquid nitrogen temper-

atures (80 K) using a Wissel spectrometer equipped with an
Oxford Variox 316 cryostat. A

Co source was moved in con-

stant acceleration mode. Velocity calibration was carried out
with respect to the centre of a Fe foil spectrum at room tem-
perature. The spectra were deconvoluted with a least-squares
computer program assuming Lorentzian line-shapes.

The complex of copper(II) triethylentetramine [Cu Trien]

was  used  for  determination  of  cation  exchange  capacity
(CEC)  of  the  samples.  The  0.01 M  solution  of  the  complex
[Cu  Trien]

was prepared according to Meier & Kahr

(1999). 100 mg of samples were added to 50 ml of distilled
water and 10 ml solution of [Cu Trien]

. The suspensions

were dispersed by an ultrasonic treatment for 5 minutes, fil-
trated and the concentration of Cu(II) complex was deter-
mined by UV-VIS spectrophotometry (Cary 100, Varian) at
578 nm (Meier & Kahr 1999). The amount of adsorbed
[Cu Trien]

was determined using molar absorption coeffi-

Fig. 1. Location of: 1 – Jelšový Potok, 2 – Kopernica, 3 – Lies-
kovec, 4 – Lastovce, 5 – Dolná Ves deposits.

537

EXPERIMENTAL INTERACTIONS OF BENTONITES WITH METALLIC IRON (SLOVAKIA)

cient

ε=0.245 mol

—1

· dm

· cm

—1

(Kaufhold & Dohrmann

2003) and the CEC values were calculated.

Element contents were determined by the flame atomic ab-

sorption  spectrometry  (AAS)  method.  Si  and  Al  were  mea-
sured  in  acetylene-nitrous  oxide  flame  (AAS  Perkin  Elmer
Model  5000),  Ca,  Na,  Mg,  K  and  Fe  in  acetylene-air  flame
(AAS  Perkin  Elmer  Model  1100).  Lanthanum  was  added
(c (La) = 1 g/l)  for  the  determination  of  Ca  and  Mg,  Fe  was
measured with deuterium background correction.

Results

X-ray diffraction

All the starting samples were analysed by XRD. Samples

J45, K45, L45 and La45 have similar mineral compositions
while sample DV45 is different. The non-clay minerals quartz,
K-feldspar, cristobalite and volcanic glass (for explanation of
how the volcanic glass was detected see discussion) were
found in samples K45, L45 and La45 (Figs. 2, 3).

The principal layer silicate in bentonites is smectite with

small  amounts  of  kaolinite and biotite (Figs. 2, 3). The pres-
ence of (060) peak at  ~ 0.149 nm indicates dioctahedral smec-
tite. The sample La45 contains the lowest amount of smectite,
the highest amounts of cristobalite, volcanic glass and a small
amount  of  calcite  compared  with  the  samples  J45,  K45  and
L45  (Table 1).  The  mineral  composition  of  sample  DV45  is
dominated by quartz, K-feldspar and volcanic glass. Instead of
smectite,  the  mixed-layer  illite-smectite  is  present  with  non-
expandable  (illite)  layers  forming  about  68 %,  and  traces  of
biotite and kaolinite (Fig. 4). The RockJock quantitative XRD
data  are  listed  in  Table 1.  The  smectite  content  in  bentonites
varies  between  81  and  42 %.  The  illite-smectite  content  of
DV45  is  34 %.  The  residual  iron  (0.202 nm)  was  present  to-
gether  with  newly-formed  magnetite  (0.253 nm)  in  the  XRD

Fig. 2. XRD patterns of random specimens of bentonite K45 (frac-
tions < 45 µm). 1 – starting bentonite, 2 – after 30 days of iron-ben-
tonite interactions, 3 – after 120 days of iron-bentonite interactions.
Sm – smectite, Bt – biotite, Qtz – quartz, Kfs – K-feldspar,
Mgt – magnetite, Fe – iron.

Fig. 3. XRD patterns of the oriented ( < 2 µm, ethylene glycolated)
samples of La45. 1 – starting bentonite, 2 – after 30 days of iron-
bentonite interactions. Sm – smectite, Kln – kaolinite, Crs – cris-
tobalite, Lp – lepidocrocite.

Fig. 4. XRD patterns of the oriented ( < 2 µm, ethylene glycolated)
samples of K-bentonite DV45. 1 – starting K-bentonite, 2 – af-
ter 30 days of iron-bentonite interactions, 3 – after 120 days of
iron-bentonite interactions. I—S – illit-smectite, Kln – kaolinite,
Qtz – quartz, Sm – smectite.

Table 1: Mineralogical compositions (in wt. %) of starting samples
(fractions < 4 5 µm) determined by RockJock computer program.

Minerals Sample

J45

(wt.%)

K45

(wt. %)

L45

(wt. %)

La45

(wt. %)

DV45

(wt. %)

Quartz

Plagioclase

K-feldspar

Biotite

Kaolinite

Smectite

81 77 56 42

Cristobalite

Volcanic glass

Calcite

Illite-smectite

538

OSACKÝ, HONTY, MADEJOVÁ, BAKAS and ŠUCHA

patterns  of  randomly  oriented  specimens  of  samples  J45,
L45, K45 and DV45 after the interactions conducted for 30
and  120  days  (Fig. 2).  The  decrease  of  intensity  of  the  iron
diffraction peak in XRD patterns after 120 days with respect
to  the  30  day  patterns,  indicate  the  higher  consumption  of
iron in experiments with longer duration. The intensity of the
diffraction  peaks  of  quartz  and  feldspars  decreased  signifi-
cantly after the experiments in all samples. The sample La45
reacted with iron after 30 days differently with respect to the
samples  mentioned  above.  The  diffraction  peaks  of  cristo-
balite  and  calcite  decreased  significantly  during  the  experi-
ments but the peaks of iron did not change significantly and
no  magnetite  was  formed.  It  indicates  that  iron  was  much
less  consumed  (oxidized)  during  iron-bentonite  interactions
and it is still present in its original oxidation state (Fe

). This

finding was also confirmed by Mössbauer spectroscopy (Ta-
ble 2). Instead of magnetite a small amount of lepidocrocite
(0.627 nm) was formed (Fig. 3).

The XRD patterns of oriented specimens ( < 2 µm) also

showed that the intensity of the smectites diffraction peaks de-
creased in all samples. Their position did not change during
the experiments (Fig. 3).

K-bentonite from Dolná Ves (sample DV45) reacted in a

similar way to smectitic bentonites (J45, L45 and K45), how-
ever the illite-smectite peaks decreased more significantly
than the smectite peaks of bentonites. Interestingly a peak of
newly formed smectite was observed at low two-theta degrees
of the EG saturated patterns (Fig. 4).

FTIR spectroscopy

The IR spectra of samples J45, K45, La45 and L45 after 30

days of iron interactions at 60 °C showed no differences with
respect to the original IR spectra J45, K45, La45, L45
(Fig. 5c,d). No significant changes were observed in the shape

and position of bands from samples before and after the exper-
iments. No changes in the structural OH groups and Si—O vi-
brations indicate that no structural changes occurred in the
tetrahedral and octahedral sheets of smectites during iron in-
teractions after 30 days in samples J45, K45, La45 and L45.
The potential formation of Fe oxyhydroxides during iron-clay
interactions cannot be detected by IR spectroscopy, because
the absorption bands of these Fe compounds are usually over-
lapped by more intense bands of smectite.

The most evident differences were observed between the IR

spectra of sample DV45 and DV45 after 30 days (Fig. 5a,b).
The intensity of the AlAlOH band at 915 cm

—1

decreased, the

AlMgOH band at 848 cm

—1

disappeared and the intensity of

the quartz doublet at 800 and 779 cm

—1

increased (Fig. 5a,b).

Sample

Bhf (T)

ISFe (mm/s)

QS (mm/s)

Area (%)

Determination

J45

Doublet 1

0.43

0.69

Montmorillonite + Fe-oxides

Sextet

51.5 0.5

–0.1

Hematite

Sextet 2

47.5

0.51

–0.05

Al-hematite

Sextet 3

43.3

0.47

–0.16

Goethite?

Sextet 4

34.4

0.11

–0.02

Metallic iron

L45

Doublet 1

0.47

0.69

Montmorillonite + Fe-oxides

Sextet 1

49.6

0.5

–0.07

Hematite

Sextet 2

46.2

0.47

–0.06

Al-hematite

Sextet 3

42.1

0.45

–0.17

Goethite?

Sextet 4

34.6

0.12

–0.01

Metallic iron

K45

Doublet 1

0.48

0.72

Montmorillonite + Fe-oxides

Sextet 1

51.9

0.51

–0.14

Hematite

Sextet 2

47.6

0.5

–0.05

Al-hematite

Sextet 3

43.3

0.49

–0.16

Goethite?

Sextet 4

34.4

0.11

–0.01

Metallic iron

DV45

Doublet 1

0.46

0.8

Montmorillonite + Fe-oxides

Sextet 1

51.8

0.51

–0.1

Hematite

Sextet 2

0.52

–0.06

Al-hematite

Sextet 3

43.5

0.5

–0.18

Goethite?

Sextet 4

34.4

0.1

–0.01

Metallic iron

La45

Doublet 1

0.54

0.68

Montmorillonite

Sextet 4

34.4

0.11

0.01

Metallic iron

Table 2: Hyperfine parameters (B

– magnetic hyperfine field, IS

– isomer shift, QS – quadrupole splitting) and relative areas (%) of

spectral components from the Mössbauer spectra of samples after 30 days of iron-bentonite interactions, measured at 80 K.

Fig. 5. FTIR spectra of samples (fractions < 2 µm) before and after
iron interactions. a – DV45 after 30 days, b – DV45, c – L45 af-
ter 30 days, d – L45.

539

EXPERIMENTAL INTERACTIONS OF BENTONITES WITH METALLIC IRON (SLOVAKIA)

The spectral changes of the sample DV45 reflected OH group
dehydroxylation and release of the central atoms (Al, Mg)
from the octahedral sheet of illite-smectite.

Mössbauer spectroscopy

The Mössbauer spectra of samples J45, K45, DV45, L45

and La45 after iron interaction at 60 °C for 30 days were tak-
en  at  liquid  nitrogen  temperatures.  The  prominent  central
doublet  in  the  spectra  of  all  samples  is  characteristic  of
Fe(III) in octahedral coordination with no magnetic ordering,
indicating that the Fe is bound primarily in the phyllosilicate
structure (Figs. 6—10). The sextets in the spectra of all sam-
ples  are  attributed  to  the  octahedral  Fe(III)  of  Fe
(oxyhydr)oxides and to the Fe(0) of residual metallic Fe, all
with magnetic ordering (Figs. 6—10).

The Mössbauer spectrum of the sample J45 shows that

41 % of the Fe is bound in paramagnetic phases (Fig. 6, Ta-
ble 2). Based on the hyperfine parameters (isomer shift, qua-
drupole  splitting;  Table 2),  this  doublet  is  assigned  to
phyllosilicate,  montmorillonite  which  is  the  most  abundant
mineral in the sample J45. Because of relatively high amount
of the total Fe (41 %) contributing to the doublet signal, we
assume  that  the  doublet  could  be  attributed  to  iron  atoms
bound  in  small  particle-size  or  Al-substituted  Fe
(oxyhydr)oxides also. Both Al substitution and small particle-
size often lead to the decrease of the magnetic ordering tem-
perature  of  Fe  (oxyhydr)oxides  (Golden  et  al.  1979;  Murad
1988,  1989;  Friedl  &  Schwertmann  1996;  Betancur  et  al.
2004).  The  Mössbauer  spectrum  of  the  sample  J45  contains
four sextets at 80 K (Fig. 6). Based on the hyperfine parame-
ters (Table 2), two sextets assigned to hematite and Al-hema-

Fig. 6. Mössbauer spectrum of sample J45 after 30 days of iron-
bentonite interactions, measured at 80 K.

Fig. 7. Mössbauer spectrum of sample K45 after 30 days of iron-
bentonite interactions, measured at 80 K.

Fig. 8. Mössbauer spectrum of sample DV45 after 30 days of iron-
bentonite interactions, measured at 80 K.

Fig. 9. Mössbauer spectrum of sample L45 after 30 days of iron-
bentonite interactions, measured at 80 K.

540

OSACKÝ, HONTY, MADEJOVÁ, BAKAS and ŠUCHA

tite account for 16 % and 13 % of the total Fe in the sample,
respectively. The sextet with B

of 43.3 T could be attributed

to goethite with partial substitution of Al for Fe in the struc-
ture. This phase accounted for 13 % of the total Fe in the sam-
ple. However, for more accurate mineral assignment the
variable temperature Mössbauer spectroscopy should be ap-
plied. The sextet in the sample J45 assigned to residual metal-
lic Fe accounted for only 17 % of the total Fe in the sample.
This finding indicates that most of the metallic Fe reacted dur-
ing iron-clay interaction and is chemically bonded.

The Fe in the sample DV45 also exists as metallic Fe in

the quantity 18 % of the total Fe in the sample; however, in
this sample more of the total Fe is bound in hematite and Al-
hematite (total 38 %) compared to the sample J45 (Fig. 8,
Table 2).

The Mössbauer spectrum of the sample K45 is similar to the

spectrum  of  J45  (Figs. 7,  6).  The  most  evident  difference  is
that  the  metallic  Fe  in  the  sample  K45  accounts  for  46 %  of
the  total  Fe  in  the  sample.  Accordingly,  the  consumption  of
metallic  Fe  is  much  lower  in  the  sample  K45  with  regard  to
the  sample  J45  during  iron-clay  interactions.  The  relative
amount of metallic Fe in the iron phase is the lowest (12 %) in
the sample L45 of all the studied samples (Table 2).

In the sample La45 only two Fe components are present –

montmorillonite and residual metallic Fe (Table 2). The cen-
tral doublet in the sample La45 is assigned to montmorillonite.
This phase accounted for 9 % of the total Fe in the sample.
The sextet in the sample is assigned to residual metallic Fe,
which accounted for 91 % of the total Fe in the sample.

The Mössbauer spectroscopy revealed that in the samples

J45,  K45,  DV45  and  L45  the  Fe  was  distributed  among  five
Fe-bearing  components,  in  the  sample  La45  only  two  Fe
components were present (Table 2). The Mössbauer spectrum
of  the  sample  La45  was  very  different  with  respect  to  the
Mössbauer spectra J45, K45, DV45 and L45. The other differ-
ence in spectra of particular samples was in the quantity of Fe
in these Fe-bearing components. The residual metallic Fe ac-
counted  for  12—46 %  of  total  Fe  in  the  samples  J45,  K45,

DV45 and L45, whereas 91 % of the total Fe was present as
metallic Fe in the sample La45.

Cation exchange capacity

The results of cation exchange capacity (CEC) are given in

Table 3. After iron-bentonite interactions the CEC values de-
creased in all samples. This change indicates the decrease in
the content of expandable structure which might correspond to
the decrease in the smectite content after the experiment. The
decrease in CEC values was more significant with prolonga-
tion of the reaction time.

Solution chemistry

The results of AAS are summarized in Table 4. The conduc-

tivity, pH and Eh measured after cooling of the experimental
solution  at  room  temperature  are  given  in  Table 5.  After  30
days  solution  pH  was  slightly  alkaline  between  7  and  8,  Eh
values ranged from —14 to  + 10 mV. After 30 days the concen-
trations of Si and Al were similar, Fe concentration was negli-
gible in all samples. After 120 days all element concentrations
increased (except for Al) in all samples.

Discussion

The results of FTIR spectroscopy and quantitative XRD

analysis (RockJock) indicate the presence of volcanic glass in
our samples. The presence of an absorption band at ~ 800 cm

—1

Sample

0 days

(meq/100g)

30 days

(meq/100g)

120 days

(meq/100g)

J45

L45

La45

n.a.

K45

102 83 85

DV45

Table 3: Cation exchange capacities (meq/100 g) of starting sam-
ples (0 days) and samples after 30 and 120 days of iron-bentonite
interactions.

Table 4: Chemistry of reaction solutions (in mg/l) after 30 and 120
days of iron-bentonite interactions determined by AAS (n.a. – not
analysed).

Sample

(mg/l)

30 days

J45

27.03 0.53 <0.265 12.19 117.66 351.92 230.02

L45

44.31 0.53 <0.265 6.04 69.43 203.52 155.82

La45

528.94 1.80 <0.265 22.26 111.30 789.70 85.97

K45

28.62 0.53 <0.265 10.39 58.51 474.88 147.34

DV45

15.69 0.64 <0.265 1.72 35.40 83.74 433.54

120 days

J45

50.24 0.53 3.40 68.37 761.08 538.48 472.76

L45

74.41 0.53 8.17 60.42 770.62 499.26 486.54

La45

n.a. n.a. n.a. n.a. n.a. n.a. n.a.

K45

69.21 0.53 2.93 54.59 418.70 881.92 289.38

DV45 151.58 1.69 283.02 23.53

83.52 201.40 519.40

Fig. 10. Mössbauer spectrum of sample La45 after 30 days of iron-
bentonite interactions, measured at 80 K.

541

EXPERIMENTAL INTERACTIONS OF BENTONITES WITH METALLIC IRON (SLOVAKIA)

was assigned to the amorphous SiO

(Fig. 5). According to

Číčel et al. (1992) the absorption band at ~ 800 cm

—1

suggests

the presence of Si bound in crystalline and/or non/crystalline
phases like volcanic glass, opal, tridymite, cristobalite. More-
over, Číčel et al. (1990) observed differences in the structural
formulas of montmorillonite from Jelšový Potok calculated
from the acid dissolution and from bulk chemical analysis.
These differences were attributed to the high extraneous SiO

amount  in  the  sample.  The  authors  observed  that  about  one
third of the mass of the sample was found to be extraneous to
the  montmorillonite  and  it  was  supposed  that  it  is  bound  in
feldspar  and  mainly  in  a  volcanic  glass.  We  determined  the
quantitative  amount  of  cristobalite  and  volcanic  glass  in  our
samples  from  XRD  patterns  using  the  RockJock  program.  If
the  volcanic  glass  was  included  in  the  calculation  of  the
mineral composition of bentonites, the degree of fit between
measured  and  calculated  XRD  pattern  was  much  better  than
the  degree  of  fit  when  the  volcanic  glass  was  excluded  from
the calculation. This could be further circumstantial evidence of
the presence of volcanic glass in the bentonite samples.

During the iron-bentonite interactions all experimental com-

ponents  may  be  altered  and  new  mineral  phases  could  be
formed  according  to  the  published  data.  Magnetite  and  1 : 1
Fe-rich  7

phyllosilicates (berthierine-like) are reported as

the main reaction products of these interactions at a tempera-
ture of about 80 °C (Habert 2000; Lantenois 2003; Perronnet
2004; Lantenois et al. 2005; Wilson et al. 2006). Lantenois et
al. (2005) and Perronnet et al. (2007) found that iron is oxi-
dized in the presence of smectite, forming magnetite. Wilson
et al. (2006) suggest that either green-rust or magnetite could
be produced from the oxidation of iron.

The behaviour of the samples J45, K45, L45 during the ex-

periments agrees with the published data. The lack of ber-
thierine in our laboratory products could be explained by the
aerobic experimental conditions. In fact, Harder (1978) found
that Fe-containing clay minerals including 1 : 1 Fe-rich 7

phyllosilicates can be synthesized in a short time at low tem-
peratures  ( ~ 20 °C)  but  only  under  reducing  conditions.  De-
crease  of  smectite  diffraction  peaks  is  a  common  feature  of
iron-clay interactions (Guillaume et al. 2003; Lantenois et al.
2005; Wilson et al. 2006). It may indicate decrease of smectite
content (which means dissolution of smectite), split of smec-
tite crystallites and/or smectite structural changes which re-
sult  in  a  decrease  of  the  number  of  coherently  diffracting
smectite domains.

We observed a clear decrease of the smectite XRD patterns

intensity  and  CEC  values  after  iron-bentonite  interactions.
These changes indicate the decrease in the smectite content af-
ter experiments. In contrast, FTIR spectra are the same before
and  after  experiments,  which  implicitly  indicates  a  process
other than dissolution. As we did not identify any changes in
FTIR spectra on the basis of Madejová et al. (1998) a study,
which  revealed  that  the  FTIR  spectroscopy  is  able  to  detect
decrease  in  10 %  of  structural  Al  in  SWy-1  montmorillonite
after  acid  treatment,  we  assumed  that  structural  changes  in
smectites did not exceed 10 %. L45 bentonite contains Al—Fe

montmorillonite with the highest octahedral Fe

of all sam-

ples used for our experiments. However, there is no clear evi-
dence that this smectite is destabilized more than smectites in

samples J45, K45 and La45 (smectites with low structural
Fe

). The conceptual model of smectite destabilization in the

presence of iron was proposed by Lantenois et al. (2005). Ac-
cording to this model, destabilization of smectites resulting
from the presence of trioctahedral (Fe

) domains in the octa-

hedral sheet of reacted dioctahedral (Fe

) smectite, because

of the inability of the tetrahedral sheet to accommodate the
larger dimensions of the newly formed trioctahedral domains.
The deprotonation of OH groups and the reduction of structur-
al Fe

in the smectite are the main processes leading to the

presence of trioctahedral (Fe

) domains in the dioctahedral

(Fe

) smectite structure. Apparently no sign of these changes

was detected by the XRD, FTIR and Mössbauer spectroscopy
in our samples. It indicates that aerobic experimental condi-
tions prevent the above mentioned processes in the system.

The formation of small amounts of lepidocrocite instead of

magnetite in the sample La45 after the experiment (Fig. 3)
could be explained by more oxidizing conditions (the highest
Eh value of all samples) in comparison with the other samples
(Table 5). It means that the Fe

produced by oxidation of iron

was oxidized very fast to Fe

therefore the concentration of

ions in the reaction solution was too low for magnetite

formation. Lantenois (2003) used XRD to identify lepidocroc-
ite as the by-product of the air exposure of the Fe-containing
gel phase after iron-clay interactions at 80 °C.

The very low level of Fe oxidation in the sample La45 in re-

spect to the other bentonites is a surprise. The low reactivity
between bentonite La45 and iron may be connected with a dif-
ferent mineral composition of the bentonite La45 with respect
to the bentonites J45, L45 and K45 (Table 1). The bentonite
La45 contains 42 % of smectite while the smectite content in
bentonites J45, L45 and K45 is much higher (56—81 %). We
suppose that the very low smectite content in the bentonite
La45 could be the reason for low reactivity with iron.

This assumption is consistent with the observations of Per-

ronnet (2004), Lantenois et al. (2005) and Perronnet et al.
(2007) who found that iron is not consumed (oxidized) in the
absence of smectite whereas in the presence of smectite, iron
is oxidized. The high affinity of iron for smectite followed by
iron-smectite interactions have been recognized as the cause
of oxidation of steel pipes when clay-containing drilling fluids
were used (Tomoe et al. 1999; Cosultchi et al. 2003).

Another interesting feature is the drastic deterioration of the

illite-smectite from the Dolná Ves deposit. In DV45 sample
the iron was consumed, magnetite was formed, the diffraction
peaks of illite-smectite decreased significantly after the exper-
iment and a new diffraction peak of smectite was detected
(Fig. 4). The results of infrared spectroscopy showed decrease
in intensities of OH-bending vibrations, which reflect release

Table 5: The properties of reaction solutions measured (at room
temperature) after 30 days of iron-bentonite interactions.

Sample

Eh (mV)

Conductivity (µS/cm)

Redistilled water

5.70

170

5.92

J45

7.59 –10

242

L45

7.30 +10

144

La45

7.94 +78

330

K45

7.77 +10

241

DV45

7.10 –14

144

542

OSACKÝ, HONTY, MADEJOVÁ, BAKAS and ŠUCHA

of the central atoms (Al, Mg) from the octahedral sheet of il-
lite-smectite after the experiment. The decrease in CEC values
indicates  the  decrease  in  the  illite-smectite  content  after  the
experiment. It is clear from the XRD, FTIR and CEC that ex-
perimental  interactions  caused  dissolution  of  the  original  il-
lite-smectite.  The  delamination  of  illite-smectite  crystals  can
be  excluded  because  it  would  have  resulted  in  the  shift  of
XRD peaks (delamination would change the ratio of expand-
able  and  non-expandable  layers).  It  seems  that  the  illite-
smectite  with  its  significantly  lower  surface  area  (1/3  of
smectite)  is  much  less  resistant  to  the  experimental  condi-
tions than the smectites.

The concentrations of alkali and alkali-earth cations, which

were initially present in smectite interlayers (Ca

, Na

, Mg

), were significantly higher after 120 days. This increase is

probably  the  result  of  exchange  processes  between  smectite
interlayer  cations  and  cations  contained  in  reaction  solution.
Increase  in  Si  and  Mg  concentrations  could  be  related  to  the
changes in the smectite structure, and higher Si concentrations
may also be related to quartz and feldspar dissolution during
iron-clay  interactions.  High  Si  concentration  in  the  sample
La45 after 30 days can be explained by dissolution of amor-
phous  SiO

(volcanic glass, cristobalite) during iron-clay in-

teractions.  It  is  in  accordance  with  XRD  data,  because  after
iron  interactions  the  diffraction  peak  of  cristobalite  disap-
peared in the sample La45 after 30 days. Significant increase
in Fe content in all samples after 120 days could be explained
by  changes  in  the  oxidation  state  of  iron  and  the  subsequent
mobilization of the iron.

Conclusions

The experiments compared with the previously published

data show that bentonites do not react with iron in aerobic en-
vironment in the same way as in anaerobic environments.

Three smectitic bentonites – J45, K45, and L45, had very

similar or almost identical reactions. The diffraction peaks of
smectite  decreased  but  were  still  clearly  visible  in  XRD  pat-
terns. The iron was consumed to a large extent and magnetite
was formed (Fig. 2). In the sample La45 the intensity of dif-
fraction  peaks  of  smectite  decreased  but  the  iron  was  much
less consumed (oxidized) during iron-clay interactions and in-
stead  of  magnetite  a  small  amount  of  lepidocrocite  was
formed (Fig. 3). The fact that iron oxidation was prevented in
La45 can be explained by the low content of smectite.

Significant decrease of the smectite XRD patterns intensity

and CEC values during the experiments cannot be assigned to
the dissolution of smectite. It represents, rather, a rearrange-
ment of the original smectite crystals.

Mixed-layer illite-smectite proved to be much less stable in

the presence of iron than smectites. Clear signs of I-S dissolu-
tion were detected along with the neoformation of smectite.

Acknowledgments: The authors are grateful to Dr. Peter
Komadel for constructive comments and help during prepara-
tion of the paper. This study was supported by the Slovak
Grant Agency VEGA (Projects 1/3072/06 and 2/0171/08) and
by the Student Research Grant UK/272/2008.

References

Arthur R., Apted M. & Stenhouse M. 2005: Comment on Long-Term

Chemical and Mineralogical Stability of the Buffer. SKI Report
2005 : 09, Swedish Nuclear Power Inspectorate, Stockholm,
Sweden.

Betancur J.D., Barrero C.A., Greneche J.M. & Goya G.F. 2004: The

effect of water content on the magnetic and structural properties
of goethite. J. Alloys. Compounds 369, 247—251.

Börgesson L., Karnland O., Hökmark H. & Sellin P. 2002: Buffer and

Safety Assessment for KBS-3H. SKB, September 2002.

Cosultchi A., Rossbach P. & Hernandez-Calderon I. 2003: XPS anal-

ysis of petroleum well tubing adherence. Surface and Interface
Analysis 35, 239—245.

Číčel B., Komadel P. & Hronský J. 1990: Dissolution of the fine frac-

tion of Jelšový Potok bentonite in hydrochloric and sulphuric ac-
ids. Ceramics 34, 41—48.

Číčel B., Komadel P., Bednáriková E. & Madejová J. 1992: Mineral-

ogical composition and distribution of Si, Al, Fe, Mg and Ca in
the fine fractions of some Czech and Slovak bentonites. Geol.
Carpathica, Ser. Clays 43, 1, 3—7.

Eberl D.D. 2003: User’s guide to Rockjock – a program for deter-

mining quantitative mineralogy from powder X-ray diffraction
data. U.S. Geol. Surv., Open-File Report 03—78, 47.

Friedl J. & Schwertmann U. 1996: Aluminium influence on iron ox-

ides: XVIII. The effect of Al substitution and crystal size on
magnetic hyperfine fields of natural goethites. Clay Miner. 31,
455—464.

Golden D.C., Bowen L.H., Weed S.B. & Bigham J.M. 1979: Möss-

bauer studies of synthetic and soil-occurring aluminum-substi-
tuted goethites. Soil Sci. Soc. Amer. J. 43, 802—808.

Guillaume D., Neaman A., Cathelineau M., Mosser-Ruck R., Peiffert

C., Abdelmoula M., Dubessy F., Villiéras F., Baronnet A. &
Michau N. 2003: Experimental synthesis of chlorite from smec-
tite at 300 °C in the presence of metallic Fe. Clay Miner. 38,
281—302.

Guillaume D., Neaman A., Cathelineau M., Mosser-Ruck R., Peiffert

C., Abdelmoula M., Dubessy F., Villiéras F. & Michau N. 2004:
Experimental study of the transformation of smectite at 80 °C
and 300 °C in the presence of Fe oxides. Clay Miner. 39, 17—34.

Habert B. 2000: Réactivité du fer dans les gels et les smectites. Thesis,

Université Paris 6, Paris, 1—227.

Harder H. 1978: Synthesis of iron layer silicate minerals under natu-

ral conditions. Clays Clay Miner. 26, 1, 65—72.

JNC 2000: Second progress report on research and development for

the geological disposal of HLW in Japan. H12: Project to estab-
lish scientific and technical basis for HLW disposal in Japan
(Project Overview Report). JNC TN1410 2000-001.

Kaufhold S. & Dohrmann R. 2003: Beyond the methylene blue meth-

od: determination of the smectite content using the Cu-Triene
method. Z. Angew. Geol. 2, 13—17.

Lantenois S. 2003: Réactivité fer métal/smectites en milieu hydraté

80 °C. Ph.D. Thesis, Université d’Orléans, Orléans, 1—188.

Lantenois S., Lanson B., Muller F., Bauer A., Jullien M. & Plançon

A. 2005: Experimental study of smectite interaction with metal
Fe at low temperature: 1. Smectite destabilization. Clays Clay
Miner. 53, 6, 597—612.

Madejová J., Bujdák J., Janek M. & Komadel P. 1998: Comparative

FT-IR study of structural modifications during acid treatment of
dioctahedral smectites and hectorite. Spectrochimica Acta, Part
A, 54, 1397—1406.

Madejová J., Kečkéš J., Pálková H. & Komadel P. 2002: Identifica-

tion of components in smectite/kaolinite mixtures. Clay Miner.
37, 377—388.

Madsen F.T. 1998: Clay mineralogical investigations related to nu-

543

EXPERIMENTAL INTERACTIONS OF BENTONITES WITH METALLIC IRON (SLOVAKIA)

clear waste disposal. Clay Miner. 33, 109—129.

Meier L.P. & Kahr G. 1999: Determination of the cation exchange

capacity (CEC) of clay minerals using the complexes of
copper(II) ion with triethylentetramine and tetraethylenepentam-
ine. Clays Clay Miner. 47, 386—388.

Müller-Vonmoos M., Kahr G., Bucher F., Madsen F.T. & Mayor P.A.

1991: Untersuchungen zum Verhalten von Bentonit in kontakt mit
Magnetit und Eisen unter Endlagerbedingungen. NTB 91-14.
Nagra, Hardstrasse 73, CH-5430, Wettingen, Switzerland.

Murad E. 1988: Properties and behaviour of iron oxides as deter-

mined by Mössbauer spectroscopy. In: Stucki J.W., Goodman
B.A. & Schwertmann U. (Eds.): Iron in soils and clay minerals.
Reidel, Dordrecht, 309—350.

Murad E. 1989: Poorly-crystalline minerals and complex mineral as-

semblages. Hyperfine Interactions 47, 33—53.

Perronnet M. 2004: Etude des interactions fer-argile en condition de

stockage géologique profond des déchets nucléaires HAVL.
Ph.D. Thesis, ENS Géologie, Nancy, 1—233.

Perronnet M., Villiéras F., Jullien M., Razafitianamahavaro A.,

Raynal J. & Bonnin D. 2007: Towards a link between the ener-
getic heterogeneities of the edge faces of smectites and their sta-
bility in the context of metallic corrosion. Geochim. Cosmochim.
Acta 71, 1463—1479.

Push R. 2001: The Buffer and Backfill Handbook. Part 2: Materials

and techniques. SKB TR 02-12. Swedish Nuclear Fuel and
Waste Management Co., Stockholm, Sweden.

Push R. 2003: The Buffer and Backfill Handbook. Part 3: Models for

calculation of processes and behavior. SKB TR 03-07. Swedish
Nuclear Fuel and Waste Management Co., Stockholm, Sweden.

SKI 2005: Engineered barrier system – Long-term stability of

buffer and backfill. SKI Report 2005 : 48. Swedish Nuclear
Power Inspectorate, Stockholm, Sweden.

Smart N.R., Blackwood D.J. & Werme L. 2002: Anaerobic corro-

sion of carbon steel and cast iron in artificial groundwaters:
Part 1 – Gas generation. Corrosion 58, 627—637.

Środoń J., Drits V.A., McCarty D.K., Hsieh J.C.C. & Eberl D.D. 2001:

Quantitative X-ray diffraction analysis of clay-bearing rocks from
random preparations. Clays Clay Miner. 49, 6, 514—528.

Thorsager P. & Lindgren E. 2004: KBS-3H – Summary report of

work done during basic design. SKB Rapport R-04-42. Swedish
Nuclear Fuel and Waste Management Co., Stockholm, Sweden.

Tomoe Y., Shimizu M. & Nagae Y. 1999: Unusual corrosion of a

drill pipe in newly developed drilling mud during deep drilling.
Corrosion 55, 706—713.

Wilson J., Cressey G., Cressey B., Cuadros J., Vala Ragnarsdottir

K., Savage D. & Shibata M. 2006: The effect of iron on mont-
morillonite stability. (II) Experimental investigation. Geochim.
Cosmochim. Acta 70, 323—336.