Drits Etal - 2002
Drits Etal - 2002
Drits Etal - 2002
V I C T OR A. D R IT S 1 , H OL G E R L I ND G R E EN 2 , * , B O R IS A. S A K HA R O V 1 , H A N S J ØR G E N J A K OB S E N 3 ,
A L F R ED L . S A L YN 1 A N D L I DI A G . D A I NY A K 1
1
Institute of Geology, Russian Academy of Science, Pyzhevsky per D.7, 109017 Moscow, Russia
2
Clay Mineralogical Laboratory, Geological Survey of Denmark and Greenland, Thoravej 8, DK2400
Copenhagen NV, Denmark
3
Instrument Centre for Solid-State NMR Spectroscopy, Department of Chemistry, University of Aarhus,
DK8000 Aarhus C, Denmark
Abstract—Illite-smectite (I-S) minerals isolated from Upper Jurassic oil-source rock shales from
Denmark and the North Sea have been investigated by X-ray diffraction, thermal analysis, infrared,
Mössbauer, and solid-state nuclear magnetic resonance spectroscopies and chemical analysis. Detailed
structures have been determined in order to reveal the diagenetic transformation mechanism in these
shales. Generally, in oil-source rocks of sedimentary basins, oil generation takes place simultaneously with
the diagenetic transformation of I-S. We demonstrate a link between the two reactions: NH3 released from
kerogen during maximum oil generation is fixed as NH+4 in the NH4-bearing mica or tobelite layers formed
from smectite or vermiculite layers in I-S, in a diagenetic interval which we name the ‘tobelitization
window’. Due to this solid-state transformation, mixed-layer structures have been formed consisting of
interstratified illite, tobelite, smectite and vermiculite layers (I-T-S-V) and having maximum ordering of
illite + tobelite and smectite layers for R = 1. The tobelitization of smectite in I-S is probably typical for all
oil-source rock shales.
Key Words—Ammonium, Diagenesis, Illite-smectite, Jurassic, North Sea, Oil-Source Rock, Shales.
formed to I-S having the maximum degree of ordering Upper Jurassic shales which is the most important oil-
possible (R = 1). Lindgreen (1994) found for the same source rock in the North Sea (Thomsen et al., 1983). For
rocks that with increasing depth of sample, the ratio comparison, samples from the immature Upper Jurassic
NH4/K for soluble cations in the porefluids of core shales in the Danish Subbasin of the Norwegian–Danish
samples increased as did the same ratio for fixed cations Basin are included. Comparison of the results obtained
in interlayers of the I-S structures and concluded that by these methods with those obtained previously by
NH4 fixation may be a general process during burial Lindgreen (1994), Drits et al. (1997a,b) and Lindgreen et
diagenesis in oil-source rocks. al. (2000) has provided a new insight into specific
Structural study of I-S in shales and especially in oil- features of the structural mechanism of I-S transforma-
source shales is difficult because clay fractions extracted tion during oil generation as well as into the origin and
from shale samples are typically a mixture of detrital and parent source of detrital material.
authigenic minerals and, moreover, the complex struc-
ture of I-S in shales cannot be determined by conven- SAMPLE MATERIALS
tional XRD techniques. A new methodological complex
consisting of three methods has been developed and Cuttings (hand-picked) and core samples of shales of
tested on several representative samples of the Upper Upper Jurassic (Kimmeridgian-Volgian-Ryazanian) age
Jurassic and Cambrian alum black oil-source shales from the Mandal, Farsund and Haugesund Formations in
(Drits et al., 1997a; Sakharov et al., 1999; Lindgreen et the Central Trough, North Sea, and from the Børglum,
al., 2000). The first, the multi-specimen method, Sauda and Tau Formations in the Norwegian–Danish
determines the detailed structural and probability para- Basin (including the Danish Subbasin) have been
meters of the interstratification in multi-component studied. The cuttings include the Danish wells M8, U1
mixed-layer minerals (Drits et al., 1997b; Sakharov et and I1 and the Norwegian wells 2/7-3 and 2/11-1; core
al., 1999). In contrast to most published simulations of samples came from wells E1, and 2/11-1 as well as from
XRD patterns for I-S models, in this method, agreement Danish wells East Rosa 3/3A, SE Igor 1, N. Jens 1, Jens
between calculated and experimental XRD patterns is 1, Iris 1, Jeppe 1 and Elin 1. Studied samples from the
obtained not only for positions, but also for intensity and Danish Subbasin include core samples from
profiles of the reflections. Furthermore, such agreement Frederikshavn 1, Uglev 1, Børglum 1 and Haldager 1
was obtained between one structural model and the XRD (Figure 1). The samples investigated represent the
patterns of different specimens prepared from each interval from onset to maximum oil generation. Details
sample subjected to different treatments (saturation by of the samples and their location were previously given
different interlayer cations, glycolation, etc.). This by Hansen and Lindgreen (1989) and Lindgreen and
method revealed the sequential structure transformation Hansen (1991).
of illite-smectite-vermiculite during diagenesis of the
North Sea Upper Jurassic oil-source shales to be a solid- METHODS
state transformation of smectite interlayers into mica
(Drits et al., 1997b). The results of this work confirmed Chemical pretreatment
the conclusions of Lindgreen (1991). The second method The samples were pretreated by removal of organic
determined: (1) the average content of NH4; and (2) the matter and Fe and Al oxides, and the mixed-layer
distribution patterns of interlayer K and NH4 in NH4- fraction was isolated by centrifugation (Hansen and
bearing I-S structure (Drits et al., 1997a). Application of Lindgreen, 1989) yielding particles predominantly
this method showed that NH4-bearing I-S from the North <500 Å in diameter and <100 Å thick (Lindgreen et
Sea and Cambrian Alum shales from the Baltic area al., 1992). By this method, detrital illite was removed.
consist of irregularly interstratified K-bearing 9.98 Å The mixed-layer fractions were saturated with K+ and
illite, NH4-bearing 10.33 Å tobelite and expandable Mg2+ by repeated washings using chloride solutions,
layers (Drits et al., 1997a; Lindgreen et al., 2000). The excess chloride being removed by washing with water.
third method provides semiquantitative determination of Oriented specimens were prepared by the pipette method
trans-vacant (tv) and cis-vacant (cv) 2:1 layers in I-S by using 2.5 mg/cm2 of specimen and analyzed air-dried
thermal analysis (Drits et al., 1998b). It is based on the and after glycolation for three days at 608C in glycol
fact that smectites and illites consisting of cv 2:1 layers vapor. Potassium-saturated samples were heated at
have dehydroxylation temperatures 1008 to 2008C higher 1508C to contract the expandable layer thickness to
than those consisting of tv 2:1 layers (Drits et al., 1995). 9.98 Å and the patterns were recorded immediately
In contrast to XRD methods, this technique can be afterwards, heating being repeated for 5 min before the
applied to I-S having a turbostratic structure, such I-S scan of each peak.
being widespread in shales.
In this work simulation of the experimental XRD X-ray diffraction
patterns, thermal analysis and IR spectroscopy have been The XRD patterns were obtained using the Philips
applied for a large number of samples from the North Sea PW3040 and PW1050 (samples investigated in
84 Drits et al. Clays and Clay Minerals
Figure 1. Locations of wells investigated in the Central Trough and the Norwegian–Danish Basin.
Denmark) and the DRON-4 (samples investigated in respectively (Drits et al., 1997a,b; Sakharov et al., 1999;
Russia) diffractometers, employing CoKa and CuKa Lindgreen et al., 2000). The K, NH4 and Na cations were
radiation. The PW 3040 instrument has a curved graphite placed in the center of mica interlayers. The thicknesses
monochromator and a fine-focus tube. For complete of the illite and tobelite layers in I-T-S-V were found to
XRD patterns of oriented specimens in the interval be 9.98 and 10.33 10.35 Å, respectively (Drits et al.,
1.5 6582y, U8 fixed divergence and anti-scatter slits 1997a) whereas for I-S-V the thickness of illite-like
were used and intensities were measured for 10 s per layers was variable. The thicknesses of coherent
0.182y step. The DRON-4 instrument has a horizontal scattering domains (CSDs) were distributed log-nor-
goniometer supplied with a graphite monochromator mally. The parameters of this distribution were deter-
which has a set of fine anti-scatter slits (0.1 0.25 mm) mined using a mean thickness of CSDs and the
together with Soller slits which have an angular aperture regression given by Drits et al. (1997c) having mean
of 1.58. Intensities were measured for 100 s per 0.0582y and maximum thicknesses of CSDs as variable
step. Fitting of the XRD patterns within the 2.5 5582y parameters.
region was carried out according to the procedure
described by Drits et al. (1997b) and Sakharov et al. Thermal analysis
(1999). In the following, I, T, S and V denote illite, A Stanton Redcroft DTA 673-674 with gas outlet to
tobelite, smectite and vermiculite-like interlayers or non-dispersive infrared (IR) H2O and CO2 detectors was
layers, respectively, and I-S-V and I-T-S-V denote used to determine quantitatively the amount of structural
mixed-layer illite-smectite-vermiculite and illite-tobe- water released during heating (evolved water analysis,
lite-smectite-vermiculite, respectively. We define EWA), (Morgan, 1977). The main advantage of this
expandable layers as smectite and vermiculite if in the technique is that the loss of hydroxyls is recorded
glycolated state their interlayers contain two or one separately and that the amount of structural water
layers of glycol molecules, respectively, independent of released during dehydroxylation can be determined
the nature of the exchangeable cations. Layers with quantitatively and with high sensitivity.
swelling properties which depend on the exchangeable Sodium-saturated samples were heated at 58C/min in
cations are referred to as ‘low-charge vermiculite or a gas flow of 300 ml N2 per min. In order to separate
high-charge smectite’ layers (V’-layers). For I-S-V the K dehydration and dehydroxylation processes, the samples
content in the illite-like interlayers was 0.75 0.5 atoms were heated first to 2008C and the temperature was then
per O10(OH)2 for different samples and decreases with kept constant until the water release had stopped. The
the degree of diagenetic transformation of the samples. samples were then heated in steps to 10008C at 58C/min
For I-T-S-V the K and NH4 contents in illite and tobelite for each step, constant temperature being kept at
interlayers were 0.75 and 1.0 cations per O10(OH)2, dehydroxylation peaks.
Vol. 50, No. 1, 2002 Tobelitization of I-S in oil-source shales 85
the difference in the layer thickness can be ignored if proportions of I, S and V interlayers vary from 65 73,
only the 001 reflection is used. The validity of this 20 25 and 0.05 0.07, respectively (Figure 2, Table 1).
conclusion follows from the calculation of Drits et al. As follows from the conjunction probability parameters
(1997) made for interstratified I-T. Środoń et al. (1992) pij (i, j = I, S, V) (Figure 2) the layer-type interstrati-
showed that the mean number of layers in CSDs, N, and fication has a significant tendency to segregation since
the number of expandable interlayers, wS+V, is related to Pii > wi (i = I, S, V). The Mg-saturated and glycolated
the mean number of 2:1 layers in illite fundamental I-S-V of group 2 (samples x1, x5, x6, x7, x12, x13, x17,
particles, n̄, i.e. the number of 2:1 layers whose 79, 87) contain 0.78 0.86 of illite, 0.10 0.14 of
interlayers contain fixed K and NH4, by the equation smectite and 0.04 0.08 of vermiculite interlayers
(Figure 3, Table 1). The illite and smectite layers in
n̄ = N/[(N 1)wS+V + 1] (4)
this phase are distributed with the maximum degree of
ordering for R = 1 because PSS = PSV = 0. However,
Atomic force microscopy vermiculite layers preserve their tendency to be segre-
Atomic force microscopy (AFM) was carried out on a
Rasterscope 4000 instrument using the non-contact mode.
The samples were dispersed by ultrasonic treatment in 112 Mg-glycol
distilled water and a drop was left to dry on a block of 17.7 Å I S V
highly-oriented pyrolytic graphite. The specimens were h 9.98 Å 17.05 Å 14.2 Å
w 0.68 0.27 0.05
scanned at room conditions at a speed of 500 nm/s. I S V
I 0.772 0.204 0.024
10.0 Å S 0.500 0.450 0.050
RESULTS 3.364 Å
7.23 Å V 0.400 0.200 0.400
3.562 Å
Simulation of the experimental XRD patterns 5.46 Å
5.03 Å 2.380 Å
2.001 Å
The XRD patterns of the samples can be divided into
three groups. Glycolated group 1 samples have XRD 0 10 20 30 40 50 60
patterns with rather intense 17 Å reflections whose
resolution from the low-angle side varies for different x3 Mg-glycol
samples. This group includes shale samples from the I S V
h 9.99 Å 17.00 Å 12.9 Å
Danish Subbasin and from shallow depth in the Central 17.7 Å w 0.75 0.20 0.05
Trough of the North Sea. Group 2 samples have two I S V
I 0.827 0.140 0.033
distinct reflections in the low-angle diffraction region
S 0.600 0.400 0.000
with 12.8 13.2 Å and 9.8 10 Å spacings, respectively 10.0 Å V 0.200 0.300 0.500
and include most samples from the Central Trough 5.36 Å 3.344 Å
7.21 Å 5.00 Å
shales taken from depths of >3000 3500 m. Group 3 3.562 Å 2.000 Å
samples have XRD patterns with a rather sharp and
intense 10 Å reflection and a broad low-angle shoulder 0 10 20 30 40 50 60
and include some of the samples from depths >3500 m.
83 Mg-glycol
For each group, one sample was studied in detail by
I S V
simulation of the experimental XRD patterns obtained 17.9 Å h 9.98 Å 16.95 Å 14.2 Å
for several specimens of the same sample subjected to w 0.66 0.31 0.03
different treatments (Drits et al., 1997b). Each sample is I S V
I 0.710 0.268 0.023
a mixture of two mixed-layer phases. The predominant S 0.570 0.430 0
phase (>90%) has alternating illite, smectite and V 0.500 0 0.500
10.1 Å
vermiculite interlayers (I-S-V) separated by Al-rich 2:1 3.348 Å
5.40 Å
layers. The second phase is a kaolinite-illite-vermiculite 7.24 Å 5.04 Å 3.57 Å 2.003 Å
(K-I-V) containing 90% kaolinite layers. In this work,
simulation of the experimental XRD patterns of the 0 10 20 30 40 50 60
samples was carried out using structural and statistical ˚2θ
parameters as initial data determined for representative Figure 2. X-ray diffraction patterns of specimens of samples
members. The main aim was to determine the most 112, x3 and 83 of group 1 fitted with three components (I, S, V).
reliable content of the layer types in the I-S-V in the For each specimen, experimental (thin line, with d-values) and
sample collection. Figures 2 5 show simulated and simulated patterns (shaded line) are shown, together with
experimental XRD patterns and structural and prob- thicknesses (h) and contents (w i ) of the layer types (I: illite; S:
smectite; and V: vermiculite layers) and the matrix for the
ability parameters providing the best agreement between
junction probability, p i,j , for layer j to follow layer i (i, j = I,S,V).
the XRD patterns compared. The reflections with spacings of 7.2 Å and 3.56 Å correspond to
In the Mg-saturated and glycolated I-S-V phase of a kaolinitic phase. Sample 83 contains a small amount (11%) of a
group 1 (samples x3, 80, 82, 83, 92, 95, 99, 112), high-illitic I-S-V. CoKa radiation.
Vol. 50, No. 1, 2002 Tobelitization of I-S in oil-source shales 87
0 10 20 30 40 50 60
0 10 20 30 40 50 60
87 Mg-glycol
x5 Mg-glycol
I S V
I S V 13.0 Å
h 10.03 Å 17.10 Å 12.90 Å
h 10.03 Å 17.1 Å 13.3 Å 10.06 Å w 0.80 0.12 0.08
3.361 Å
w 0.82 0.09 0.09 I S V
I S V I 0.810 0.130 0.060
12.7 Å I 0.841 0.110 0.049
10.1 Å S 1 0 0
3.351 Å S 1 0 0 V 0.400 0.200 0.400
5.05 Å
V 0.450 0 0.550 7.23 Å 3.361 Å
5.05 Å 2.000 Å
x17 Mg-glycol
I S V
h 10.03 Å 17.05 Å 13.2 Å
w 0.78 0.10 0.12
12.7 Å
10.1 Å I S V
3.349 Å I 0.795 0.128 0.077
S 1 0 0
V 0.500 0 0.500
7.25 Å 5.05 Å
3.58 Å
2.003 Å
0 10 20 30 40 50 60
Figure 3. X-ray diffraction patterns of specimens of samples x1,
x12 Mg-glycol
x5, x17, x12 (left), and x13 and 87 (above) of group 2 fitted with
three components (I, S, V). Explanation and symbols as for
I S V
h 10.08 Å 17.0 Å 13.3 Å
Figure 2. The reflections with spacings of ~7.2 Å and ~3.5 Å
w 0.82 0.09 0.09 correspond to a kaolinitic phase. CoKa radiation except for
12.6 Å I S V sample 87: CuKa radiation.
10.2 Å I 0.835 0.110 0.055
S 1 0 0
3.370 Å V 0.500 0 0.500
7.29 Å 5.13 Å
3.59 Å
2.017 Å
0 10 20 30 40 50 60
˚2θ
gated (WV < PVV). It is remarkable that the thickness of vermiculite layers in the actual I-S-V structure lead to
illite layers for the I-S-V increases up to 10.03 Å overestimation of the content of illite layers. Note that
(Figure 3) in comparison with 9.98 Å typical for illites transition from group 1 to group 3 is accompanied by a
(Moore and Reynolds, 1989). Group 3 I-T-S-V patterns decrease in the amount of K in illite-like interlayers from
(x14, x15, 86, 89, 101) are different mainly because the 0.75 to 0.5 atoms per O10(OH)2. Both the increase in the
content of illite layers increases (0.84 0.86) and the illite-layer thickness and the decrease in the amount of K
illite layer thickness increases to 10.05 Å (Figure 4). indicate the presence of NH4 in illite-like interlayers.
Comparison of the amount of non-expandable illite The I-S-V are, in fact, the mixed-layer four-component
layers in I-S-V of the studied samples (Table 1) with system I-T-S-V (Drits et al., 1997b), because the non-
those presented by Lindgreen et al. (1991, Table 1) expandable layers in the mixed-layer phases are 9.98 Å
shows that for some samples the proportion of illite illite and 10.33 10.35 Å tobelite layers. Figure 5
layers determined by Lindgreen et al. is greater mainly compares the experimental XRD patterns with those
because they simulated XRD patterns in terms of a two- calculated for I-T-S-V models. It is remarkable that the
component I-S model and therefore the presence of 13 Å amounts of tobelite layers determined by the simulation
88 Drits et al. Clays and Clay Minerals
86 Mg-glycol
11.5 Å
87 Mg air-dried
I S V I T S V
10.14 Å h 10.04 Å 17.2 Å 13.5 Å h 9.98 Å 10.35 Å 14.00 Å 12.50 Å
w 0.84 0.06 0.10 w 0.65 0.15 0.12 0.08
I S V I T S V
3.347 Å I 0.851 0.071 0.077 I 0.645 0.19 0.09 0.075
S 1 0 0 T 0.70 0 0.30 0
V 0.650 0 0.350 4.96 Å S 0.90 0.10 0 0
5.01 Å 3.355 Å
7.16 Å
V 0.225 0.175 0.20 0.40
3.572 Å
2.000 Å 2.007 Å
0 10 20 30 40 50 60
0 10 20 30 40 50 60 70
89 Mg-glycol 13.0 Å
87 Mg-glycol
10.14 Å I S V 3.361 Å I T S V
3.361 Å
h 10.05 Å 17.1 Å 12.9 Å h 9.98 Å 10.35 Å 17.10 Å 13.00 Å
10.06 Å
w 0.84 0.08 0.08 w 0.65 0.15 0.12 0.08
I S V I T S V
I 0.857 0.095 0.048 I 0.645 0.19 0.09 0.075
S 1 0 0 5.05 Å T 0.70 0 0.30 0
5.05 Å V 0.500 0 0.500 S 0.90 0.10 0 0
V 0.225 0.175 0.20 0.40
7.24 Å 3.58 Å
2.012 Å 2.000 Å
0 10 20 30 40 50
0 10 20 30 40 50 60
˚2θ 89 Mg-glycol
Figure 4. X-ray diffraction patterns of specimens of samples 86 10.14 Å 3.361 Å I T S V
h 9.98 Å 10.35 Å 17.1 Å 12.9 Å
and 89 of group 3 fitted with three components (I, S, V).
w 0.65 0.19 0.08 0.08
Explanation and symbols as for Figure 2. The reflections with
I T S V
spacings of ~7.2 Å and 3.5 Å correspond to a kaolinitic phase. I 0.645 0.229 0.071 0.056
For sample 86: CoKa radiation, and for sample 89: CuKa 5.05 Å T 0.717 0.084 0.179 0.020
radiation. S 0.900 0.100 0 0
3.58 Å V 0.284 0.216 0 0.500
2.012 Å
coincides with those determined independently for the
same samples using the technique of Drits et al. (1997a). 0 10 20 30 40 50 60
For example, both the simulation (Figure 5) and the °2θ
2θ
basal reflection width ratios (Table 4 in Drits et al.
(1997a)) gave the same content of tobelite layers in I-T- Figure 5. X-ray diffraction patterns of specimens of samples 87
S-V phase, 0.15 and 0.19 for samples 87 and 89, of group 2 and 89 of group 3 fitted with four components (I, T, S,
V). For sample 87, the diffraction patterns for both the air-dried
respectively. and the glycolated specimens are shown. Explanation and
The conjunction probability parameters for the three- symbols as for Figure 2. The patterns from two different Mg-
and four-component models for samples 87 and 89 are saturated specimens, one air-dried (CoKa radiation) and one
shown by the matrices presented in Figures 3, 4 and 5. intercalated with ethylene glycol (CuKa radiation), demonstrate
From these matrices, the layer pair probabilities wij = that one structural model fits all the totally different patterns
wipij (i, j = I, T, S, V) in the four-component model and obtained by different treatments of one sample. The reflections
with spacings of 7.23 Å and 3.55 Å correspond to a kaolinitic
the probabilities wij = wipij (i, j = I, S, V) of the three- phase. 87 Mg air-dried: CoKa radiation, 87 Mg-glycol and 89
component model are calculated and new matrices Mg-glycol: CuKa radiation.
consisting of wij may be written as:
I T S V I S V
I 0.419 0.124 0.059 0.048 I 0.648 0.104 0.048
T 0.105 0 0.045 0 > S 0.12 0 0 sample 87
S 0.108 0.012 0 0 V 0.032 0.016 0.032
V 0.018 0.014 0.016 0.032
I T S V I S V
I 0.411 0.149 0.046 0.044 I 0.712 0.08 0.048
T 0.136 0.016 0.034 0.004 > S 0.08 0 0 sample 89
S 0.072 0.008 0 0 V 0.048 0 0.032
V 0.031 0.017 0 0.032
Vol. 50, No. 1, 2002 Tobelitization of I-S in oil-source shales 89
Table 1. Number of the layer types in I-T-S-V determined from simulation of XRD patterns.
Sample Well Depth (m) wI+T1 wT2 wI3 wS+V4 w S5 wV6
Group 1
x3 SE Igor 1 2502 0.75 0.08 0.67 0.25 0.20 0.05
80 U1 2542 0.70 0.05 0.65 0.30 . .
1
Number of non-expandable interlayers
2
Number of tobelite interlayers
3
Number of illite interlayers
4
Number of expandable interlayers
5
Number of smectite interlayers
6
Number of vermiculite interlayers
For both samples the parameter sums wII + wIT + wTI + (Vedder, 1965; Yamamoto and Nakahira, 1966;
wTT, wIS + wTS, wIV + wTV, wSI + wST and wVI + wVT for Karyakin et al., 1973; Higashi, 1978; Juster et al.,
the four-component I-T-S-V model are equal to the sums 1987). For group 1, the spectra either have rather weak
wII, wIS, wIV, wSI and wVI for the three-component I-S-V or no maxima in this region.
model. Note that for the four-component models of both For the IR spectrum of sample 89, the decomposition
samples wIT & wTI, wTS > wST, whereas the wTT, wTV, in the region of the OH-stretching vibration into
wSS, wSV and wVS are either very low or equal to zero. It individual bands of the OH, H and NH4 vibrations is
should be noted that simulation of the experimental XRD shown in Figure 8. The areas of the bands corresponding
patterns was performed for the I-S-V and I-T-S-V without to the OH and NH4 (at 1430 cm 1) vibrations (Figures 7
illite as an individual phase. and 8) were transformed to optical densities of IOH and
INH4, respectively. The INH4/IOH values and the amount
Thermally evolved water of NH4 taken from Table 1 in Lindgreen (1994) and
The step-heating EWA curves (Figure 6) of most from Table 4 in Drits et al. (1997a) are almost the same
studied samples are quite similar and have one strong (Table 3) showing that the technique is accurate for
and broad dehydroxylation maximum at 510 5508C. determination of NH4 in the NH4-bearing mixed-layer
However, for some samples (x3, x5, 89) the step-heating minerals.
curves contain additional maxima at temperatures
>6008C. The relative area of the high-temperature
dehydroxylation calculated as the ratio of this area to Table 2. EWA analysis, stepwise heating.
the sum of the high- and low-temperature maxima was
Sample T(8C) Area (%) T(8C) Area (%)
<0.04 for most samples, but was 0.15 for some of them
(Table 2). x3 510 85 600 15
x1 503 100
Fixed NH4 content x5 506 84 596 16
x13 496 100
Infrared spectra of samples from groups 2 and 3 have x14 503 100
an intense absorption band near 1430 cm 1 (Figure 7) 89 500 96 620 4
corresponding to NH4 absorption maxima in tobelites
90 Drits et al. Clays and Clay Minerals
900 250
800
8C
600
x5
˚C
500 100
400 596˚C Evolved
water 50
300
200 0
0
Time (s) 10000
900 300
510˚C
600 150
500 Evolved 100
620˚C
water
400 50
300 0
0 Time (s) 8000
Figure 6. Evolved water curves obtained by step-wise heating for sample x5 of group 2 and sample 89 of group 3.
Mössbauer spectra decomposition Lindgreen et al. (1991) and following the procedure
The Mössbauer spectra have been decomposed into described above, we have calculated the average
two doublets for Fe 3+ and Fe2+, respectively. Parameters formulae for the cuttings samples and the proportions
of the inner doublets are reflected in visual splitting of of I-S-V and kaolinite (Table 4). The fact that Si/IVAl
the spectra and vary from 0.58 to 0.63 mm/s, typical of and IV Al/ VI Al values calculated from the average
Fe3+-bearing dioctahedral layer silicates (Johnston and formulae and measured by NMR almost coincide
Cardile, 1987). The ratios of areas corresponding to the (Table 4) shows that both techniques provide reliable
Fe3+ and Fe2+ doublets are used for calculation of the results.
I-T-S-V structural formulae. Table 5 contains the structural formulae of I-T-S-V
calculated from equations 2 and 3. In all samples the I-T-
Average structural formulae for the I-T-S-V S-V contain almost the same amount of fixed interlayer K
For the cuttings samples, Lindgreen et al. (1991)
determined total chemical composition, Si/IV Al and
IV Table 3. Relationship between INH4/IOH and NH4 content in
Al/VIAl values using AAS and NMR methods but not
the I-T-S-V.
accurate structural formulae for the I-S-V phase because
of the kaolinitic phase content. Using the data of Sample INH4/IOH1 cNH42 cNH43 wT4 2cNH45
Group 1
0 x3 0.06 0.07 0.10 0.08 0.08
82 0.07 0.08 0.07 .
0.07
2 95 0.04 0.06 0.08 0.06
89 .
4 Group 2
x1 0.04 0.03 0.03 0.00 0.03
Absorption
6 x13
x12 0.20 0.20 0.22 0.19 0.20
8 x13 0.21 0.21 0.22 0.19 0.21
x17 0.08 0.10 0.10 0.08 0.09
10
x1 87 0.18 .
0.15 0.15 0.16
12 Group 3
1630 89 0.18 0.22 0.18 0.17 0.19
14 1430 x17
1
Ratio between IR optical densities of the NH4 band at
16 1430 cm 1 and the OH band at 3631 cm 1
1800 1700 1600 1500 1400 1300 1200 2
Data by the isotopic dilution method (Lindgreen, 1991)
3
cm -1 Data from 005 reflection positions of K-saturated and
dehydrated I-T-S-V (Table 4, Drits et al., 1997b)
4
Figure 7. Sections of the IR spectra containing the NH 4 Data from the FWHH(005)/FWHH(002) values (Table 4,
(1410 1440 cm 1 ) band for K-saturated and heated (1508C) Drits et al., 1997b)
5
samples x1, x13 and x17 of group 2 and 89 of group 3. Average of the four cNH4 determinations
Vol. 50, No. 1, 2002 Tobelitization of I-S in oil-source shales 91
0
analysis of K-saturated and dehydrated samples and the
2852 mean number of 2:1 layers in illite fundamental particles
1
3573 3048 (n̄) in the I-T-S-V calculated from equation 4. With a
3699 3153
2 3468 decrease of the expandable interlayers, the values of N
Absorption
3307
3622 increase from 5.3 to 6.4 layers and the values of n̄ from
3
3657 3622 2.2 to 3.4 layers (Table 6).
4
DISCUSSION
5
NH4- and K-bearing non-expandable interlayers
6 For I-S a quantitative determination of content and
3800 3600 3400 3200 3000 2800 2600
distribution of NH4 by diffraction methods is difficult
cm-1 because different proportions and distributions of layer
types modify the positions, intensity and profile of basal
Figure 8. Decomposition into individual bands for sample 89 (K- reflections more strongly than the presence of NH4.
saturated and heated to 1508C) of the complex maximum of the However, diffraction sensitivity to structural NH 4
IR spectrum in the stretching region. The bands at 3657, 3622
increases significantly if the samples are saturated with
and 3573 cm 1 correspond to OH, those at at 3468 and
3153 cm 1 to water molecule vibration and those at 3307 and K and dehydrated because then the thickness of the K-
3048 cm 1 to NH 4 . The band at 3699 cm 1 is due to kaolinite. saturated expandable layers will be equal to the
thickness of illite layers and the I-S thus becomes a
(0.38+0.02) whereas the NH4 content increases from 0 two-component structure of interstratified 9.98 Å K-
to 0.22 cations per O10(OH)2. The increase of NH4 is containing and 10.33 Å NH4-containing mica layers.
accompanied by a significant increase of Al in tetrahedral In NH4-bearing I-S-V, K and NH4 in mica interlayers
sheets. For example, for samples taken from the shallow may be distributed as follows: (1) homogeneously, i.e.
depth (70, 80, 82, 83), the amount of IVAl per O10(OH)2 interlayers have identical amounts of (K + NH4) and
varies from 0.33 to 0.40 atoms whereas for diagenetically NH4/K ratios; (2) interlayers contain either K or NH4; or
transformed samples the IV Al content increases to (3) interlayers have different NH4/K ratios. For K-
0.54 0.66 atoms per O10(OH)2. The octahedral cation saturated and dehydrated samples, the amount of fixed
content of the 2:1 layers also changes significantly during NH4 per O10(OH)2, cNH4, is determined by the d(005)
diagenesis. For example, the VIAl content increases from value and is independent of the K and NH4 distribution,
1.31 per O10(OH)2 for sample 70 to 1.69 for sample 89 which is determined by the ratio between full width at
whereas the Fe and Mg contents decrease from 0.35 to half heights (FWHH) of 00l reflections. In particular,
0.16 and from 0.34 to 0.14, respectively. FWHH(005)/FWHH(002) determines the content of
NH4-bearing 10.33 Å layers, wNH4, in the mixed-layer
Mean number of layers in the CSDs and illite structure of type 2. It must be emphasized that the
fundamental particles functional relationships between cNH4 and d(005) and
Table 6 contains the mean numbers of layers in the between wNH4 and FWHH(005)/FWHH(002) are inde-
CSDs (N) determined by the 001 reflection profile pendent. Therefore, samples having equal amounts of K
Table 4. Averaged cation composition of the cutting samples (per O10(OH)(8 6p)) and Si/IVAl and IV
Al/VIAl calculated from
the formulae and determined by NMR.
IV VI
Sample Si Al Al Fe3+ Fe2+ Mg Soct p1 K NH4 Na Ca Si/IVAl IV
Al/VIAl
(2) (3) (2) (3)
704 3.70 0.30 1.60 0.28 0.07 0.27 2.22 0.89 0.33 0 0.24 0.03 12.5 11.7 0.17 0.18
80 3.62 0.38 1.51 0.30 0.05 0.28 2.14 0.93 0.35 0.05 0.24 0.01 9.6 9.6 0.25 0.24
82 3.67 0.33 1.63 0.23 0.05 0.30 2.21 0.89 0.36 0.08 0.24 0.01 11.0 11.0 0.20 0.21
83 3.67 0.33 1.57 0.25 0.04 0.30 2.16 0.92 0.35 0.04 0.26 0.02 11.2 11.2 0.21 0.21
113 3.56 0.44 1.62 0.24 0.05 0.23 2.14 0.93 0.38 x6 0.28-x 0.02 8.0 7.5 0.27 0.27
795 3.52 0.48 1.73 0.22 0.04 0.23 2.22 0.89 0.35 0.13 0.25 0.01 7.3 7.3 0.28 0.28
87 3.36 0.54 1.55 0.21 0.07 0.20 2.00 0.98 0.37 0.16 0.19 0.02 6.4 6.4 0.35 0.35
86 3.43 0.57 1.64 0.16 0.04 0.16 2.03 1.00 0.36 0.21 0.21 0 6.0 5.8 0.35 0.35
89 3.35 0.65 1.69 0.12 0.04 0.14 2.00 1.00 0.40 0.22 0.15 0.03 5.1 5.0 0.38 0.38
1
Proportion of the I-S-V or I-T-S-V phase
(2) From chemical formulae
(3) From NMR
4
Well G2, 2609 m
5
Well Lulu 1, 3420 m
6
not determined
92 Drits et al. Clays and Clay Minerals
and NH4 and the same d(00l) values may have different NH4, independent of distribution of K and NH4 over
00l basal reflection widths and samples having equal non-expandable mica interlayers in NH4-bearing I-S.
basal reflection widths may have different K and NH4 The fact that INH4 normalized for the number of OH
contents. For the studied mixed-layer phases, the cNH4 groups in the unit-cell (IOH) is proportional to the
and w NH 4 values determined from the d(005) and amount of tobelite layers determined from the
FWHH(005)/FWHH(002), respectively, are almost the FWHH(005)/FWHH(002) values is additional evidence
same (Drits et al., 1997a) and coincide with the content for separate distribution of K and NH4 among mica
of fixed NH4 per O10(OH)2 determined by the accurate interlayers.
isotopic dilution method (Lindgreen, 1994). These The total content of non-expandable mica interlayers,
results demonstrate that, first, the K-saturated and wI+’T, determined by the experimental XRD simulation
dehydrated phase has a mixed-layer structure consisting and the amount of tobelite interlayers, wT, determined
of 9.98 Å and 10.33 Å layers, and, second, the from the basal reflection positions and widths are
diagenetically transformed phase consists of interstrati- compared in Table 1. The difference between these
fied illite, tobelite, smectite and vermiculite. Simulation values is equal to the content of illite interlayers, wI, in
of the experimental XRD patterns using the I-T-S-V the I-T-S-V studied. Table 1 shows that wI is equal to
model confirms the second conclusion and shows a (0.65+0.05) independent of sample location, burial
tobelite content equal to cNH4 and wNH4. The direct link depth and temperature. It is remarkable that according
between total amount of NH4 and content of tobelite to the structural formulae (Table 4) the I-T-S-V contain
layers in the NH4-bearing I-S studied follows from the almost the same amount of K per O 1 0 (OH) 2
relationship between INH4/IOH and wT (Figure 9). The IR (0.38+0.02). This means that diagenesis does not
band at 1430 cm 1 is sensitive to the total amount of change illite interlayers. Accordingly, diagenetic trans-
Table 6. Number of layers in CSDs and in fundamental particles and content of K per
O10(OH)2 in the I-T-S-V.
Sample N1 wS+V2 n̄ 3 wT4 wI5 K6 K7
cK = 0.75 cK = 0.90
0.25
0.2
0.15
wT
0.1
0.05
0
0 0.05 0.1 0.15 0.2 0.25
INH4/IOH
Figure 9. Content of tobelite layers, w T , in I-T-S-V vs. the ratio between IR optical densities, I NH 4 /I OH , of the NH 4 band at 1430 and
the OH band at 3631 cm 1 .
formation of I-S-V in the Upper Jurassic oil-source rock finding that only a little soluble NH4 is present in the
is accompanied not by an illitization of smectite shale compared to K, any NH4 formed during transfor-
involving fixation of K as generally accepted but solely mation of organic matter probably having been flushed
by a formation of tobelite layers due to fixation of NH4 out of the shale by the migrating solutions. Previously,
in former smectite and/or vermiculite interlayers. such migrating solutions were indicated by Jensenius
Tobelitization of smectite is a characteristic feature of (1987) who investigated fluid inclusions in the Upper
the I-T-S formed simultaneously with oil generation. In Cretaceous chalk of the North Sea Skjold oil field and
particular, our investigation of samples from Cambrian found indications that hot water solutions (87 1128C, as
black shales has shown that the I-S and I-T-S have the indicated from inclusions) from deeper layers had
same K content and increase of layer charge because the migrated through the chalk.
increase of IVAl is compensated by NH4 (Lindgreen et
al., 2000). As a consequence, the K content in mixed- Fixed K content in illite interlayers of the I-T-S-V
layers from oil-source rocks cannot be used to indicate Meunier and Velde (1989) and Środoń et al. (1992)
the increase in amount of mica layers, because the newly suggested that the fixed K+ content in illite interlayers,
formed mica layers are in fact tobelite layers. cK, of I-S is 0.9 per O10(OH)2. Simulation of the XRD
Furthermore, conventional K/Ar dating cannot be patterns of the I-S-V and I-T-S-V from the Upper
applied to investigate the timing of diagenesis but Jurassic shales performed by Drits et al. (1997b) and in
represents mainly the time of formation of the initial or the present study has shown that 0.75 K per O10(OH)2
mica material (later transformed to I-S-V by weathering) provided much better agreement between the experi-
in the parent rock. mental and calculated XRD patterns than 0.9 K per
An exception from the tobelitization is, however, O10(OH)2. The 0.75 K is supported by calculations of the
sample x1 taken from the East Rosa 3/3A well at 1520 m amount of interlayer K from the mean number of layers
depth. For this sample, the amount of NH4 determined in illite fundamental particles (Table 6). The number of
by the isotopic dilution method (Lindgreen, 1994) and illite interlayers in illite fundamental particles is n̄ 1
from the positions and widths of basal reflection (Drits and the K content per O 10(OH)2 in the structural
et al., 1997a) does not exceed 0.03 per O10(OH)2. formulae of the I-S-V having only illite interlayers is
However, the experimental XRD pattern corresponds to
(n̄ 1)cK/n̄ (5)
group 2 and an I-S-V phase containing the largest
amount of illite layers (0.86), I and S being distributed If the non-expandable mica interlayers are illite, wI,
with maximum possible degree of ordering at R = 1 and and tobelite, wT, then the amount of K per O10(OH)2 in
V layers have a tendency to segregation. This sample is the I-T-S-V should be equal to
special in the way that the Upper Jurassic shale in this
cKwI(n̄ 1)/[(wI + wT)n̄] (6)
well is comparatively thin (10 m) and that it is
positioned immediately above a Zechstein salt structure The K contents calculated for cK = 0.75 and 0.90 are
and below a thin Upper Cretaceous chalk (89 m). This given in Table 6. The number of K calculated for cK =
indicates that the salt movement started in Cretaceous 0.75 is, except for sample 17, practically equal to the
time and that the Upper Jurassic shale was never buried number determined from the structural formulae
to a depth sufficient to initiate oil formation or (Tables 4 and 5). The same value, 0.75 K per
diagenesis of smectite layers. Instead, it is probable O10(OH)2, was previously determined for illite inter-
that salt solutions have migrated from the Zechstein salt layers in I-S and I-T-S from Cambrian Black shale of the
through the thin Upper Jurassic shale, having supplied Baltic area (Lindgreen et al., 2000). This coincidence,
sufficient heat to cause tetrahedral substitution of obtained by two independent methods is additional
smectite transforming it into illite and supplying the K evidence that tobelitization of the expandable layers is
for formation of illite layers. This agrees with the not accompanied by K fixation.
94 Drits et al. Clays and Clay Minerals
Structural mechanism of an I-T-S-V transformation having mean numbers of layers equal to 2.2 2.4
(Table 6).
Several main structural mechanisms of transforma-
tion have been proposed for the smectite illitization: Selective sorption and fixation of NH4
solid state, neoformation and Ostwald ripening (Altaner Why is NH+4 but not K+ fixed in newly formed mica
and Ylagan, 1997; Środoń et al., 2000). Each of these layers? Analysis of soluble plus exchangeable NH4 and
reactions may result in fixation of NH4 in newly formed K in the core samples of the Central Trough shows that
mica layers. However, only the solid-state transforma- the amount of soluble K decreases with depth of the
tion can explain the constant amount of pure illite layers sample whereas the amount of NH 4 increases
when the amount of mica layers increases due to (Lindgreen, 1994). These results correlate with the
smectite tobelitization. observed increase of fixed NH4 in the studied I-T-S-V
Neoformation and recrystallization as well as (Table 1). However, in the Central Trough, concentrations
Ostwald ripening are accompanied by an increase of
particle size and, in addition, the newly formed particles
a
show well-developed crystal faces. In contrast, the solid-
state reaction does not change the size and the
morphology of the particles significantly. For the I-S-V
from the North Sea area, atomic force microscopy
(Figure 10) shows that the size and the shape of the
individual particles do not change significantly with
degree of diagenesis and that crystals with well-
developed faces are not appearing in the samples
transformed during diagenesis, two facts that support
the conclusion of a solid-state transformation. An
increase in the mean number of layers in CSDs from
5.3 5.9 for I-S-V from the Danish Subbasin and the
shallow Central Trough samples to 6.0 6.4 for I-T-S-V
from the diagenetically transformed central Trough
shales takes place because the transformation of former
smectite interlayers into tobelite is probably accompa-
nied by increasing numbers of layers parallel to each
other.
The tobelitization reaction may be imaged as a
sequential substitution of a smectite by a tobelite
interlayer. In the detrital mixed-layer crystals, NH4 and
Al migrate from solution to the interlayer space of
smectite interlayers. Substitution of Si for Al then
increases the negative charge of the tetrahedral sheets
in the 2:1 layers of smectite. Direct evidence of the
b
increase of tetrahedral Al in the 2:1 layers in the I-S-V
and I-T-S-V with increasing temperature and burial
depth has been obtained by solid-state high-resolution
MAS NMR spectroscopy (Lindgreen et al., 1991) and is
seen in the structural formulae (Table 5). This increase
in tetrahedral charge combined with fixation of dehy-
drated NH+4 between these illite tetrahedral sheets results
in formation of tobelite layers. Adjacent tetrahedral
sheets of smectite are thus transformed into vermiculite
tetrahedral sheets and further or directly into illite
tetrahedral sheets. In addition to substitution of Si for
Al, the transformation is accompanied by significant
modification of octahedral cation composition of the 2:1
layers due to a substitution of Fe and Mg for Al. The
migration of octahedral cations may take place through
smectite interlayers without disturbing the illite inter- Figure 10. Atomic force microscopy images of I-S fractions
layers. This is due to the fact that the illite fundamental deposited on graphite. 0.186 nN, non-contact mode. Mean value
particles in I-S-V prior to tobelitization are very thin filter applied. (a) Sample 112; (b) Sample 89.
Vol. 50, No. 1, 2002 Tobelitization of I-S in oil-source shales 95
0.25
of soluble plus exchangeable K, 0.8 to 7.6 meq/100 g, is
several times greater than for NH 4, 0.07 to 0.64. Thus, in
0.2
spite of the predominance of K in solution, a selective
absorption and fixation of NH4 in the former smectite
interlayers takes place. This preferential absorption of 0.15
cs
different reaction of different cations to increasing layer 0.1
charge (Eberl, 1986; Lindgreen, 1994). For a given layer
112
charge, the larger cation in the dehydrated state is more 0.05
easily dehydrated. At low layer charge all cations are
hydrated. With increasing interlayer charge due to
0
substitution of Si for Al in smectite tetrahedral sheets, 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
NH4 is dehydrated and fixed first because of its larger cf
size (1.43 Å) compared to K (1.33 Å) and possibly also
because of the formation of hydrogen bonds with oxygen Figure 11. Amount of soluble plus exchangeable NH4 , c(s), in
atoms in the interlayer surface and because of the cation samples from cores vs. the fixed NH 4 content, c(F), in I-T-S-V in
the same samples, with the plot for sample 112 identified. Data
demixing effect (Gaultier and Mamy, 1979, 1982). from Lindgreen (1994).
Analysis of the data presented by Lindgreen (1994)
for soluble plus exchangeable (cSNH4) and fixed (cFNH4) trans-vacant (tv) 2:1 layers. Accordingly, a formation of
NH4 content for the studied samples shows that these illite (tv) layers from smectite (cv) layers should lead to
values are interrelated: the higher the value of cSNH4, the interstratification of tv and cv layers in I-S and to an
higher the value of cFNH4 (Figure 11). An exception is increase in proportion of tv 2:1 layers, whereas I-S
sample 112 of group 1 taken from well E1 at 2984 m formed during weathering of illite (tv) independent of
depth which has the highest soluble NH 4 content the layer types should have only tv octahedra. Drits et al.
(0.56 meq/100 g) but the I-T-S-V phase contains only (1995) showed that tv and cv 2:1 layers have dehydroxyl-
0.06 of NH4 per O10(OH)2 (Figure 11). The sample is at ation temperatures of 500 6008C and 650 7508C,
the onset of oil generation (Thomsen et al., 1983); respectively. Therefore, I-S-V formed from smectite
therefore this NH4 probably originates from deeper should have a DTA-EWA curve with two endothermic
layers and has migrated through the shale to 2984 m peaks corresponding to dehydroxylation of the tv and cv
depth. Sample 113 from the same well at 3938 m layers (Drits et al., 1996, 1998b; Cuadros and Altaner,
contains 0.17 tobelite interlayers (Table 1). Accordingly, 1998; Lindgreen et al., 2000). In contrast, DTA-EWA
fixation of NH3 at temperatures lower than maximum oil curves of I-S-V formed by weathering of illite (tv)
generation does not take place, because of the lack of should have only one dehydroxylation peak below
increase in tetrahedral charge. The link between 6008C. For the North Sea I-S-V and I-T-S-V, the main
tobelitization and maximum oil generation is probably thermal decomposition occurs below 5508C for samples
a link between breaking of bonds in organic material of groups 1, 2 and 3 (Table 2), and therefore the North
(e.g. C C and C N bonds) and, at the same tempera- Sea I-S-V and I-T-S-V consist mainly of tv layers. Some
ture, breaking of Si O bonds and formation of Al O samples contain a small amount of cv layers, but this
bonds, and, with the resulting increased tetrahedral content does not correlate with the amount of non-
charge, fixation of released NH3. expandable mica layers or with the degree of diagenesis.
The predominance of tv layers in the studied I-S-V and I-
Source areas and parent material for the I-S-V T-S-V is in accord with the fact that these minerals
Hansen and Lindgreen (1989), Lindgreen (1991) and formed by weathering of illite.
Drits et al. (1997b) showed that I-S-V samples of group 1 The fact that all the mixed-layer minerals studied,
have detrital origin, and have not been transformed independent of their location and burial depth, contain
significantly by diagenesis. The mixed-layer minerals the same amount of pure illite layers (Table 1) finally
probably formed during weathering of illite and were confirms that the I-S-V and I-T-S-V have formed from
transported from the Fennoscandian Shield. The I-S-V of the same parent material and have the same source area.
group 1 have a significant tendency to segregation which
is quite unusual for diagenetically transformed I-S. This Oil generation and tobelite transformation windows
tendency to segregation of the layer types formed during In the Norwegian-Danish Basin, the amount of uplift
illite weathering was probably inherited in the I-S-V. In during the Tertiary was ~1 km. If the geothermal
addition, the I-S-V is associated with mixed-layer gradient was ~258C/km (Thomsen et al., 1983), the
kaolinite-illite-vermiculite which probably also formed present-day temperature of 24 518C had reached a
through weathering of illite (Drits et al., 1997b). maximum of 50 758C. Except for sample x1, samples
Tsipursky and Drits (1984) showed that normally from the Central Trough are experiencing their max-
Al-rich smectites consist of cis-vacant (cv) and illites of imum temperature during burial history at present,
96 Drits et al. Clays and Clay Minerals
180
window’ occurs within the ‘oil window’ and is restricted
160 30
Tobelite window to the physical-chemical interval where maximum
15 17
140 release of NH4 coincides with intense formation of
8 20
120 20 22 tobelite layers. For the North Sea oil-source rocks, the
20 4
19 tobelitization reaction probably takes place during oil
T (˚C)
100
5 6 generation and oil expulsion at vitrinite reflectance
80 8 3
values of 0.55 0.7 Ro and T & 100 1208C (Figure 12).
60
7
40 I-T-S-V as an indicator of oil-generating source rocks
20 3 in sedimentary basins
5
0 The tobelitization of smectite in I-S or I-S-V is
0.4 0.5 0.6 0.7 0.8 0.9 1 probably typical for all oil-source rock shales. Therefore
Ro (%) the presence of I-T-S or I-T-S-V in sedimentary rocks
Figure 12. Relationship between proportion of tobelite layers, may be used as independent evidence for oil formation
w T , formation temperature of the samples, T, and vitrinite even after migration or thermal decomposition of the oil
reflectance values, R o , for North Sea and Norwegian–Danish because once formed, the tobelite interlayers should be
basin I-T-S-V minerals. Labels for each point correspond to the
stable during low-temperature metamorphism as the
w T value for the I-T-S-V. The tobelite window is limited by the
vitrinite reflectance values Ro & 0.55 0.7%. release of fixed NH4 from NH4-bearing layers takes
place above 4008C (Higashi, 1978). The results obtained
because of a large subsidence during the Tertiary. for I-S from Cambrian Alum black shales of both early
Figure 12 shows how the content of newly formed diagenetic and anchimetamorphic grade may serve as an
tobelite layers in I-T-S-V depends on the formation example (Lindgreen et al., 2000). In the early diagenetic
temperature, T (T. Bidstrup, pers. comm.; Lindgreen, samples, the I-S is composed of two phases, one of which
1991) and the vitrinite reflectance values, Ro (Thomsen contains 0.05 and the other 0.25 smectite interlayers. The
et al., 1983; Lindgreen, 1991; C. Guvad, pers. comm.). first highly illitic phase does not change during meta-
The I-S-V of samples from the Norwegian–Danish Basin morphism. In the second phase, 0.10 0.18 smectite layers
and from shallow depth (<3000 m) in the Central Trough are transformed to tobelite layers. Thus, the number of T
(with T <1008C and Ro <0.70%) contain <0.10 tobelite layers formed in this phase corresponds to that found in
layers (Figure 12, Table 1). At greater depths in the diagenetically transformed I-T-S-V from the Central
Central Trough the number of tobelite layers is generally Trough. It is remarkable that, as for the North Sea I-S-V
much larger, for 0.55 < Ro < 1.0% and 110 < T < 1408C and I-T-S-V, the number of I layers in the second phase is
the number of tobelite layers is 0.15 0.22 (Table 1). not changed and all illite interlayers contain 0.75 K per
Before oil generation, the I-S-V from the Central O10(OH)2. As in the case of the North Sea, transition of
Trough had a limited number of expandable interlayers I-S to I-T-S in the Cambrian shales takes place through a
(S+V) available for tobelitization (0.30 0.35, Table 1). solid-phase transformation. However, in contrast to I-T-S-V
Maximum oil generation and expulsion in the Central from the Central Trough, the I-T-S from Cambrian shale
Trough took place at vitrinite reflectance values of have been transformed strongly during anchimetamorph-
0.6 0.7 Ro (Thomsen et al., 1983), the stage at which ism. Therefore, organochemical generation parameters
maximum release of NH4 during the generation process cannot be applied. Furthermore, the absence of material
occurs (Hunt, 1979; Barth et al., 1996). At this stage (at from terrestrial vegetation in the Cambrian renders coal
T & 100 1208C) 50 70% of the expandable interlayers petrography useless. However, the presence of a
have been transformed into tobelite interlayers significant amount of I-T-S in the highly transformed
(Figure 12). Thus, once the shale has passed into the Cambrian samples in combination with a high potential
stage that we call the ‘tobelite window’, oil generation for oil generation in early diagenetic samples containing
and expulsion is accompanied by intense tobelitization I-S with a small amount (<0.05) of tobelite layers leads
of previously expandable interlayers. This conclusion is to the conclusion that the tobelitization in the Cambrian
in agreement with the observed relationship between the Alum black shales is related to oil generation which took
contents of soluble and fixed NH4 (Figure 11). With place before the rocks were transformed during
further increasing temperatures and vitrinite reflectance Caledonian folding, including heating to 2008C.
values, the tobelite layer formation decreases or even For the North Sea I-S-V and I-T-S-V, structures
ceases, probably because the amount of available NH4 in intermediate between those of group 1 and 2 have not
solution decreases with increasing Ro. The ‘oil window’ been found. This indicates that the tobelitization is an
for oil generation in different source rocks corresponds almost instantaneous process once the temperature for
to the vitrinite reflection interval 0.6 < Ro < 1.3 (Hood et the breaking of Si O bonds has been reached. Thus,
al., 1975; Dow, 1977; Tissot and Welte, 1978; Thomsen samples 82 and 112 from depths 2899 and 2983 m,
et al., 1983) and the temperature interval 80 < T < 1408C respectively, are not tobelitized but samples x5 and x6
(Hunt, 1979; Barth et al., 1996). The ‘tobelitization from ~2960 m depth are.
Vol. 50, No. 1, 2002 Tobelitization of I-S in oil-source shales 97