SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Data from the literature
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Iron silicate minerals are a significant component of sedimentary systems but their modes of formation remain controversial. Our analysis of published data identifies end-member compositions and mixtures and allows us to recognize controls of formation of different mineral species. The compositional fields of glaucony, Fe-illite, Fe–Al smectites are determined in the M+/4Si vs. Fe/Sum of octahedral cations (M+ = interlayer charge). Solid solutions could exist between these phases. The Fe–Al and Fe-rich clay minerals form two distinct solid solutions. The earliest phases to be formed are Fe–Al smectites or berthierine depending on the sedimentation rate. Reductive microsystems appear in the vicinity of organic debris in unconsolidated sediments. The Fe is incorporated first in pyrite and then in silicates after oxidation. Potassium ions diffuse from the sea-water–sediment interface. If not interrupted, the diffusion process is active until reaction completion is reached, i.e. formation of Fe-illite or glauconite or a mineral assemblage (berthierine–nontronite) according to the available Al ion amounts in the microsystem. Mixed-layer minerals are formed when the diffusion process is interrupted because of sedimentation, compaction or cementation. Despite the common belief of their value as palaeoenvironment indicators, these minerals can form in a variety of environments and over a period of millions of years during sediment burial.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Data from the literature
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Clay minerals formed in marine or hypersaline environments are constantly Fe–K-rich species: glauconite, Fe-illite, Fe-smectite, nontronite. They have been described in contrasted sedimentation rate environments such as big river mouths (Giresse et al., 1988; Michalopoulos and Aller, 1995; Wiewiora et al., 1999), oceanic floor (Steinberg et al., 1987), accretion prisms (Gaudin et al., 2005), marine hydrothermal vents (Buatier et al., 1989), continental shelves (Odin, 1988) or evaporitic systems (Pédro et al., 1978; Baker, 1997; Huggett and Gale, 1997; El Albani et al., 2005). Their formation may be very rapid; a few months for K–Fe–Mg beidellite-type clay minerals in the Amazon river mouth (Michalopoulos and Aller, 1995) or very slow, a few million years for glaucony (Smith et al., 1998). Most often, these clay species grow on solid supports (detrital minerals or biogenic debris) from the sea-water–sediment interface to a variable depth inside the unconsolidated sediments.

In spite (or because) of the huge literature dedicated to the description of these K–Fe–Mg marine clays, many points remain controversial or unclear. Their formation mechanism is still an open question. Using published data, the purpose of this paper is to reconsider some of these points. First, the extension of the solid solutions for each mineral phase is investigated. Special attention is paid to the relation between composition domain overlaps and mixed layering. Such relationships could indicate a compositional continuity between the mineral species concerned. In a second step, the mechanism of formation of these Fe-bearing clay phases is discussed in order to identify the factors which determine their species.

Method

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Data from the literature
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Data sources

The data (Table 1) used here have been selected from published papers using the following criteria: (1) the chemical compositions must be coherent with the definition given by AIPEA (Bailey, 1980); (2) the structural formulae must be compatible with a dioctahedral structure. Celadonite is not considered here, because it is a high-temperature phase, typical of basalt alteration and not of authigenesis in sediments. The chemical compositions presented in the selected data have been obtained by wet chemical methods (including the measurements of FeO and Fe2O3 amounts), electron microprobe or analytical electron microscopy (AEM).

Table 1.   General characteristics of Fe-bearing clay minerals whose chemical compositions are used here. The references are given for each data source.
Geological dataReference
FormationSampling depthStratigraphical age
Glaucony
MiscellaneousHendrix and Ross (1941)
MiscellaneousFoster (1969)
MiscellaneousWeaver and Pollard (1973)
MiscellaneousBuckley et al. (1978)
MiscellaneousThompson and Hower (1975)
MiscellaneousNewman and Brown (1987)
Bohemian formations???Upper CretaceousCimbálníková (1971)
First Wilcox sandstone3030 mEoceneStrickler and Ferrell (1990)
SandstonesOutcropProterozoicMendes Guimarães et al. (2000)
Clayey-siltstone Conglomerate sandstones435–795.5 mAlbian CenomanianRousset et al. (2004)
Fe-illite
 Limestones Calcareous sandstonesOutcropMiddle Cambrian Lower OrdovicianBerg-Madsen (1983)
 Shallow-marine, brackish, lacustrine, hypersaline faciesOutcropPurbeckianDeconninck et al. (1988)
 Fluvial litharenites2258.9–2266.9 mEarly TriasBaker (1997)
 Shallow-marine, brackish, lacustrine, hypersaline faciesOutcropPurbeckianEl Albani et al. (2005)
Nontronite & Fe-montmorillonite
 Galapagos moundsSea-water interface 3000 m, core 0.13–3.75 mRecentBuatier et al. (1989)
 Calcareous biogenic sediments Costa Rica marginSea-water interface 3000 m, core 0.13–3.75 mRecentGaudin et al. (2005)
Fe–Al beidellite-montmorillonite
 Black shalesDSDP leg 11AlbianSteinberg et al. (1987)
 Sandstones, siltstones0–700 mMiocene to RecentWiewiora et al. (1999)
Greenalite, Nontronite & Berthierine
 Congo River estuary12–16 mMiocene to RecentWiewiora et al. (1996)

The M+/4Si vs. Fe/Sum octa. chemiographic projection

In spite of careful sample preparation described in the selected papers, the clay mineral separates, which have been analysed using wet chemical methods, were sometimes not totally purified. Consequently, the calculated structural formulae suffer from experimental errors due to mineral contamination. As the most common source of contamination for glauconite or Fe-smectites in sediments is the presence of quartz, pyrite or calcite, all the analyses used in the paper were previously projected in M+–4Si–R2+ coordinates (Meunier and Velde, 1989). The polluted analyses plot on mixing lines between the glauconite, Fe-illite or Fe-smectite domains and the 4Si, R2+ or M+ poles, respectively. These data have been eliminated. A similar treatment was also applied to the chemical compositions obtained using electron microprobe, because clay minerals are frequently smaller than the spot size and contamination is always possible. To solve this problem, Ireland et al. (1983) and Curtis et al. (1987) preferred to use AEM. In spite of difficulties in determination of the ‘k factors’, the AEM analytical precision was roughly similar to that of electron microprobe equipped with an energy dispersion system. We used the chemical compositions obtained by Ireland et al. (1983) to check the validity of those obtained by wet chemical and electron microprobe methods.

The presence of other phyllosilicates such as trioctahedral 1 : 1 or 2 : 1 : 1 phases (sometimes interstratified) leads to erroneous calculations of the cation proportions in both the tetrahedral, octahedral and interlayer sheets. This drawback can be partially reduced if ratios are used rather than absolute cation amounts. Indeed, ratios being independent of the oxygen calculation basis, the artefact due to contamination by 1 : 1 or 2 : 1 : 1 clay phases is eliminated. As structural formulae are generally calculated for Si4O10, the M+/4Si vs. Fe/Sum octahedral cations coordinates were chosen. The M+/4Si ratio corresponds to the interlayer charge (Na + K + 2Ca) divided by Si amount/4. For instance, the coordinates of celadonite are theoretically: M+/4Si = 1/1; Fe/Sum octa. = 1/2. The coordinates discriminate the Fe-bearing clay species (Fe-montmorillonite, nontronite) but not the Al-bearing ones (montmorillonite, beidellite).

The end-member compositions used here are presented in Table 2. A theoretical composition of glauconite has been proposed on the mica model (Bailey, 1980) assuming a layer charge of −1 per Si4O10. It is close to that of celadonite. However, another composition has been considered because glauconite is a low temperature mineral which could be considered closer to illite than to mica (Meunier and Velde, 2004). Thompson and Hower (1975) have shown that glauconies having less than 5% expandable layers are near 0.75 K per Si4O10. Using this value, a formula was derived from that proposed by the AIPEA (Table 2). In spite of the fact that berthierine is a trioctahedral species, it will be represented here because it is frequently associated with glauconitic minerals. Its compositional domain has been determined using data from Brindley (1982).

Table 2.   Projection of the compositions of the Fe-bearing clay mineral end-members in the M+/4Si vs. Fe/Sum octahedral cations diagram.
End-membersSymbolStructural formulae
CeladoniteCel[Si4−xAlx]O10(Fe3+Mg)(OH)2K with x < 0.2
Glauconite (mica)Glm[Si3.67Al0.33]O10(Rinline imageRinline image(OH)2K, Fe3+>Al Mg>Fe2+
Glauconite (illite)Gli[Si3.75Al0.25]O10(Rinline imageRinline image(OH)2K0.75
Nontronite (high charge)No Hc[Si3.4Al0.6]O10(Feinline image(OH)inline image
Nontronite (low charge)No Lc[Si3.7Al0.3]O10(Feinline image(OH)inline image
Fe-montmorillonite (high charge)FeMo HcSi4O10(Feinline imageMg0.6)(OH)inline image
Fe-montmorillonite (low charge)FeMo LcSi4O10(Feinline imageMg0.3)(OH)inline image
Beidellite (high charge)Be Hc[Si3.4Al0.6]O10(Al2)(OH)inline image
Beidellite (low charge)Be Lc[Si3.7Al0.3]O10(Al2)(OH)inline image
Al-montmorillonite (high charge)AlMo HcSi4O10(Al1.4Mg0.6)(OH)inline image
Al-montmorillonite (low charge)AlMo LcSi4O10(Al1.7Mg0.3)(OH)inline image

Data from the literature

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Data from the literature
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The chemical composition domains of glauconite, Fe-illite, Fe–Al smectites and Fe-smectites

Data from Hendrix and Ross (1941), Foster (1969), Weaver and Pollard (1973), Thompson and Hower (1975), Buckley et al. (1978), Newman and Brown (1987) and Cimbálníková (1971) have been used to determine the chemical compositional domain of glaucony. This domain extends from the glauconite-mica (Glm) and glauconite-illite (Gli) end-members towards the Al-smectite compositions (Fig. 1a). The glauconite composition field appears to be roughly limited to Fe/Sum octa. = 0.25 on its ‘smectite side’ (grey zone in Fig. 1a). The chemical compositions obtained using AEM (Ireland et al., 1983) plot inside that field (Fig. 1b). This means that the effects of mineral contamination in the selected data obtained by wet chemical methods are negligible. Therefore, these analyses can be used safely. This allows us to take into account the FeO and Fe2O3 amounts which are not measurable using electron microprobe or AEM analyses.

image

Figure 1.  Plots of the chemical compositions of the Fe-bearing clay phases in the M+/4Si vs. Fe/Sum octahedral cations diagram. (a) Glauconies (wet chemical analyses). (b) Glauconies (AEM). (c) Fe-illite. (d) Fe–Al smectites.

Download figure to PowerPoint

The composition domain of Fe-illites (or aluminous glauconites) is represented in Fig. 1(c). A part of the compositions published by Berg-Madsen (1983) and Baker (1997) lie in the glauconite domain. Consequently, the corresponding ‘Fe-illite’ cannot be distinguished from glauconites. On the contrary, the other compositions are comparable to those published by Deconninck et al. (1988) and El Albani et al. (2005). They plot in a field whose limits are: 0.12 ≤ Fe/Sum octa. ≤ 0.25 and 0.95 ≤ M+/4Si ≤ 0.55 (grey zone in Fig. 1c). It is to be noticed that data from El Albani et al. (2005) are those of Fe-illite/smectite mixed layer minerals.

The compositions of Fe–Al smectites given by Steinberg et al. (1987) and Wiewiora et al. (1999) are largely scattered (Fig. 1d). Part of them lie in the Fe-illite and glauconite composition fields. However, if Fe–Al smectites are considered to have ordinary layer charges, i.e. 0.3–0.6 per Si4O10, their composition domain is bordered as follows: 0.12 ≤ Fe/Sum octa. ≤ 0.25 and 0.3 ≤ M+/4Si ≤ 0.7 (grey zone in Fig. 1d).

The glaucony, Fe-illite and Fe–Al smectite composition domains as they are defined here are contiguous. These domains define the type compositions for each species. However, data from the selected literature show that the fields for each species extend over several contiguous compositional types. Even if we ignore their exact extension, one can infer that there is no composition gap between these three phases. This is not the case for the Fe-rich smectites whose compositions are scattered in a distinct domain (Fig. 2). Most of the published compositions lie outside the theoretical smectite composition domain limited by the Nohc, Nolc, FeMohc and FeMolc end-members. Except for the Fe-rich smectites analysed by Wiewiora et al. (1999), the other compositions (Buatier et al., 1989; Gaudin et al., 2005) plot in a domain bordered by the FeMohc and FeMolc end-members and oriented towards the berthierine composition field (grey zone in Fig. 2).

image

Figure 2.  Plot of the chemical compositions of the Fe-smectites in the M+/4Si vs. Fe/Sum octahedral cations diagram.

Download figure to PowerPoint

The iron and potassium contents of glauconies are not correlated (Fig. 3a). Iron is present in the two valency states; the more oxidized one (Fe3+) predominates. The Fe2+/Fe3+ ratio is highly variable; it is obviously not correlated to Si or K amounts (Fig. 3b,c). This was discussed by Velde (2003) who considered that all iron cannot be found entirely within the sediment included in proto-glauconite pellets, a part of it being necessarily introduced from outside.

image

Figure 3.  Iron content of glauconies. (a) Total Fe content vs. K. (b) Fe2+/Fe3+ vs. 4Si. (c) Fe2+/Fe3+ vs. K.

Download figure to PowerPoint

The variability of glaucony chemical compositions

It has been constantly observed that the composition of glauconies varies from place to place in a given formation and from grain to grain in a given place. Figure 4 shows the dispersion of two Fe-rich glaucony sample series from the upper Cretaceous Bohemian formations (Cimbálníková, 1971) and sandstones and siltstones of the Albian-Cenomanian formations of south-eastern France (Rousset et al., 2004) and an Al-rich glaucony series from the Proterozoic sandstones of the Paranoa Group, mid-western Brazil (Mendes Guimarães et al., 2000). Neither the increasing stratigraphical age, nor the intensity of the diagenetic conditions experienced by the glaucony-bearing rocks modify the compositions of glaucony grains. This means that the measured composition is acquired early in the post-sedimentary stage and does not change during the burial stage. However, this is a controversial point as several papers argue that the glauconitization process is active over long periods of time. For instance, Strickler and Ferrell (1990) showed that the composition of glaucony grains varies according to the degree of cementation of the sediment (Fig. 5). The most ‘mature’, i.e. Fe-rich, grains are found in the uncemented zones, while the most ‘immature’ ones have been observed inside concretions. The authors estimated that the ‘maturation process’ (incorporation of Fe3+ cations) is active at depths from 0.6 to 1.8 km. They claimed that glauconitization is a time-dependent process. This is supported by Smith et al. (1998) who showed from 40Ar/39Ar ages of glaucony grains that glauconitization could be active over periods of several million years.

image

Figure 4.  Variability of the chemical composition of glauconies formed in similar conditions in different geological formations.

Download figure to PowerPoint

image

Figure 5.  Variation of the composition of glauconies with the degree of cementation of the Wilcox sandstones, Upper Palaeocene to Lower Eocene (from Strickler and Ferrell, 1990).

Download figure to PowerPoint

Because of the great importance of glaucony for the geological time scale, the time-dependence of the glauconitization process has been investigated early. Using several sample series, Odin and Dodson (1982) have shown that the apparent K–Ar age decreases with increasing K2O content of glaucony. The apparent K–Ar ages depend on the relative proportions of detrital vs. authigenic clays in the analysed samples. The ‘zero age’ is obtained by extrapolation of the detrital component to zero. The example of the Gulf of Guinea sample series shows that the amount of detrital clays (det.clay) decreases with increasing K2O according to an exponential function (Fig 6). Consequently, the detrital clay + K + Fe [RIGHTWARDS ARROW] glaucony mineral reaction is a first order one:

  • image
image

Figure 6.  Relation between apparent K–Ar age and the K2O amount in glauconies from the Gulf of Guinea (from Odin and Dodson, 1982).

Download figure to PowerPoint

The conditions of formation of the Fe- and Fe–Al smectites

Figures 1 and 2 show that the composition domains of Fe- and Fe–Al smectites do not overlap: no intermediate compositions between the Fe–Al beidellite–montmorillonite and the Fe-montmorillonite–nontronite composition fields. The origin of the composition gap between the two smectite groups is related to the conditions of formation. The compositions of the hosting sediments and the sedimentation rate could be the factors that control the formation of one or the other smectite groups.

The Holocene to Miocene deltaic sediments of the Congo river contain smectites of the nontronite or Fe–Al beidellite group, which form in low or high sedimentation rate environments, respectively (Wiewiora et al., 1996, 1999). In both cases, they are associated with berthierine. Two mineral reactions may be inferred from these observations:

1. kaolinite + quartz + goethite [RIGHTWARDS ARROW] nontronite + Fe-rich 7 Å phase (berthierine).

2. Fe-kaolinite + quartz [RIGHTWARDS ARROW] Fe–Al beidellite + Fe-rich 7 Å phase (berthierine).

In both cases, kaolinite is destabilized and becomes a source of aluminium for new phases. Aluminium is consumed first in berthierine. Nontronite is formed instead of Fe–Al smectites, because goethite dissolves and provides large amounts of Fe into the system. In the absence of goethite, two aluminium-rich phases are formed concomitantly: Al–Fe smectite and berthierine.

The Albian black shales collected in the Atlantic ocean (Steinberg et al., 1987) are aluminium-rich sediments. The clay fraction is composed of illite and smectite particles which exhibit Fe–Al beidellite or Fe–Al montmorillonite overgrowths, respectively. The neoformation of these Fe–Al smectites was considered to be due to the reaction of detrital clay particles with pore-water during early diagenesis. The chemical conditions prevailing during the mineral reactions could be estimated through the associated mineral phases. The presence of clinoptilolite, siderite, iron sulphide and opal indicate that the solutions were alkaline, reducing and saturated with respect to amorphous silica.

The sediments overlying the Galapagos spreading centre (Buatier et al., 1989) or the Costa Rica margin (Gaudin et al., 2005) are both calcareous sediments mostly composed of pelagic ooze. Calcite is the predominant mineral, but the sediments contain opal also and siliceous biogenic remains. No detritals such as clay minerals, feldspars or mica have been observed. The sediments are aluminium poor. In both cases, the green material was seen to be composed of Fe-rich, Al-depleted smectites. Lath-shaped particles observed in the sediments overlying the Galapagos hydrothermal systems were described as glauconite particles because of their high charge. However, they are far from the definition given by Bailey (1980) as total Al varies from 0 to 0.17 per Si4O10. The Fe-rich smectites are mostly Fe-montmorillonites but their compositions are scattered in a large domain, roughly oriented towards the berthierine composition (Fig. 2).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Data from the literature
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The numerous published data devoted to Fe-bearing clay mineral species formed in marine or hypersaline environments show that the transition from expandable phases to non-expandable ones, if it exists, does not depend on the age of the hosting formation, nor on burial depth. This was discussed by Velde (2003). From the results presented above, it appears that all these minerals form in open reductive microsystems which attract some chemical components from outside them by ion diffusion in the pore solutions. The potential relation between the species of the Fe-clay mineral which is formed and the local conditions such as the composition of the sediment must be investigated first. Then, as chemical diffusion is involved in the process, the influence of time should be studied.

Relation between Fe-bearing clays, sediment composition and time

Two groups of Fe-bearing clays formed in marine or hypersaline environments can be distinguished: (1) the Fe–Al smectite–Fe-illite–glauconite group in which no composition gap separates the different phases; (2) the Fe-montmorillonite–nontronite–berthierine group in phases are separated by composition gaps (Fig. 7). The absence of a composition gap in the first group is to be related to the capacity of glauconite to be interlayered with smectite as shown by Thompson and Hower (1975). The extension of the glaucony domain towards the Fe–Al smectite composition suggests that the expandable layers in the mixed layered structures have Fe–Al smectite compositions. This reinforces the assumption of a complete solid solution between Fe–Al smectites and glauconite. A similar solid solution may exist between Fe–Al smectites and Fe-illite because mixed layer minerals have been observed (El Albani et al., 2005). In other words, a complete glaucony-Fe illite–Fe,Al smectite solid solution may exist.

image

Figure 7.  Compositional domains of Fe-bearing clay phases in the M+/4Si vs. Fe/Sum octa. coordinates. The mixed layer minerals (MLM) which have been identified are indicated in their compositional fields.

Download figure to PowerPoint

The second group does not show any overlap of the composition domains of Fe-montmorillonite or nontronite on the one hand and glaucony or Fe-illites on the other. Besides, there is no evidence of mixed-layering between these phases. On the contrary, the fact that mixed-layering has been observed between 1 : 1 Fe-rich layers (Wiewiora et al., 1999) could explain the extension of the Fe-montmorillonite composition field towards the berthierine type composition (Fig. 2).

Two factors may control the formation of one or the other Fe-bearing clay groups: the composition of the sediment and the conditions of sedimentation. If the influence of the composition of the sediment is obvious for the formation process of the Fe-montmorillonites, it is not determinative for the other clay species. Indeed, Fe-montmorillonites which are almost completely Al-depleted are formed in calcareous biogenic ooze deposits while glaucony, Fe-illite, Fe–Al smectites, crystallize in sediments containing Al-bearing phases. However, this is not apparently totally self-consistent because nontronite and berthierine are also formed in clay-bearing sediments (Wiewiora et al., 1996). Why is a Fe-rich smectite such as nontronite, which is not very different from Fe-montmorillonites, formed in spite of abundant aluminium sources in the environment? This could be due to the fact that berthierine crystallizes first in the youngest sediments (upper part of the depositional column) and then co-precipitates with nontronite in the older ones (lower part of the column). Berthierine and nontronite are the Al- and Fe-bearing phases of an authigenic mineral assemblage. In the absence of berthierine, Fe–Al smectites would form in the place of nontronite. The presence or absence of berthierine seems to be the decisive factor which determines the composition of the smectite in the green grains.

Considering the geological environments in the Congo river mouth, the berthierine–nontronite assemblage and Fe-beidellite are formed in the low and high sedimentation rate zones, respectively (Wiewiora et al., 1996, 1999). A low sedimentation rate favours the slow chemical exchanges between pore water and neoformed clays. This is consistent with the long maturation time needed for glauconitization (up to 5 Ma according to Smith et al., 1998). On the contrary, a high sedimentation rate gives rapid burial. Consequently, the Fe-beidellite has to be formed quickly near the sea-water–sediment interface. This was experimentally shown by Michalopoulos and Aller (1995) in the Amazon shelf sediments. These authors claimed that K–Fe–Mg clay minerals were formed on various solid substrates after only a few months.

Summarizing, the Fe–K–Mg clay minerals in marine or hypersaline environments crystallize according to five different reactions (Fig. 8):

image

Figure 8.  Summary of the different processes of formation and the corresponding environment conditions for Fe–Al smectites (1), glaucony (2), Fe-illite (3), Fe-montmorillonite (4) and nontronite-berthierine (5).

Download figure to PowerPoint

  • 1
    Fe–Al smectites: rapid overgrowth on detrital phyllosilicates at the sediment–sea-water interface. The duration of the reaction is limited by the high sedimentation rate.
  • 2
    Glaucony: the glauconitization process (arrow 1–2 in Fig. 8) is time-dependent: the shorter it is, the lower the amount of potassium incorporated, then the higher the smectite layer content in the mixed layer structures. Glauconitization which is controlled by chemical diffusion, goes to completion only in low sedimentation-rate environments. The continuity between the Fe–Al smectite and the glaucony domains suggests that the smectite forms first and behaves as a ‘nucleus’ on which glauconite Glm or Gli grow.
  • 3
    Fe-illite: it forms preferentially in hypersaline environments where the sedimentation rate is low (arrow 1–3 in Fig. 8). Even if its occurrences are less well documented than the glaucony ones, comparable mixed layered structures have been described. The Fe-illite and Fe–Al smectite composition domains being overlapping, this suggests that both are the components of the MLMs.
  • 4 & 5.
    Fe-montmorillonite and the nontronite-berthierine assemblage: both are formed in low sedimentation-rate environments by a chemical diffusion process. High resolution electron microscopy observations suggest that nontronite can be interlayered with berthierine.

The reducing microsystem

The formation of green clays in marine environments, i.e. glaucony or Fe-rich smectites, is made possible by the migration of K and Fe ions from sea-water or surrounding sediments to a given microsystem centred on organic debris. The crystallization process at low temperature inside the microsystem is facilitated by the presence of Mg2+ ions (Harder, 1978, 1980). Whatever its shape (pellets, worm burrows, grain coatings) or its mineral structure (glauconite, MLM, Fe-smectite), the green material is frequently chemically zoned (Parron, 1989; Gaudin et al., 2005). Both ion migration and zonation suggest that the formation of green minerals depends on chemical diffusion. The amount of K ions is considered as an index of progression of the mica-forming process (Thompson and Hower, 1975). The absence of correlation with Fe amounts (Fig. 3) shows that K and Fe ions are provided to the microsystem by independent processes.

The sources of K and Fe ions are sea-water and detrital minerals, respectively (Fig. 9). Most of the iron contained in detrital minerals is in the Fe3+ state (oxides, Fe-kaolinite). Iron must be in the soluble Fe2+ state to migrate towards the proto-green mineral pellet. Thus, the source of iron is limited to the micro-environment in which organic debris or bacterial activity impose reducing conditions. The Fe is incorporated either in the silicate or the sulphide phases. Gaudin et al. (2005) have shown that framboidal pyrite is formed first in such microenvironments. The silicate phase cannot incorporate iron when sulphide present. It forms when the sulphide is oxidized. To crystallize, Fe-rich smectites and glauconite need a change in the Redox conditions at the scale of the microsystem: oxidation occurs as soon as the organic matter is totally mineralized and the bacterial activity stops (Gaudin et al., 2005). This explains why a part of the Fe2+ ions is oxidized when incorporated into the green grains. Then, glauconite form by a nucleation-crystal growth process (Gago-Duport et al., 2000).

image

Figure 9.  Schematic representation of the formation of green materials in a clay-rich marine sandstone where organic debris are present (black area). Potassium diffuses from sea-water through the sediment–sea-water interface while Fe2+ ions diffuse inside the reductive microenvironment centred on the organic debris.

Download figure to PowerPoint

Potassium chemical diffusion and glauconitization

Odin and Dodson (1982) claimed that ‘Glauconitization may go to completion in close contact with the sea-water. This means that exchanges with the sea-water must be possible during the whole process; such exchanges can only occur in the first centimetres of a mud, in the first decimetres of a sandy mud and the first few metres of pure coarse sand’. These observations indicate that the physical characteristics (porosity, permeability, tortuosity) of the hosting sediment are decisive in the glauconitization process. Considering that glauconitization is controlled by chemical diffusion, the Odin and Dodson's observation suggests that two parameters should be taken into account: the flux of ion species per time (f) and the length of diffusion paths (l). The thickness of the diffusion zone being obviously dependent on the grain size in the sediments, this means that the tortuosity is one of the main parameters to be considered in the chemical process involved in glauconitization of mud, sandy mud and coarse grain sands (Fig. 10).

image

Figure 10.  Schematic representation of the diffusion pathways in different sediments: (a) mud; (b) sandy mud; (c) sand. (d) Relation between the flux of species q (f = dq/dt) with the average length of diffusion paths (l). l* and f*: critical value of l and f, respectively (see text for explanation).

Download figure to PowerPoint

According to Odin and Dodson (1982), one can consider that the progress of the glauconitization reaction is controlled by the diffusion of K+ ions in the solution from the sediment–sea-water interface towards the glaucony grain. The flux of K+ ions can be correctly depicted by the equation from Holchella and Banfield (1995):

  • image(1)

where dq/dt is the flux (f) of species q per time t; β the geometric constant taking into account the porosity or tortuosity; cq the constant that scales the radius of species q; T the absolute temperature; Ci and Ce the concentration of species q in the internal (outer surfaces of glauconite particles) and external (pore water) parts of the system, respectively; A the effective pore area; η the viscosity; and l is the average length of diffusion paths.

Let us take an example: the glauconitization of a given sediment, for instance a coarse grain sand. The effective pore area and the pore geometry do not change much during the whole process. We can approximate that the β and A parameters are roughly constant. Because the glauconitization process occurs near the sea-water–sediment interface, it can be considered as isothermal, thus T is constant. Isothermal and iso-saline conditions mean that the viscosity of sea-water (η) does not change. The driving force of the diffusion is the concentration difference Ci − Ce. This parameter can be considered as constant because the sea-water reservoir is infinite and the concentration of the element q at the glaucony grain is fixed by the physico-chemical conditions at surface of the growing mineral. These considerations allow us to reduce Eq. (1) to:

  • image(2)

where K is a constant.

Equation (2) shows that the flux of species q per time t varies proportionally to the inverse of the average length of diffusion paths. The thickness of the sediment affected by the glauconitization process depending on the mud/sand ratio (Odin and Dodson, 1982) means that there is a critical value for l (l*). When too long, the diffusion pathways reduce the flux to a critical value (f*) which is inefficient for the growth of the glaucony grain under consideration. Inefficient means that the incorporation and liberation rates of the ion species at the surface of glauconite crystals are equivalent. Consequently, the glauconitization in that given microsystem stops. It may continue in another microsystem for which l has not reached l*. This explains the different age of grain populations observed by Smith et al. (1998).

Why glauconitization may stop before completion? Thompson and Hower (1975) have shown that the per cent of expandable layers in glaucony is inversely related to the amount of K ions in the crystal structure. The amount of the diffusing ion can be obtained by integrating the flux f for a time t. If there is no modification of the sediment texture during the glauconitization process, the diffusion should be active until l* is reached. Then, the process stops at a time t* (value of t for f*). The glauconitization process goes to completion as it is in the case of the Gulf of Guinea (Fig. 6) described by Odin and Dodson (1982). However, the process may stop before completion if the flux decreases. Any modification of the porosity during the early diagenetic stage in cemented sandstones changes the composition of the glaucony (Strickler and Ferrell, 1990). Consequently, two phenomena may change the flux: (1) increasing l* by addition of sediments (the interface with sea-water progresses upward), (2) interruption of the diffusion paths due to compaction or cementation.

According to Odin and Dodson (1982), the formation of a glaucony grain having more than 8% of K2O (highly evolved stage) needs several hundred thousands to 1 Myr. From a chemical point of view, this means that the conditions at the micro-environment scale remain stable for those long periods of time. This is hardly understandable. Single-grain 40Ar–39Ar analyses reveal that the glaucony grains in a given sediment have not been formed simultaneously (Smith et al., 1998). The glaucony populations from Cenomanian, Bartonian and Burdigalian sediments exhibit a wide range of ages which suggests that glaucony has been formed in several genetic steps during a long period of time of up to 5 Myr.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Data from the literature
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

In spite of unavoidable uncertainties due to Fe-rich nano-inclusions which could be intercalated between the layers, the illite and glauconite solid solution compositions undoubtedly are extended towards Fe- or Al-rich end-members, respectively. The distribution of Fe ions in the octahedral sheets of illite and glauconite has been investigated through infrared spectroscopy studies and Monte Carlo simulations (Cuadros et al., 1999; Sainz-Diaz et al., 2003a,b).

From a mineralogical point of view, the analyses of the published data suggest that a continuous solid solution exists between the Fe–Al smectite, Fe-illite and glauconite compositional end-members. On the contrary, nontronite and Fe-montmorillonite composition fields are clearly separated from those of glauconite. Whatever their composition, the formation of these minerals is a response to the change of chemical conditions when sediments interact with sea-water or hypersaline solutions. The quickest response is the overgrowth of Fe–Al smectites on detrital clays. When reductive microsystems are formed in the unconsolidated sediments around organic debris, Al-bearing (glauconite, Fe-illite) or Al-depleted (Fe-montmorillonite, nontronite) species begin to form according to the sediment composition. Nucleation and crystal growth are controlled by chemical diffusion of K+ and Fe2+ ions. If not interrupted, the diffusion process is active until reaction completion is reached, i.e. formation of the stable phase (Fe-illite or glauconite) or mineral assemblage (berthierine–nontronite). Intermediate steps are observed as mixed-layer minerals when the diffusion process is interrupted because of sedimentation rate, compaction or cementation.

The mechanism described above can be activated in sediments whatever their depositional environment. Green minerals have been observed from coastal to deep sea conditions and from hypersaline to more diluted waters. Consequently, these minerals, and particularly glauconies are not as good palaeoenvironmental indicators as commonly thought (Huggett and Gale, 1997). They can form during long periods of time (several million years) as attested by the wide range of single-grain ages.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Data from the literature
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

This study has been supported by the UMR 6532 CNRS and Alexander von Humboldt Foundation. The authors warmly thank the scientific editor, Dr Fritz and two anonymous referees for their constructive remarks.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Data from the literature
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • Bailey, S.W., 1980. Summary of recommendations of AIPEA nomenclature committee. Clay Mineral., 15, 8593.
  • Baker, J.C., 1997. Green ferric clay in non-marine sandstones of the Rewan Group, southern Bowen Basin, Eastern Australia. Clay Mineral., 32, 499506.
  • Berg-Madsen, V., 1983. High-alumina glaucony from the middle Cambrian of Öland and Bornholm, southern Baltoscandia. J. Sed. Petrol., 53, 875893.
  • Brindley, G.W., 1982. Chemical compositions of berthierines – a review. Clay Clay Mineral., 30, 153155.
  • Buatier, M., Honnorez, J. and Ehret, G., 1989. Fe-smectite–glauconite transition in hydrothermal green clays from the Galapagos spreading center. Clay Clay Mineral., 37, 532541.
  • Buckley, H.A., Bevan, K.M., Brown, M., Johbson, L.R. and Farmer, V.C., 1978. Glauconite and celadonite: two separate mineral species. Mineral. Mag., 42, 373382.
  • Cimbálníková, A., 1971. Chemical variability and structural heterogeneity of glauconites. Am. Mineral., 56, 13851398.
  • Cuadros, J., Sainz-Diaz, C.I., Damirez, R. and Hernandez-Laguna, A. (1999) Analysis of Fe segregation in the octahedral sheet of bentonitic illite-smectite by means of FTIR, Al-27 MAS NMR and reverse Monte Carlo simulations. Am. J. Sci. 299, 289308.
  • Curtis, C.D., Hughes, C.R., Ireland, B.J., Warren, E.D., Whiteman, J.A. and Whittle, C.K., 1987. Analytical transmission electron microscopy in the study of diagenetic clay minerals. Mineral. Mag., 51, 123124.
  • Deconninck, J.F., Strasser, A. and Debrabant, P., 1988. Formation of illitic minerals at surface temperatures in Purbeckian sediments (Lower Berriasian, Swiss and French Jura). Clay Mineral., 23, 91103.
  • El Albani, A., Meunier, A. and Fürsich, F., 2005. Unusual occurrence of glauconite in a shallow lagoonal environment. Terra Nova, 17, 537544.
  • Foster, M.D., 1969. Studies of celadonite and glauconite. Geol. Surv. Prof. Pap., 614-F, 17.
  • Gago-Duport, L., Fernandez-Bastero, S., Pimentel, F., Villar, P., Santos, A., Serra, C. and Vilas, F., 2000. Glauconite nucleation in silica tubular microstructures from low-temperature solution experiments. Goldschmidt 2000. J. Conf. Abst., 5, 418419.
  • Gaudin, A., Buatier, M.D., Beaufort, D., Petit, S., Grauby, O. and Decarreau, A., 2005. Characterization and origin of Fe3+-montmorillonite in deep-water calcareous sediments (Pacific Ocean, Costa Rica margin). Clay Clay Mineral., 53, 452465.
  • Giresse, P., Wiewiora, A. and Lacka, B. 1988. Mineral phases and processes within green peloids from the recent deposits near the Congo river mouth. Clay Mineral., 23, 447458.
  • Harder, H., 1978. Synthesis of iron layer silicate minerals under natural conditions. Clay Clay Mineral., 26, 6572.
  • Harder, H., 1980. Synthesis of glauconite at surface temperatures. Clay Clay Mineral., 28, 217222.
  • Hendrix, S.B. and Ross, C.B., 1941. Chemical composition and genesis of glauconite and celadonite. Am. Mineral., 26, 683708.
  • Holchella, M.F., Jr and Banfield, J.F., 1995. Chemical weathering of silicates in nature: a microscopic perspective with theoretical considerations. In: Chemical Weathering Rates of Silicate Minerals (A.F.White and S.L.Brantley, eds). Mineralogical Society of America. Rev. Mineral., 31, 353406.
  • Huggett, J.M. and Gale, A.S., 1997. Petrology and palaeo environmental significance of glaucony in the Eocene succession at Whitecliff Bay, Hampshire Basin, UK. J. Geol. Soc., 154, 897912.
  • Ireland, B.J., Curtis, C.D. and Whiteman, J.A., 1983. Compositional variation within some glauconites and illites and implications for their stability and origins. Sedimentology, 30, 769786.
  • Mendes Guimarães, E., Velde, B., Hillier, S. and Nicot, E., 2000. Diagenetic/anchimetamorphic changes on the Proterozoic glauconite and glaucony from the Paranoà group, mid-western Brazil. Rev. Brasil. Geocienc., 30, 363366.
  • Meunier, A. and Velde, B., 1989. Solid solution in illite/smectite mixed layer minerals and illite. Am. Mineral., 74, 11061112.
  • Meunier, A. and Velde, B., 2004. Illite. Origins, Evolution and Metamorphism. Springer, Berlin, 286 pp.
  • Michalopoulos, P. and Aller, R.C., 1995. Rapid clay mineral formation in Amazon delta sediments: reverse weathering and oceanic elemental cycles. Science, 270, 614617.
  • Newman, A.C.D. and Brown, G., 1987. The chemical constitution of clays. In: Chemistry of Clays and Clay Minerals (A.C.D.Newman, ed.). Mineral. Soc. Monogr., 6, 1128.
  • Odin, G.S. (ed.), 1988. Green Marine Clays. Oolithic Ironstone Facies, Verdine Facies, Glaucony Facies and Celadonite-bearing Facies – A Comparative Study. Elsevier, Amsterdam, 445 pp.
  • Odin, G.S. and Dodson, M.H., 1982. Zero isotopic age of glauconies. In: Numerical Dating in Stratigraphy (G.S.Odin, ed.), Chapter 14, pp. 277305. Wiley Interscience, New York.
  • Parron, C. 1989. Voies et mécanismes de cristallogenèse des minéraux argileux ferrifères en milieu marin. Le processus de glauconitization: évolutions minérales, structurales et géochimiques. Thesis. University of Aix-Marseille, 411 pp.
  • Pédro, G., Carmouze, J.P. and Velde, B., 1978. Peloidal nontronite formation in recent sediments of lake Tchad. Chem. Geol., 23, 139149.
  • Rousset, D., Leclerc, S., Clauer, N., Lancelot, J., Cathelineau, M. and Aranyossi, J.F., 2004. Age and origin of Albian glauconites and associated clay minerals inferred from a detailed geochemical analysis. J. Sed. Res., 74, 631642.
  • Sainz-Diaz, C.I., Palin, E.J., Nernandez-Laguna, A. and Dove, M.T., 2003a. Octahedral cation ordering of illite and smectite. Theoretical exchange potential determination and Monte Carlo simulation. Phys. Chem. Miner., 30, 382392.
  • Sainz-Diaz, C.I., Palin, E.J., Dove, M.T. and Nernandez-Laguna, A., 2003b. Ordering of Al, Fe and Mg cations in the octahedral sheet of smectites and illites by means of Monte Carlo simulations. Am. Mineral., 88, 10331045.
  • Smith, P.E., Evensen, N.M., York, D. and Odin, G.S., 1998. Single-grain 40Ar/39Ar ages of glauconies: implications for the geologic time scale and global sea level variations. Science, 279, 15171519.
  • Steinberg, M., Holtzapffel, T. and Rautureau, M., 1987. Characterization of overgrowth structures formed around individual clay particles during early diagenesis. Clay Clay Mineral., 35, 189195.
  • Strickler, M.E. and Ferrell, R.E., Jr, 1990. Fe substitution for Al in glauconite with increasing diagenesis in the first Wilcox sandstone (lower Eocene), Livingston Parish, Louisiana. Clay Clay Mineral., 38, 6976.
  • Thompson, G.R. and Hower, J., 1975. The mineralogy of glauconite. Clay Clay Mineral., 23, 289300.
  • Velde, B., 2003. Green clay minerals. In: Treatise on Geochemistry, Volume 7, Sediments, Diagenesis, and Sedimentary Rocks (F.T.Mackenzie, ed.), pp. 309324. Elsevier, Amsterdam.
  • Weaver, C.E. and Pollard, L.D., 1973. The Chemistry of Clay Minerals. Elsevier, Amsterdam, 213 pp.
  • Wiewiora, A., Lacka, B. and Giresse, P., 1996. Characterization and origin of 1:1 phyllosilicates within peloids of the recent, Holocene and Miocene deposits of the Congo Basin. Clay Clay Mineral., 44, 587598.
  • Wiewiora, A., Giresse, P., Jaunet, A.M., Wilamoswski, A. and Elsass, F., 1999. Crystal chemistry of layer silicates of the Miocene green grain (Congo Basin) from transmission electron microscopy (TEM) and analytical electron microscopy (AEM) observations. Clay Clay Mineral., 47, 582590.