GEOLOGICA CARPATHICA, 50, 3, BRATISLAVA, JUNE 1999
257272
DIAGENETIC TRIOCTAHEDRAL PHYLLOSILICATES FROM
SEDIMENTS OF THE AMBRON ZONE (EASTERN SLOVAKIA):
XRD, SEM, AND EMPA STUDY
ADRIAN BIROÒ*, JÁN SOTÁK and JURAJ BEBEJ
Geological Institute, Slovak Academy of Sciences, Bratislava; Branch: Severná 5, 974 01 Banská Bystrica, Slovak Republic;
*biron@gu.bb.sanet.sk
(Manuscript received March 24, 1998; accepted in revised form March 17, 1999)
Abstract: Trioctahedral clay minerals occurring as phyllosilicate cement in diagenetically altered serpentinitic graywackes
have been studied using X-ray powder diffraction, scanning electron microscopy, and microprobe analyses. Two distinct
cement assemblages were observed: (1) saponite ± calcite ± dolomite ± opal-CT ± pyrite, which is restricted to quartz-
rich graywackes, and (2) ordered mixed-layered chlorite/smectite (C/S) + saponite ± calcite ± dolomite ± opal-CT ±
pyrite characteristic of serpentinite-rich graywackes. Swelling properties as well as chemical analyses reveal a predomi-
nance of high-charge layers both in saponite and C/S. Saponite and C/S show a unique composition of 2:1 layers with
respect to Si/(Si+Al) ratio. The decrease in Mg/(Mg+Fe) ratio from saponite to C/S with respect to the increasing Al
IV
content implies that Fe-Mg substitution was controlled by distortion of tetrahedral sheets caused by Si-Al substitution.
The lack of correlation between whole-rock and phyllosilicate Mg/(Mg+Fe) ratios supports this interpretation. Textural
as well as compositional evidence suggests that both saponite and C/S originated by the interaction of the sediment with
pore-fluids during burial as direct precipitates. It is inferred that a different bulk rock composition, and consequently, a
different chemistry of pore-fluids played an important role during authigenesis. Particularly, the availability of Al may
have been a primary factor controlling whether saponite or C/S formed, while the role of Mg (or Mg/(Mg+Fe) ratio) as
well as the role of temperature were only of secondary importance.
Key words: ambron Zone, serpentinitic sandstones, diagenesis, saponite, mixed-layer chlorite/smectite, corrensite.
(without saponite precursor) has also been reported (Almon et
al. 1976; Shau et al. 1990; Bettison-Varga et al. 1991; Bautier et
al. 1995).
This study is concerned with saponite and corrensite-like
chlorite/smectite mixed layer (C/S) clay from serpentinitic
sandstones of the flysch formations of the ambron Zone of
the Central Carpathian Paleogene Basin (Levoèa sub-Basin)
(Soták & Bebej 1996). The sandstones are significantly en-
riched with unstable lithic components serpentinized ul-
tramafic rocks. This specific composition gave rise to the
formation of mafic expandable clay minerals during burial
diagenesis. The exposed sedimentary sequence with serpen-
tinitic sandstones represents a narrow stratigraphic range
(only a few meters in thickness), so that physical conditions
of post-sedimentary alteration were equal for each sandstone
bed. Nevertheless, they show significantly different mineral-
ogy of the authigenic clay phases and individual sandstone
beds have either saponitic or corrensitic phyllosilicate ce-
ment. Apparently, the conventional scheme of temperature
controlled saponite to chlorite series cannot explain such
spatial distribution of authigenic assemblages. This indi-
cates, that instead of temperature, other factors influenced
the origin of the clay minerals in serpentinitic sandstones
from the ambron Zone.
In this paper, it is our intention to: (1) demonstrate the oc-
currence, structure and chemistry of saponite and C/S; and
(2) propose a model that explains their formation in the ser-
pentinitic sandstones of the ambron Zone.
Introduction
Trioctahedral clay minerals such as saponite and corrensite (1:1
regularly interstratified chlorite/smectite) are abundant in di-
verse geological settings. They widely form under diagenetic to
subgreenschist facies conditions in regionally metamorphosed
and hydrothermally altered mafic igneous rocks (e.g. Krist-
mannsdóttir 1978; Bettison & Schiffman 1988; Shau et al. 1990;
Schiffman & Fridleifsson 1991; Shau & Peacor 1992; Robinson
et al. 1993; Schiffman & Stautigel 1995), evaporite and carbon-
ate sequences (e.g. April 1981; Bodine & Madsen 1987; Fisher
1988; Janks et al. 1992; Masaryk et al. 1995), or clastic sedi-
mentary rocks containing significant amounts of mafic volcano-
clastic material (e.g. Iijima & Utada 1971; Galloway 1974; Al-
mon et al. 1976; Helmond & van de Kamp 1984; Inoue et al.
1984; Chang et al. 1986; Inoue & Utada 1991; Bautier et al.
1995). In all the above mentioned environments, the origin of
chlorite/smectite minerals is related to high concentrations of
Mg. Corrensite generally occurs as an intermediate product dur-
ing the smectite-to-chlorite transition. Most commonly, this
transition occurs through the discontinuous sequence of three
phases saponite, corrensite, and chlorite, concurrent with in-
creasing temperature of formation (e.g. Inoue & Utada 1991;
Meunier et al. 1991; Schiffman & Stautigel 1995). Only a few
examples have been interpreted as a more or less continuous se-
ries of mixed-layer chlorite/smectite minerals (Schultz 1963;
Helmond & van de Kamp 1984; Chang et al. 1986). Alternative-
ly, a direct precipitation of corrensite from concentrated solution
258 BIROÒ, SOTÁK and BEBEJ
Geological setting
The serpentinitic graywackes were found in the ambron
Zone of the Central Carpathian Paleogene in Eastern Slovakia
(localities near Kamenica Fig. 1; Soták & Bebej 1996).
They occur within a thinly bedded sandstoneclaystone se-
quence resembling the zebra-type flysch. The sandstones are
grey in colour, fine- to medium grained and usually laminated
in Bouma Td intervals. The claystones are weakly calcareous
(average content of CaO and MgO in carbonate fraction is
5.12 and 3.68 wt. %, respectively) and with their brownish co-
lour close to the Menilite shales. The sandstoneclaystone se-
quence is also interbedded with thin layers of whitish dolo-
mites. The age corresponds to the Late Oligocene (NP 2425
nanoplanktone zone; Soták & Bebej 1996). The graywacke
flysch deposits near Kamenica are refolded into recumbent to
moderately plunging folds and intersected by overthrust faults.
The occurrence of serpentinite-rich flysch deposits in the am-
bron Zone indicates a suture zone of the Tertiary collision
running along the junction of the Central Carpathian Paleo-
gene with the Pieniny Klippen Belt. The ambron Zone is
thought to be a perisutural part of the Central Carpathian Pa-
leogene Basin enriched with ophiolitic detritus.
Analytical methods
Samples were examined by the following methods: optical
microscopy, X-ray powder diffraction (XRD), scanning elec-
tron microscopy (SEM), back-scattered electron imaging
(BSE). The chemical composition of clay minerals was
determined by electron microprobe analysis (EMPA) and
whole-rock chemical composition by X-ray fluorescence
spectrometry.
XRD. Preparation of clay size fractions for XRD analyses
utilized standard procedures. Samples were crushed to obtain
chips about 5 mm in diameter. Rock fragments were then dis-
aggregated in distilled water using an ultrasonic bath for 10
min to extract phyllosilicates from pore spaces. Suspensions
were flocculated with NaCl and subsequently treated with
sodium acetate buffer, H
2
O
2
and with sodium dithionite to re-
move carbonate minerals, organic matter and Fe, Mn-oxides,
respectively (Jackson 1975). The <2
µ
m fraction was sepa-
rated by sedimentation in distilled water. Mg
2+
and K
+
forms
of clay fractions were prepared by exchange (three times,
overnight) with 1 M chloride solutions. Finally, clay frac-
tions were repeatedly washed with distilled water, centri-
fuged to remove excess salts, and then air dried.
Oriented samples for XRD analysis were prepared by sedi-
mentation of a clay suspension on glass slide (10 mg/cm
2
)
and analyzed in the air dried state and after saturation by eth-
ylene glycol vapour for 8 h at 60
o
C. In addition, Mg-saturat-
ed clay was also analyzed after solvation with glycerol va-
pour at 100
o
C for 24 h (Reynolds 1988). The K-saturated
samples were heated at 550
o
C for 1 h and then analyzed.
XRD patterns of randomly oriented specimens were also ob-
tained in order to determine the d(06,33) value of phyllosili-
cates. Silicon powder was used as an internal standard.
XRD analyses were performed on a Philips PW 1710 dif-
fractometer (35 kV, 20 mA, CuK
α
, Ni-filter, divergence slit:
1
o
, receiving slit: 0.1 mm, range: 250
o
2
Θ
) and a DRON 2.0
diffractometer (35 kV, 20 mA, CuK
α
, Ni-filter, divergence
Fig. 1. Geological sketch-map of the Levoèské vrchy Mts. depicting the location of serpentinitic sandstone occurrences near the junction
of the Central Carpathian Paleogene with the Pieniny Klippen Belt (westwards of Kamenica village, black arrow).
TRIOCTAHEDRAL PHYLLOSILICATES: XRD, SEM AND EMPA STUDY 259
slit: 1 mm, receiving slit: 0.1 mm, range: 150
o
2
Θ
and 58
63
o
2
Θ
). Oriented preparations were scanned at goniometer
speed 1
o
2
Θ
/min, unorientated powders at 0.5
o
2
Θ
/min.
SEM. Specimens for SEM were prepared as freshly frac-
tured, air dried, gold-coated rock fragments or polished, car-
bon-coated thin sections. Both secondary electron and back-
scattered electron imaging were employed. SEM
observations were carried out on a JEOL JSM 840 scanning
electron microscope equipped with energy-dispersive spec-
trometer and a Tesla BS 300 scanning electron microscope.
EMPA. Quantitative, wavelenght-dispersive analyses of sa-
ponite and C/S were performed on a JEOL JXA-733 Super-
probe electron microprobe at 15 keV accelerating voltage
and 3 nA beam current using an electron beam approximate-
ly 3
µ
m in diameter. The following silicate and oxide stan-
dards were used: albite (Na), orthoclase (K), Al
2
O
3
(Al),
quartz (Si), wollastonite (Ca), MgO (Mg), hematite (Fe),
rhodonite (Mn), TiO
2
(Ti) and chromite (Cr). Corrections
were made with the ZAF computing program.
Materials
The detailed petrographic description of sandstones has
been given by Soták & Bebej (1996). The sandstones are
mostly graywackes (Füchtbauer 1959) with lithic grain con-
tent between 40 and 80 %. Lithic grains include predomi-
nantly fragments of serpentinized ultramafic rocks with the
typical mesh and loop textures. The coarse-grained, flaky liz-
ardites and fragments of fibrous chrysotiles are also present.
In addition, the detrital framework contains a subordinate
proportion of glassy fragments, large detrital flakes of mus-
covite, biotite or chlorite, carbonate and phyllitic rocks. Intr-
aclasts from the surrounding shales are relatively frequent.
Other major detrital constituents are quartz grains (10 to
50 %). Feldspars form only an accessory component of the
sandstones (up to 10 %). Both alkali feldspars and plagio-
clases (Na- or Ca-dominated) are present.
The sandstones contain abundant phyllosilicate cement.
Clay minerals are present both as a pore-filling cement, and
as an alteration phase in lithic and feldspar grains. Non-phyl-
losilicate authigenic minerals are volumetrically insignifi-
cant and they are not uniformly distributed. Considering de-
trital modes and cement mineralogy, the sandstones can be
divided into two distinct types:
(A) Quartzolithic graywackes with significant content of si-
liciclastic material (Q
37
F
12
L
51
) and authigenic assemblage: sa-
ponite and subordinate dolomite, calcite, opal-CT and pyrite.
(B) Lithic graywackes with absolute prevalence of the lithic
component (Q
19
F
7
L
74
) and authigenic assemblage: C/S + sa-
ponite and subordinate dolomite, calcite, opal-CT and pyrite.
The average whole-rock chemical compositions are shown
in Table 1.
Results
Habit of phyllosilicates
Quartzolithic graywackes. The only authigenic clay miner-
al in quartzolithic graywackes is trioctahedral smectite-sapo-
nite. Saponite occurs as a pore-filling cement and usually
completely occludes the intergranular spaces. Observed un-
der the SEM, pore-filling saponite exhibits radial alignment
of densely packed, up to 50
µ
m long platy crystals (or aggre-
gates of crystals ?) attached approximately perpendicularly
to the detrital grain surfaces. Pore-filling aggregates show
well-developed medial sutures (Fig. 2A), indicating growth
of crystal platelets inward from surrounding framework
grains toward the center of interstitial voids (Dickinson
1970; Wilson & Pittman 1977). If seen perpendicularly to
their prolongation, crystals form an irregular network with a
cellular pattern (Fig. 2B). The similar boxwork-like appear-
ance of saponite have been found in the open pores (Fig. 2C).
Saponite is absent at the detrital grain contacts (Fig. 2D).
The characteristic feature of quartzolithic graywackes is
scarcity of grain-alteration phyllosilicates (in contrast to the
lithic graywackes described below). Only saponite infilling
large pores in the vicinity of feldspar grains may be interpret-
ed as originating by their partial replacement (Fig. 2D).
Within lithic constituents, which, in these sandstones, are
represented mainly by completely serpentinized ultramafic
rocks, replacement textures are very rare. They can be identi-
fied in the BSE images as a zones with darker image show-
ing shrinking cracks due to dehydration of smectitic interlay-
ers in the vacuum chamber (Fig. 2E). Also microprobe
analyses obtained from these zones indicate the presence of
smectite-like layers (increased Al and Ca contents) in other-
wise serpentine-dominated material. Saponite was also found
in dolomite beds. It fills small cavities and displays radiating
textures (Fig. 2F).
Lithic graywackes. The dominant authigenic phyllosilicate
in lithic graywackes is C/S. The SEM study reveals that C/S
occurs in many textural habits. The main form of occurrence,
petrographically very similar to saponite, is pore-filling ce-
ment. C/S crystals are aligned radially with respect to the
framework grains with distinct sutures in the center of pores
(Fig. 3A). If pore space is not entirely infilled, aggregates
show honeycomb patterns of curved, interlocking crystal
platelets (Figs. 3B and 3C). Rarely, C/S forms flat and isomet-
Table 1: Whole-rock chemical analyses of serpentinitic graywackes.
Type
Q uartzolith ic g rayw ack es
L ithic g rayw ack es
Mean d etrital mod e
Q
37
F
12
L
51
Q
19
F
7
L
74
Samp le
2 5 8 /V
K -I
2 5 5 /II
2 5 6/TM 1
SiO
2
5 6.42
5 9.65
49.5 0
47.70
47.80
47.60
A l
2
O
3
7.0 8
7.8 3
7.7 2
10.62
11.5 5
10.06
Mg O
11.98
8 .4 0
14.47
16.68
14.31
11.39
Fe
2
O
3
4.3 93.7 4
4.5 1
6.4 2
7.3 2
5 .2 4
Mn O
0.0 5
0.0 4
0.0 3
0.0 2
0.0 3
0.0 5
TiO
2
0.4 0
0.3 7
0.6 1
0.5 3
0.4 3
0.6 4
CaO
6.3 2
5 .98
7.6 5
3.6 9
3.1 3
8 .91
N a
2
O
0.0 1
0.0 4
0.6 90.0 1
1.0 6
0.6 9
K
2
O
1.2 3
2 .1 0
1.2 2
1.4 5
1.91
1.90
Total
8 7.87
88.14
86.41
8 7.13
87.5 3
8 6.47
Si/(Si+A l)
0.931
0.92 8
0.916
0.8 8 4
0.8 75
0.8 8 9
Mg /(Mg +Fe)
0.8 44
0.8 17
0.8 64
0.8 38
0.7 95
0.8 12
260 BIROÒ, SOTÁK and BEBEJ
TRIOCTAHEDRAL PHYLLOSILICATES: XRD, SEM AND EMPA STUDY 261
tal feldspars are pseudomorphously replaced by saponite, C/S
cementation is restricted to the original outlines of the grains
(Fig. 4D). This textural relationship clearly reflects intense in
situ dissolution.
X-ray diffraction
Two types of authigenic layer silicates were identified by
XRD study: discrete saponite and R1 ordered mixed-layered
chlorite/smectite.
Saponite. XRD traces of Mg-solvated, air-dried saponite
(Fig. 5A) exhibit an intense 001 reflection at about 14.6 Å,
and weaker higher order (00l) reflections at 4.87 (003), 3.65
(004), 2.920 (005), 2.433 (006) and 2.080 Å (007). The basal
spacing of 14.6 Å indicates two-water-interlayer complex
(MacEwan & Wilson 1980). The glycerol solvation follow-
ing Mg-saturation produced collapse of the saponite structure
(Fig. 5A). The higher order reflections at 4.77 (003), 3.58
(004), 2.868 (005) and 2.043 Å (007) indicate basal spacing
of approximately 14.3 Å, characteristic of single-layer glyc-
erol complex (MacEwan & Wilson 1980). However, distinct
intensity reduction and low-angle-side asymmetry of the 001
reflection suggest heterogeneity in layer charge of this clay.
The estimated position of the first-order basal reflection of
this additional phase at about 17 Å is indicative of two-layer
glycerol complex (MacEwan & Wilson 1980). This conclu-
sion is supported by the existence of weak reflections at 5.97
and 3.02 Å, representing 003 and 006 basal reflections, re-
spectively (Fig. 5Aarrows). In contrast, swelling behav-
iour of the saponite upon Mg-saturation and ethylene glycol
solvation is uniform. The XRD patterns show one series of
rational 00l reflections with d(001) spacings of 17.02 Å
(Fig. 5A), which is indicative of two-layer glycol complex
(MacEwan & Wilson 1980). In the K-saturated, air-dried
samples first-order basal reflection at 12.4 Å (not shown
here) suggests the presence of one-water layer hydrate
(MacEwan & Wilson 1980). Heat treatment to 550
o
C for
one hour caused a full dehydration of this clay and collapse
of the structure to 10.2 Å (Fig. 7A). XRD analysis of ran-
domly oriented <2
µ
m fractions gave 06,33 reflections at
1.535 (± 0.001) Å (Fig. 5B) and a calculated unit-cell dimen-
sion of 9.21 Å. In all XRD patterns the basal reflections of
saponite show a rational series of spacings, characteristic for
pure mineral without structural interstratification. Clay frac-
tions usually contain small admixtures of dioctahedral illite,
chlorite and a mineral of serpentine group, which are consid-
ered to be detrital in origin.
The swelling behaviour displayed by the authigenic clay
from quartzolithic graywackes is intermediate between that of
smectite (i.e. two-layer glycol complex and no structural col-
lapse upon K-saturation) and vermiculite (i.e. one-layer glyc-
erol complex), indicating a high negative charge of the 2:1
layers. Such swelling properties are characteristic both of
high-charge saponite and low-charge vermiculite, which
makes categorization of this phase difficult (Malla & Douglas
1987; de la Calle & Suquet 1988). However, the layer charge
of 1.05 per O
20
(OH)
4
calculated from EMPA (see below) is
below the limiting value of 1.20 per O
20
(OH)
4
, proposed by
the AIPEA Nomenclature Committee for distinguishing smec-
Fig. 2. Photographs of authigenic saponite from quartzolithic
graywackes. A SEM photomicrograph of pore-filling aggregate
of saponite showing radial alignment of crystallites and well-devel-
oped medial suture. Crystallites are perpendicular to the detrital
grain surfaces; Cal calcite. B SEM photo of pore-filling sapo-
nite viewed perpendicularly to the prolongation of crystallites.
Highly folded platelets form a cellular pattern, sometimes with a
concentric organization (upper left). C The boxwork-like appear-
ance of saponite in the open pore. Saponite is overlain by opal-CT
lepispheres (SEM). D A large pore entirely infilled by saponite.
Saponite partly replaces detrital albite grain (Ab). Surrounding
framework consist of quartzose grains. Note the absence of saponite
at the grain-to-grain contacts, e.g. lower right (BSE image). E
BSE image of detrital serpentinite grain. The darker-contrast zones
are partly replaced by saponite (see text for further explanation). F
Saponite exhibiting radial arrangement of crystallites infills
rounded cavity in metasomatically (?) dolomitized sandstones. Cav-
ity is rimmed by euhedral crystals of Fe-dolomite (BSE image).
ric (in ab plane) crystals with rounded outlines, tangentially at-
tached to the grain surface. Particles range from 1 to 3
µ
m in
diameter and sometimes form rosette-like clusters (Fig. 3D).
The authigenic C/S also occurs as an alteration phase in
lithic grains. Replacement textures are best developed within
partially serpentinized detrital grains where C/S infills
rounded voids after completely dissolved primary mafic min-
eral (olivine ?). In thin sections (BSE image), original detri-
tal grains are recognizable through preserved mesh texture,
since the serpentinized parts of clasts remained conserved
(Fig. 3E). The void-filling C/S exhibit characteristic textural
zonation, with coarser-grained, up to 20-
µ
m-wide rims and
fine-grained interiors. Initial parallelly stacked platelets are
normal to the inner wall of voids, while later platelets display
rather random arrangement (Fig. 3F). Considering the physi-
cally unstable character of clay replacements, alteration un-
doubtedly occurred in situ, as grains could not withstand
transport and subsequent compaction.
In many instances lithic grains were replaced completely
and their original character is unknown. However, taking
into account the frequent presence of apatite and/or opaque
minerals in these aggregates it is assumed that they represent
altered fragments of fresh igneous minerals or vitritic clasts.
Observed under the SEM, alteration C/S commonly shows
multi-layered, curvilinear textures (Figs. 4A and 4B). Exten-
sive replacement has locally resulted in almost complete de-
struction of the detrital framework of the sandstones. In such
domains, massive aggregates of C/S commonly contain
floating grains of chemically stable detrital constituents (e.g.
quartz, serpentinites or large flakes of detrital phyllosilicates,
Fig. 4B). On the other hand, C/S rims never occur if grain-to-
grain contact is preserved.
The second diagenetic clay mineral present in lithic
graywackes is saponite. Saponite is developed exclusively as a
replacement phase after detrital feldspars. It usually infills
small dissolution voids (Fig. 4C), but nearly complete pseudo-
morphs after grains were also observed (Fig. 4D). In this type
of graywacke, saponite was never found as pore-filling ce-
ment. From petrographic observations, it appears that saponite
replacements post-date earlier formation of C/S. Even if detri-
▲
262 BIROÒ, SOTÁK and BEBEJ
TRIOCTAHEDRAL PHYLLOSILICATES: XRD, SEM AND EMPA STUDY 263
Fig. 4. Photographs illustrating C/S and saponite from lithic graywackes. A SEM photo of a large pore filled by C/S showing multi-
layered, curvilinear texture. B The same situation as in (A) seen in the BSE image. Extensive dissolution and replacement partly de-
stroyed detrital framework of the sandstones producing massive aggregates of C/S with floating grains of chemically stable detrital con-
stituents. C BSE image of saponite filling dissolution voids in detrital grain of K-feldspar (Kfs). D Detrital grain of Ca-dominated
plagioclase (Plg) almost entirely replaced by saponite. Note that the formation of saponite post-dates precipitation of C/S since C/S ce-
ment respects the original outlines of detrital grains (BSE image).
Fig. 3. Photographs of diagenetic C/S from lithic graywackes. A Authigenic pore-lining (or pore-filling) C/S showing growth inward
from opposing sides of pores. Crystals intersect along distinct sutures near the center of the pore (SEM). B and C Curved, interlocking
crystallites of C/S organized in honeycomb arrangement (SEM). D Rosette-like cluster of C/S crystallites (SEM). E A detail of
partly serpentinized (dark strips) detrital grain exhibiting characteristic mesh texture. Primary mafic mineral (olivine ?) was entirely re-
placed by C/S forming irregular-shaped aggregates with lighter contrast (BSE image). F C/S filling dissolution voids in lithic grain.
Note the distinct textural and compositional zonation of aggregates. The rim material consists of platy and parallelly aligned crystallites
with a lighter BSE image (high Fe content) while the core material displays rather random arrangement.
▲
264 BIROÒ, SOTÁK and BEBEJ
tites from vermiculites (Bailey 1980). This defines this ex-
pandable phase as high-charge saponite. On the other hand,
the ability to form simultaneously two types of glycerol com-
plexes indicates heterogeneity in layer charge distribution
(Malla & Douglas 1987; April & Keller 1992), i.e. the pres-
ence of a small proportion of a low-charge smectite.
Mixed-layer chlorite/smectite. XRD patterns of Mg-satu-
rated, air-dried, glycerolated and glycolated mixed-layer
chlorite/smectite are shown in Fig. 6. In clay fraction, C/S al-
ways occurs with discrete saponite and traces of detrital
phyllosilicates (illite, chlorite, serpentine) which hampers in-
terpretation of XRD profiles due to interference of reflec-
tions. Nevertheless, the following characteristics have been
observed.
The presence of superstructure reflection near 29 Å in the
air-dried pattern which shifts to 31 Å after glycolation indi-
cates the presence of R1 ordered C/S (Reynolds 1988). Gen-
erally, these reflections as well as higher order reflections are
weak and broad (particularly after ethylene glycol solvation).
This suggests that the ratios of component layers in this C/S
differ significantly from the ideal corrensite, and consequent-
ly, some amount of disorder. The percentage of chlorite lay-
ers was estimated on the basis of migration of basal reflec-
tions in response to ethylene glycol solvation (Reynolds
1980, 1988). Comparison of observed and calculated basal
spacings indicates a chlorite content of approximately 60 %,
when 004 vs. 009 (or 002/002 vs. 004/005) peak combination
is used (Fig. 6A). However, it must be noted that XRD pat-
terns of R1 ordered mixed-layer chlorite (0.60)/smectite cal-
culated with the NEWMOD program (Reynolds 1985; not
shown here) do not fit satisfactorily with experimental XRD
patterns. For example, (1) higher odd-order superstructure
reflections were not observed in the Mg-saturated, air-dried
patterns (particularly the 003 peak near 9.6 Å should be visi-
ble even in the presence of a small admixture of illite); (2)
the 006 and 007 reflections occurring in the 1621
o
2
Θ
CuK
α
range after ethylene glycol solvation are not resolved.
Hillier (1995) showed that any R1 chlorite/smectite always
has a mixed-layer (R0) chlorite/corrensite equivalent. Using
a partially segregated arrangement of component layers, he
was able to simulate an XRD pattern which still has a peak at
31 Å (EG) but other diffraction maxima are broad and shift-
ed. On the basis of these findings, Hillier (1995) concluded
that the shoulder at 31 Å is not related to any kind of order-
ing but rather to the presence of a 31 Å component layer (i.e.
corrensite). This might suggest that crystallites with partially
segregated packets of corrensite and chlorite layers are also
present in chloritic material from lithic graywackes, which
could be consistent with textural heterogeneity observed in
the SEM study. However, no attempt was made to simulate
such a type of structure.
The swelling properties of C/S are very similar to those
previously described for the saponite. Mg-saturated, air-dried
C/S shows basal reflections at ~29.4, 14.54, 7.26, 4.82, 3.60,
2.893 and 2.057 Å (Fig. 6A). Upon glycerol solvation of this
same sample, a structural contraction occurred, as indicated
Fig. 5. A XRD patterns of <2
µ
m size fractions of saponite oriented preparations in MgCl
2
solvated + air-dried, MgCl
2
+ glycerol sol-
vated, and MgCl
2
+ ethylene glycol solvated state; B 06,33 reflection of randomly oriented preparation. I illite, Srp serpentine, C
chlorite. d-spacings in Angströms.
TRIOCTAHEDRAL PHYLLOSILICATES: XRD, SEM AND EMPA STUDY 265
Fig. 7. XRD patterns of <2
µ
m size fractions of (A) saponite and (B) mixed-layer chlorite/smectite after K-saturation and heating to 550
o
C
for 1 hr. C/S mixed-layer chlorite/smectite, Sap saponite, I illite, Srp serpentine, C chlorite. d-spacings in Angströms.
Fig. 6. A XRD patterns of <2
µ
m size fractions of mixed-layer chlorite/smectite (C/S) oriented preparations in MgCl
2
solvated + air-
dried, MgCl
2
+ glycerol solvated, and MgCl
2
+ ethylene glycol solvated state; B XRD pattern of 06,33 reflection of randomly oriented
preparation. Sap saponite, I illite, Srp serpentine, C chlorite. d-spacings in Angströms.
266 BIROÒ, SOTÁK and BEBEJ
by the shift of reflections towards the higher degrees two-
theta (Fig. 6A). With ethylene glycol solvation Mg-saturated
C/S expands to ~31.5 Å. Glycol solvation also produces res-
olution of overlapping peaks. Similarly, heat treatment of K-
saturated samples at 550
o
C for one hour resulted in two se-
ries of reflections with 001 spacings of ~25 and ~10.2 Å,
respectively, the former representing collapsed C/S and the
latter collapsed saponite (Fig. 7B). This means, that charge
of expandable interlayers must be high enough to produces
both vermiculite-like and smectite-like swelling behaviour.
By analogy with the above described expansion properties of
discrete saponite, the expandable component of this mixed-
layer mineral can be tentatively categorized as high-charge
saponite.
Randomly orientated C/S-rich clay fractions display strong
06,33 reflection at 1.537 (±0.001) Å (Fig. 6B). However, this
value cannot be safely attributed to C/S because of substantial
admixture of saponite and probable peak interference.
Electron microprobe data
Authigenic phyllosilicates examined in this study, occur in
exceptionally well developed and densely packed aggregates
(pore-fillings, pseudomorphic replacements within lithic
grains). Therefore, microprobe analyses can be used to deter-
mine their compositional characteristics with reasonable
confidence. The following discussion assumes that total Fe is
in the Fe
2+
state.
Saponite. Representative chemical analyses of saponite are
given in Table 2. The structural formulae have 6.156.31 oc-
tahedral cations (i.e. Al
VI
+ Mg + Fe + Mn), which is consid-
erably higher than the theoretical trioctahedral limit of 6 cat-
ions per O
20
(OH)
4
. The high octahedral occupancy indicates,
that the smectite is either structurally interstratified with a
small percentage of chlorite component or intergrown with
another chloritic mineral. Considering the XRD data, a
mixed-layering can be eliminated (see X-ray diffraction sec-
tion). The existence of intimate intergrowths with chloritic
phase is difficult to detect, however, contamination of EMPA
should result in a positive correlation between total Al con-
tents and octahedral cation totals, which is not the case (r =
0.09; see also Fig. 9B). The second possible interpretation is
that the saponite contains a small amount of Mg ions located
in the interlayer position. This alternative was recently veri-
fied by Beaufort & Meunier (1994) as well as Schiffman &
Southard (1996) for saponites from metabasaltic rocks. Sapo-
nite from serpentinitic graywackes have formed in environ-
ments exceptionally rich in Mg (see Table 1), therefore, the
presence of the interlayer Mg is also highly probable. More-
over, assignment of the excess Mg (~0.20 cations/
O
20
(OH)
4
) to the exchangeable sites presented in Table 2 in-
creases the interlayer charge of the saponite. This approach
seems reasonable as the prevalence of high-charge layers
was deduced from the XRD results.
The saponites from the quartzolithic graywackes exhibit
only negligible compositional variations between samples.
Table 2: Representative electron microprobe analyses of saponite.
NOTE:
1
total Fe as FeO,
2
calculated from stoichiometry,
3
interlayer charge including Mg(IL)
Sample
258/V
258/V
258/V
K-I
K-I
K-I
255/II
255/II
255/II
256/TM1
Oxygen basis
O
20
(OH)
4
Habit
pore
pore
pore
pore
pore
pore
pore
pore
pore
P lg-replace
SiO
2
45.02
42.87
41.47
47.15
46.38
44.57
46.61
44.25
44.00
45.11
Al
2
O
3
5.80
5.795.65
6.48
6.17
5.73
6.195.78
5.88
6.57
MgO
25.09
23.41
23.39
24.95
23.60
23.71
24.81
23.01
23.94
23.36
FeO
1
2.86
3.31
2.30
4.494.28
4.06
4.32
3.82
4.49
5.48
MnO
0.07
0.08
0.02
0.00
0.00
0.00
0.00
0.00
0.00
0.00
CaO
2.12
2.51
1.48
1.66
1.791.80
1.36
1.491.68
1.88
Na
2
O
0.06
0.05
0.04
0.03
0.00
0.05
0.06
0.00
0.04
0.05
K
2
O
0.00
0.00
0.00
0.03
0.08
0.00
0.01
0.06
0.02
0.00
Total
81.02
78.02
74.35
84.79
82.30
79.92
83.36
78.41
80.05
82.45
Si
6.948
6.911
6.945
6.976
7.066
7.004
7.008
7.062
6.930
6.925
Al(IV)
1.053
1.089
1.055
1.024
0.934
0.996
0.992
0.938
1.070
1.075
Sum IV
8.000
8.000
8.000
8.000
8.000
8.000
8.000
8.000
8.000
8.000
Al(VI)
0.002
0.011
0.060
0.107
0.174
0.065
0.104
0.1490.021
0.114
Mg(VI)
5.620
5.5295.615
5.340
5.281
5.400
5.348
5.341
5.388
5.183
Fe
0.3690.446
0.322
0.556
0.545
0.534
0.543
0.510
0.59
1
0.704
Mn
0.0090.011
0.003
0.000
0.000
0.000
0.000
0.000
0.000
0.000
Sum VI
6.000
6.000
6.000
6.000
6.000
6.000
6.000
6.000
6.000
6.000
Mg(IL)
2
0.150
0.097
0.225
0.160
0.080
0.154
0.213
0.134
0.233
0.164
Ca
0.351
0.434
0.266
0.263
0.292
0.303
0.219
0.255
0.284
0.309
Na
0.018
0.016
0.013
0.0090.000
0.015
0.018
0.000
0.012
0.015
K
0.000
0.000
0.000
0.006
0.016
0.000
0.002
0.012
0.004
0.000
Sum IL
0.5190.546
0.504
0.438
0.388
0.472
0.451
0.401
0.533
0.488
ILC
3
1.020
1.076
0.994
0.860
0.760
0.929
0.883
0.789
1.049
0.961
Si/(Si+Al)
0.868
0.863
0.862
0.861
0.864
0.868
0.865
0.867
0.864
0.853
Mg/(Mg+Fe)
0.940
0.927
0.948
0.908
0.908
0.912
0.911
0.915
0.905
0.884
TRIOCTAHEDRAL PHYLLOSILICATES: XRD, SEM AND EMPA STUDY 267
The Si/(Si + Al) ratio is almost uniform, ranging from 0.85 to
0.87, while the total Mg/(Mg + Fe) ratio varies from 0.89 to
0.95 (Table 2). The average Al
VI
content ranges from 0.04 to
0.11 cations per O
20
(OH)
4
, which is in agreement with the
essentially trioctahedral character of these smectites indicat-
ed by XRD data. Sample 256/TM1 represents saponite co-
existing with the C/S in lithic graywackes. Its Mg/(Mg + Fe)
ratio of 0.88 is the lowest found in this study (Table 2).
The sum of interlayer cations (i.e. calculated MgIL+Ca+
Na+K) ranges between 0.35 and 0.55 cations/O
20
(OH)
4
. In
addition to the only presumed interlayer Mg, Ca is the most
abundant interlayer cation with very minor to negligible
amounts of Na and K (Table 2). Taking into account the ex-
changeable Mg, saponite shows a large variation of interlay-
er charge ranging from 0.69 to 1.08/O
20
(OH)
4
, with mode of
1.05/O
20
(OH)
4
. A large interval may reflect a heterogeneity
in layer charge distribution also inferred from the swelling
properties of this phase. The layer charge is derived solely
from tetrahedral substitution of Al for Si.
Mixed-layer chlorite/smectite. The C/S-rich sample 256/
TM1 was investigated in detail. Results of 42 spot analyses
normalized on the O
20
(OH)
10
basis (i.e. corrensite formula)
are presented as a series of histograms in Fig. 8. In addition,
selected analyses are shown in Table 3. In contrast to the
XRD data, composition of this phase suits the composition
of ideal corrensite fairly well. The number of cations in the
octahedral sites is between 8.70 and 9.18 per O
20
(OH)
10
(with mode 8.93/O
20
(OH)
10
), which is consistent with com-
bination of one chlorite-like layer (6 octahedral cations per
O
10
(OH)
8
) and one saponite-like layer (3 octahedral cations
per O
10
(OH)
2
). However, BSE imaging revealed textural as
well as chemical zoning of C/S aggregates. The composition-
al heterogeneity is also confirmed by the EMPA data. Two
compositions were detected in sample 256/TM1, as evident
from the bimodal distribution of major elements in the histo-
grams (Fig. 8). The first set of analyses was obtained from
fine-grained phyllosilicates occurring in the central parts of
pore-fillings as well as aggregates replacing detrital grains.
The second group of analyses represents platy and prolonged
crystallites, which form thin rims around the core material
(Figs. 3F and 4B). The composition of core C/S differs from
the composition of rim C/S by having a higher Si/(Si + Al)
ratio (0.72 vs. 0.68) and a higher Mg/(Mg + Fe) ratio (0.84
vs. 0.80; see also Fig. 9 open squares vs. filled squares
and Table 3). The opposite trend was observed by Shau et al.
(1990) in chloritic minerals in amygdaloidal metabasalts
from the northern Taiwan. Using EMPA and TEM/AEM
data, they observed that textural and chemical zonation of
aggregates also indicates structural changes in phyllosili-
cates. They concluded that increase in total Al content and
decrease in Mg/(Mg + Fe) ratio and Ca content from rim to
core of amygdales correlate with the increasing chlorite com-
ponent in the mixed-layered chlorite/smectite (or mixed-lay-
ered chlorite/corrensite). This may suggest the similar in-
Table 3: Selected electron microprobe analyses of mixed-layer chlorite/smectite.
NOTE:
1
total Fe as FeO;
2
calculated from stoichiometry
Sample
256/TM1
O xygen basis
O
20
(O H )
10
H abit
pore
pore
pore
pore
replace
replace
replace
replace
replace
replace
core
core
core
core
core
core
core
rim
rim
rim
SiO
2
38.77
39.60
39.14
38.80
38.85
36.78
39.91
36.53
36.81
37.55
A l
2
O
3
13.33
13.6913.40
13.0913.5914.4913.36
15.50
17.00
16.54
MgO
27.49
26.85
28.12
26.94
26.82
25.67
26.74
21.82
22.02
23.91
FeO
1
8.97
9.18
9.09
8.81
9.60
9.51
8.92
11.97
12.22
10.90
MnO
0.25
0.090.00
0.11
0.00
0.15
0.00
0.00
0.00
0.16
CaO
0.68
0.64
0.70
0.82
0.62
0.97
0.58
0.50
0.36
0.73
N a
2
O
0.00
0.00
0.13
0.01
0.06
0.08
0.04
0.14
0.00
0.09
K
2
O
0.00
0.00
0.00
0.00
0.00
0.03
0.00
0.01
0.03
0.00
Total
89.49
90.05
90.58
88.58
89.54
87.68
89.56
86.60
88.44
89.88
Si
6.4396.5196.422
6.49
96.455
6.270
6.589
6.349 6.255
6.254
A l(IV)
1.561
1.481
1.578
1.501
1.545
1.730
1.411
1.651
1.745
1.746
Sum IV
8.000
8.000
8.000
8.000
8.000
8.000
8.000
8.000
8.000
8.000
A l(VI)
1.048
1.175
1.013
1.083
1.117
1.181
1.188
1.525
1.660
1.501
Mg(VI)
6.671
6.5496.740
6.667
6.5496.442
6.579
5.654
5.578
5.9
37
Fe
1.246
1.264
1.247
1.234
1.334
1.356
1.232
1.740
1.737
1.518
Mn
0.035
0.013
0.000
0.016
0.000
0.022
0.000
0.000
0.000
0.023
Sum VI
9.000
9.000
9.000
9.000
9.000
9.000
9.000
8.930
8.974
8.978
Mg(IL)
2
0.135
0.040
0.1390.060
0.09
4
0.081
0.002
0.000
0.000
0.000
Ca
0.121
0.113
0.123
0.147
0.110
0.177
0.103
0.093
0.066
0.130
N a
0.000
0.000
0.041
0.003
0.0190.026
0.013
0.047
0.000
0.029
K
0.000
0.000
0.000
0.000
0.000
0.007
0.000
0.002
0.007
0.000
Sum IL
0.256
0.153
0.303
0.210
0.224
0.291
0.117
0.143
0.072
0.159
Si/(Si+A l)
0.712
0.711
0.713
0.716
0.708
0.683
0.717
0.648
0.648
0.658
Mg/(Mg+Fe)
0.845
0.8390.846
0.845
0.833
0.828
0.842
0.763
0.763
0.79
6
268 BIROÒ, SOTÁK and BEBEJ
Fig. 9. Composition of diagenetic phyllosilicates from the serpentinitic graywackes obtained from microprobe analyses: (A) Si vs. Al,
(B) octahedral occupancy vs. Al, (C) Mg/(Mg+Fe) vs. Si, (D) Ca+Na+K vs. Si. All recalculated on the O
20
(OH)
10
basis; total iron as
Fe
2+
. The compositions of ideal low-charge (L.c.) and high-charge (H.c.) saponite and corrensite, as well as the composition of cli-
nochlore, are given for reference (filled diamonds). Open circles saponite; open squares core C/S, filled squares rim C/S.
Fig. 8. Results of microprobe analyses of zoned C/S (sample 256/TM1) presented as histograms. Normalization on the basis of O
20
(OH)
10
.
Open columns core material, filled columns rim material.
TRIOCTAHEDRAL PHYLLOSILICATES: XRD, SEM AND EMPA STUDY 269
crease of the chlorite component in the rim material found in
lithic graywackes. Fig. 9A shows that the analysis points of
rim C/S are truly shifted towards the chlorite composition. In
addition, Beaufort et al. (1997) observed that the honeycomb
morphology of C/S diminishes and the resolution of individ-
ual crystals increase as these phases become more chloritic
(see also Wilson & Pittman 1977). Thus, a distinct platy hab-
itat of rim crystals may be regarded as a further (but indirect)
evidence in favour of this interpretation. On the other hand,
the rim C/S has octahedral occupancy comparable with core
C/S, ranging from 8.70 to 8.98/O
20
(OH)
10
(see also Figs. 8
and 9B). Furthermore, both types of phyllosilicates show ap-
proximately equal (average) sums of interlayer cations (i.e.
Ca+Na+K = 0.153 vs. 0.150/O
20
(OH)
10
; Fig. 9D). Consid-
ering all these facts, we are not able to identify with certainty
the real physical nature of the rim material. However, the
possible coexistence of chlorite-dominated phase along with
corrensite may explain the poorly defined XRD traces ob-
served in this study (Hillier 1995; Beaufort et al. 1997).
Compositional relationships between phyllosilicates. The
diagrams shown in Fig. 9 illustrate basic compositional dis-
tinctions between the studied phyllosilicates. To facilitate this
comparison, both saponite and C/S analyses were normalized
on the basis of the corrensite formula (i.e. 25 oxygens). In gen-
eral, the total Al content increases and the Si content decreases
from saponite to C/S, without compositional overlap (Fig.
9A). The saponite exhibits a higher total Mg/(Mg + Fe) ratio
(Fig. 9C) as well as higher content of interlayer cations than C/
S (Fig. 9D). In addition, Figs. 9A and 9B clearly illustrate that
the C/S (particularly core material) observed in lithic
graywackes is compositionally very close to a 50:50 mixture
of chlorite and trioctahedral smectite. The observed composi-
tional trend of increasing Al content and decreasing Mg/(Mg +
Fe) in the order of saponite to C/S is consistent with composi-
tional relations between these minerals in basalt alteration
parageneses (Shau et al. 1990; Shau & Peacor 1992).
Discussion
Origin of the phyllosilicate cement
In the graywackes from the ambron Zone, clay minerals
are clearly authigenic. There is no evidence, such as coatings
composed of clay platelets oriented tangentially to grain sur-
faces, fluid internal fabric and polymineralic composition of
aggregates or presence of fine-grained impurities, that the
phyllosilicates may have been formed by diagenetic recrys-
tallization of a pre-existing (syndepositional or mechanically
infiltrated) detrital clay matrix (Dickinson 1970; Wilson &
Pittman 1977; Moraes & Ros 1992). On the contrary, several
facts lead us to suggest that both saponite and C/S originated
by the interaction of the sediment with pore-fluids during
burial as a direct precipitates: (1) perpendicular alignment of
crystallites to detrital grain boundaries, indicating growth in
open pores; (2) the delicate habit of crystallites in the pre-
served open pore spaces; (3) the chemical and structural (?)
zonation of C/S aggregates, most probably reflecting a tem-
poral and spatial evolution of the pore-fluid chemistry (Shau
et al. 1990; Shau & Peacor 1992; Bautier et al. 1995); (4) the
monomineralic, either saponitic or corrensitic, composition
of the clay cement; and (5) in situ dissolution and replace-
ment of lithic fragments.
In the case of C/S, however, another feasible model of its
origin should be considered diagenetic alteration of an
earlier saponite precursor. Such a prograde conversion of tri-
octahedral smectite to chlorite through the ordered interstrat-
ified chlorite/smectite (corrensite) during burial diagenesis
has been recorded by a number of authors from various geo-
logical environments (see Introduction section). The process
seems to be strongly temperature dependent, but little is
known about the mechanisms governing the reaction path-
ways. A simple progressive interstratification of the brucitic
interlayers in the original saponite (i.e., solid-state transfor-
mation), proposed by Bodine & Madsen (1987), is unlikely.
A large compositional gap between saponite and C/S ob-
served in this study (Fig. 9) indicates that the 2:1 layers in
these two phases have completely different chemical compo-
sition. It means, that the structure and chemistry of the 2:1
layers in C/S could not be inherited from the crystalline
structure of the original smectite. Therefore, the crystalliza-
tion of C/S at the expense of saponite would require a disso-
lution/precipitation mechanism, as proposed by Meunier et
al. (1991). The presence of saponite in the C/S-rich
graywackes supports this idea. However, petrographic evi-
dence shows that in this type of graywackes saponite post-
dates the formation of C/S and from some reason it occurs
only as a replacing phase within the detrital feldspars. Fur-
thermore, the dissolution/precipitation mechanism should re-
sult in modification of early diagenetic textures of clay ce-
ment, which is not the case (see above). Considering these
facts, a direct precipitation of C/S from solution is suspected
rather than the smectite-to-corrensite conversion.
Factors controlling precipitation of phyllosilicates
The development of trioctahedral Mg-rich clay minerals is
undoubtedly related to abundant lithic material in the sand-
stones. Its gradual chemical dissolution may have released
plentiful supplies of Si, Al, Mg and Fe into the solution, so
that the activities of the required ions in porewaters were suf-
ficiently high for direct precipitation of clay minerals. Evi-
dence for this includes the large quantities of alteration tex-
tures and local metasomatic (?) dolomitization of sandstones.
However, the important question is: Why do sandstones lo-
cated within one small outcrop have either saponitic or cor-
rensitic cementation. It is obvious that both types of sand-
stones underwent the same burial history therefore the
temperature factor can be neglected. As described earlier,
however, there are significant differences in the modal com-
position of the host sandstones (Table 1). This suggests that a
different bulk rock composition, and consequently, a differ-
ent chemistry of pore-fluids played an important role during
authigenesis, being favourable either for saponite or C/S sta-
bility. As can be seen in Fig. 10, there is a very strong posi-
tive correlation (r = 0.96) between phyllosilicate Si/(Si + Al)
ratios and corresponding whole-rock values. It indicates, that
Al-rich C/S has formed only in the graywackes with suffi-
270 BIROÒ, SOTÁK and BEBEJ
Fig. 11. Mg/(Mg + Fe) ratios of phyllosilicates plotted against cor-
responding whole-rock values. Note the weak correlation between
these two variables. Structural formulae calculated on the
O
20
(OH)
10
basis. Circles saponite : quartzolithic graywackes;
squares C/S : lithic graywackes.
Fig. 10. Plot of phyllosilicate Si/(Si + Al) vs. whole-rock Si/
(Si + Al), demonstrating a strong positive correlation (r = 0.96)
between these two variables. Structural formulae calculated on the
O
20
(OH)
10
basis. Circles saponite : quartzolithic graywackes;
squares C/S : lithic graywackes.
cient Al contents. In other words, the amount of available Al
may have been a primary factor in controlling the precipita-
tion of phyllosilicates. The importance of the whole-rock Al
content for the formation of the trioctahedral phyllosilicates
in Mg-rich environments has been pointed out by Shau &
Peacor (1992), Hillier (1993) and Beaufort & Meunier
(1994). Experimental studies by Velde (1977) also imply that
the occurrence of the regularly interstratified (R1) phase in-
stead of smectite or chlorite may be controlled by the amount
of R
3+
component in the system (the Al
3+
or Fe
3+
content of
the assemblage). However, it should be noted that different
availability of Al cannot be simply attributed to the percent-
age of lithic grains in the individual sandstone beds because
they are mostly represented by the Al-poor serpentinites. As-
suming low mobility of Al-ions during diagenesis (Wintsch
& Kvale 1994), there must have been an other potential
source of Al, which was present in some sandstones at the
time of deposition. The role of feldspars is not considered
here, due to their accessory amounts in sediments. In addi-
tion, textural relationships suggest that the main phase of
feldspar dissolution post-dates formation of the pore-filling
phyllosilicates. However, the graywackes, in which C/S
dominates, are characterized by large domains with an exten-
sively altered detrital framework. It has been proposed that
they mostly represent replaced glassy fragments or mafic
minerals. As the equivalent replacement textures have not
been observed in the quartzolithic graywackes (with saponit-
ic cementation), it is possible to suggest that presence or ab-
sence of these highly unstable detrital components in original
sediments was crucial for their diagenetic evolution.
Several authors have also discussed an influence of bulk-
rock or fluid Mg/(Mg + Fe) ratio on the occurrence of smec-
tite, corrensite or chlorite (Almon et al. 1976; Velde 1977;
Brigatti & Poppi 1984; Shau et al. 1990). Results obtained
from different geological and rock environments infer that
saponite is favoured by a high bulk-rock or fluid Mg/
(Mg + Fe) ratio whereas corrensite (or chlorite) is favoured
to form in rocks with a lower Mg/(Mg + Fe) ratio (Brigatti &
Poppi 1984). The data presented here do not show such rela-
tionship. In fact, differences in whole-rock Mg/(Mg + Fe)
values between saponite- and C/S-rich graywackes are negli-
gible, ranging from 0.817 to 0.864, and 0.795 to 0.838, re-
spectively (Table 1). Furthermore, as shown in Fig. 11, there
is only weak correlation between phyllosilicate and whole-
rock Mg/(Mg + Fe) ratios. This implies that the composition
of saponite and C/S with respect to Mg/(Mg + Fe) ratio is
unique and independent of whole-rock chemistry. Shau et al.
(1990) have proposed that the contents of Fe (relative to Mg)
and Al
IV
(relative to Si) increase in the order of saponite, cor-
rensite, to chlorite as a consequence of minimization of the
misfit between octahedral and tetrahedral sheets within the
2:1 layers (i.e. due to unique crystal-chemical relations in
each mineral). The substitution of Al
IV
for Si causes an in-
crease in size of the tetrahedral sheets, which requires a sub-
stitution of the larger Fe
2+
ions for Mg
2+
in the octahedral
sheets in order to compensate for such changes. Our results
seem to offer further evidence in favour of this idea.
Acknowledgements: The authors are grateful to ¼ubica
Pukelová (Geological Institute of Slovak Academy of Sci-
ences) for XRF and XRD analyses. We also wish to express
special thanks to Pavol Siman and his staff at the Geological
Survey of Slovak Republic for their technical assistance in
the SEM/EMPA work. Financial support for this study was
provided by the Scientific Grant Agency (VEGA), grants
No. 2/4077/97 and No. 95/5305/520.
TRIOCTAHEDRAL PHYLLOSILICATES: XRD, SEM AND EMPA STUDY 271
References
Almon W.R., Fullerton L.B. & Davies D.K., 1976: Pore space re-
duction in Cretaceous sandstones through chemical precipita-
tion of clay minerals. J. Sed. Petrology, 46, 8996.
April R.H., 1981: Trioctahedral smectite and interstratified chlorite/
smectite in Jurassic strata of the Connecticut Valley. Clays and
Clay Miner., 29, 3139.
April R.H. & Keller D.M., 1992: Saponite and vermiculite in
amygdales of the Granby basaltic tuff, Connecticut Valley.
Clays and Clay Miner., 40, 2231.
Bailey S.W., 1980: Summary of recommendations of AIPEA no-
menclature committee. Clays and Clay Miner., 28, 7378.
Bautier M.D., Früh-Green G.L. & Karpoff A.M., 1995: Mechanism
of Mg-phyllosilicate formation in a hydrothermal system at a
sedimented ridge (Middle Valley, Juan de Fuca). Contr. Miner-
al. Petrology, 122, 134151.
Beaufort D. & Meunier A., 1994: Saponite, corrensite and chlorite-
saponite mixed-layer in the Sancerre-Couy deep drill-hole
(France). Clay Miner., 29, 4761.
Beaufort D., Baronnet A., Lanson B. & Meunier A., 1997: Corren-
site: A single phase or a mixed-layer phyllosilicate in the sapo-
nite-to-chlorite conversion series? A case study of
Sancerre-Couy deep drill hole (France). Amer. Mineralogist,
82, 109124.
Bettison L.A. & Schiffman P., 1988: Compositional and structural
variations of phyllosilicates from the Point Sal ophiolite, Cali-
fornia. Amer. Mineralogist, 73, 6276.
Bettison-Varga L.A., MacKinnon I.D.R. & Schiffman P., 1991: Inte-
grated TEM, XRD, and electron microprobe investigation of
mixed-layer chlorite/smectite from the Point Sal ophiolite, Cal-
ifornia. J. Metamorphic Geol., 9, 697710.
Bodine M.W. & Madsen B.M., 1987: Mixed-layer chlorite/smectites
from a Pennsylvanian evaporite cycle, Grand County, Utah.
Proc. Int. Clay Confer. 1985, Denver., Clay Miner. Soc.,
Bloomington, Indiana, 8596.
Brigatti M.F. & Poppi L., 1984: Crystal chemistry of corrensite: a
review. Clays and Clay Miner., 32, 391399.
Chang H.K., Mackenzie F.T. & Schoonmaker J., 1986: Comparisons
between the diagenesis of dioctahedral and trioctahedral smec-
tite, Brazilian offshore basins. Clays and Clay Miner., 34, 407
423.
De la Calle C. & Suquet H., 1988: Vermiculite. In: Bailey S.W.
(Ed.): Hydrous Phyllosilicates (exclusive of micas). Rev. in
Mineralogy, Vol. 19, Mineral. Soc. Amer., 455496.
Dickinson W.R., 1970: Interpreting detrital modes of graywacke
and arkose. J. Sed. Petrology, 40, 659707.
Fisher R.S., 1988: Clay minerals in evaporite host rocks, Palo Duro
Basin, Texas Panhandle, J. Sed. Petrology, 58, 836844.
Füchtbauer H., 1959: Zur Nomenklatur der Sedimentgesteine. Erdöl
u. Kohle, 12, 605613.
Galloway W.E., 1974: Deposition and diagenetic alteration of sand-
stone in northwest Pacific arc-related basins: Implications for
graywacke genesis. Geol. Soc. Amer. Bull., 85, 379390.
Helmond K.P. & van de Kamp P.C., 1984: Diagenetic mineralogy
and controls on albitization and laumontite formation in Paleo-
gene arkoses, Santa Ynez Mountains, California. In: Mc-
Donald D.A. & Surdam D.C. (Eds.): Clastic Diagenesis. Amer.
Assoc. Petrol. Geol. Mem., 37, 239276.
Hillier S., 1993: Origin, diagenesis, and mineralogy of chlorite min-
erals in Devonian lacustrine mudrocks, Orcadian Basin, Scot-
land. Clays and Clay Miner., 41, 240259.
Hillier S., 1995: Mafic phyllosilicates in low-grade metabasites.
Characterization using deconvolution analysis discussion.
Clay Miner., 30, 6773.
Iijima A. & Utada M., 1971: Present-day diagenesis of the Neogene
geosynclinal deposits in the Niigata oilfield, Japan. In: Gould
R.F. (Ed.): Molecular sieve zeolites-I (Advances in chemistry,
Series 101). Amer. Chem. Soc., Washington DC, 34349.
Inoue A., Utada M., Nagata H. & Watanabe T., 1984: Conversion of
trioctahedral smectite to interstratified chlorite/smectite in
Pliocene acidic pyroclastic sediments of the Ohyiu district, Ak-
ita Prefecture, Japan. Clay Sci., 6, 103116.
Inoue A. & Utada M., 1991: Smectite-to-chlorite transformation in
thermally metamorphosed volcanoclastistic rocks in the Kami-
kita area, northern Honshu, Japan. Amer. Mineralogist, 76,
628640.
Jackson M.L., 1975: Soil chemical analysis advanced course.
Madison, Wisconsin, 1386.
Janks J.S., Yusas M.R. & Hall C.M., 1992: Clay mineralogy of the
interbedded sandstone, dolomite, and anhydrite: The Permian
Yates Formation, Winkler County, Texas. In: Houseknecht
D.W. & Pittman E.D. (Eds.): Origin, diagenesis, and petro-
physics of clay minerals in sandstones. SEPM Spec. Publ., 47,
145157.
Kristmannsdóttir H., 1979: Alteration of basaltic rocks by hydro-
thermal activity at 100300
o
C. In: Mortland M.M. & Farmer
V.C. (Eds.): Proc. Int. Clay Confer. 1978, Elsevier, Amsterdam,
359367.
MacEwan D.M.C. & Wilson M.J., 1980: Interlayer and intercalation
complexes of clay minerals. In: Brindley G.W. & Brown G.
(Eds.): Crystal structures of clay minerals and their X-ray iden-
tification. Mineral. Soc., (London), 197249.
Malla P.B. & Douglas L.A., 1987: Identification of expanding layer
silicates: Layer charge vs. expansion properties. Proc. Int. Clay
Confer. 1985, Denver., Clay Miner. Soc., Bloomington, Indiana,
8596.
Masaryk P., ucha V. & Lintnerová O., 1995: Is the volcanic materi-
al present in the Middle Triassic basin sediments of the Hronic
Unit (Choè Nappe, Western Carpathians)? Geol. Carpathica,
46, 175181.
Meunier A., Inoue A. & Beaufort D., 1991: Chemographic analysis
of trioctahedral smectite-to-chlorite conversion series from the
Ohyu caldera, Japan. Clays and Clay Miner., 39, 409415.
Moraes M.A.S. & de Ros L.F., 1992: Depositional, infiltrated and
authigenic clays in fluvial sandstones of the Jurassic Sergi For-
mation, Recôncavo Basin, northeastern Brazil. In: House-
knecht D.W. & Pittman E.D. (Eds.): Origin, diagenesis, and
petrophysics of clay minerals in sandstones. SEPM Spec.
Publ., 47, 197208.
Reynolds R.C., 1980: Interstratified clay minerals. In: Brindley
G.W. & Brown G. (Eds.): Crystal Structures of Clay Minerals
and their X-ray Identification. Mineral. Soc., (London), 249
303.
Reynolds R.C., 1985: NEWMOD A computer program for the cal-
culation of the one-dimensional patterns of mixed-layered
clays. R.C. Reynolds, 8 Brook Rd. Hanover, New Hampshire.
Reynolds R.C., 1988: Mixed layer chlorite minerals. In: Bailey S.W.
(Ed.): Hydrous phyllosilicates (exclusive of micas). Rev. in
Mineralogy, Vol. 19, Miner. Soc. Amer., 601629.
Robinson D., Bevins R.E. & Rowbotham G., 1993: The character-
ization of mafic phyllosilicates in low-grade metabasalts from
eastern Greenland. Amer. Mineralogist, 78, 377390.
Schiffman P. & Fridleifsson G.O., 1991: The smectite-chlorite tran-
sition in drillhole NJ-15 Nesjavellir geothermal field, Iceland:
XRD, BSE, and electron microprobe investigations. J. Meta-
morphic Geol., 9, 679696.
Schiffman P. & Stautigel H., 1995: The smectite to chlorite transi-
tion in a fossil seamount hydrothermal system: The basement
complex of La Palma, Canary Islands. J. Metamorphic Geol.,
272 BIROÒ, SOTÁK and BEBEJ
13, 487498.
Schiffman P. & Southard R.J., 1996: Cation exchange capacity of
layer silicates and palagonitized glass in mafic volcanic rocks:
A comparative study of bulk extraction and in situ techniques.
Clays and Clay Miner., 44, 624634.
Schultz L.G., 1963: Clay minerals in the Triassic rocks of the Colo-
rado Plateau. U.S. Geol. Surv. Bull., 1147-C, 171.
Shau Y.H., Peacor D.R. & Essene E.J., 1990: Corrensite and mixed-
layer chlorite/corrensite in metabasalts from northern Taiwan:
TEM/AEM, EMPA, XRD, and optical studies. Contr. Mineral.
Petrology, 105, 123142.
Shau Y.H. & Peacor D.R., 1992: Phyllosilicates in hydrothermally
altered basalts from DSDP Hole 504B, Leg 83 a TEM and
AEM study. Contr. Mineral. Petrology, 112, 119133.
Soták J. & Bebej J., 1996: Serpentinitic sandstones from the am-
bron-Kamenica Zone in Eastern Slovakia: Evidence of deposi-
tion in a Tertiary collisional belt. Geol. Carpathica, 47,
227239.
Velde B., 1977: Clays and clay minerals in natural and synthetic
systems. Elsevier, Amsterdam, 1211.
Wilson M.D. & Pittman E.D., 1977: Authigenic clays in sandstones:
Recognition and influence on reservoir properties and paleoen-
vironmental analysis. J. Sed. Petrology, 47, 331.
Wintsch R.P. & Kvale C.M., 1994: Differential mobility of elements
in burial diagenesis of siliciclastic rocks. J. Sed. Res., A64,
349361.