Abstract

Detrital, volcanic and diagenetic origins have been used to explain the smectite clay assemblage that characterizes the Upper Cretaceous Chalk of Europe. To further the understanding of how clays of different origins may have converged to this characteristic clay mineral assemblage a new approach is put forward for their investigation. This is based upon (1) the correlation that exists between the trace element and stable isotope geochemistry of the calcite cements preserved within Chalk brachiopods and the various diagenetic phases of early lithification and cementation recognized in the Chalk, and (2) an understanding of the process of late diagenetic cementation that has caused regional differences in the hardness of the Chalk. It is suggested that each phase of lithification and associated calcite cementation may preserve the different clay assemblages at various stages in their convergence to the characteristic Chalk smectite assemblage.

The Upper Cretaceous Chalk is a thick (up to 2000 m) and exceptionally pure fine-grained marine limestone that occurs widely in northern Europe. It is the main water, oil and gas reservoir and it is the preferred substrate for many major engineering projects. The Chalk’s physical consistency ranges from a soft friable uncemented chalky sediment to a dense fine-grained limestone. The understanding of its nature and properties is of considerable economic and scientific significance. The clay contents and the nature of the clay are particularly important in the development of new Chalk reservoirs and for its engineering properties (Hardman, 1982; Mortimore, 2012). In spite of the important role played by clay minerals, our knowledge is limited to generalities and restricted qualitative investigations. This reflects not a lack of interest but the inherent difficulties of dealing with the Chalk as a material – the very small size of the clay crystals (generally less than 0.5 μm) and the frequently friable, uncemented and fine-grained nature of the host Chalk which consists predominantly of very fine silt (2–4 μm) and clay-grade (<2 μm) low-magnesium calcite crystals derived from coccoliths, the scales of coccospheres having been secreted by the coccolithophoridae that lived as plankton in the photic zone of the Upper Cretaceous Chalk sea.

The Chalk Smectite Question

Millot (1949) was first to consider that there was something rather special about the clay assemblage in the Chalk. Millot et al. (1957), Duplaix et al. (1960) and Millot (1964) suggested that a montmorillonite-dominated clay assemblage sometimes with a little mica characterized the Chalk of the Paris Basin and was of neoformational origin. Thus the problem of the Chalk montmorillonite was born. In this discussion it will be referred to as the Chalk smectite assemblage. Subsequent research has largely confirmed its widespread nature (Deconinck et al., 2005; Jeans, 2006) except where local detrital sources have contributed kaolin- and mica-rich clay material (e.g. Heim,1957; Jeans, 2006, fig. 46). Evidence for the neoformational origin is more limited as direct petrographic evidence for the development of the smectite crystals are limited either to studies by transmission electron microscopy that may show euhedral lath-shaped crystal forms as well as evidence of euhedral overgrowths on detrital cores (e.g. Weir & Catt, 1965), or to studies by scanning electron microscopy of silicified patches where bushes of neoformed crystals and opal-CT hemispheres have grown on the walls of pores which have then been coated by a diffuse cloud of very fine clay crystals (e.g. Jeans, 1978). Various authors have suggested that some of the smectite is of volcanic origin, or is detrital derived from smectite-rich soils related to the rising sea levels and water tables as the Upper Cretaceous sea transgressed and drowned low-lying landmasses. There is clear evidence of more or less argillized volcanic ash and phenocrysts (Valeton, 1959, 1960; Schöner, 1960; Dorn & Bräutigam, 1959; Bräutigam, 1962). The argillized volcanic ash is usually in the form of thin marl seams of wide extent, characterized by europium anomalies in their rare earth element patterns (Wray, 1995, 1999; Wray & Wood, 1995, 1998; Wray et al., 1995, 1996). Many other marl seams of wide extent in the Chalk contain no such evidence and are considered to be of detrital origin (Wray & Wood, 1998); however, they could represent the argillization products of unevolved magmas in which europium has not been depleted by its selective uptake in the early crystallization of feldspar. To what extent are the fine euhedral clay laths and overgrowths illustrated by Weir & Catt (1968) the direct precipitate from the dissolution of ash in the porewaters of the sediment is an open question. Deconinck & Chamley (1995) have differentiated between volcanic and detrital smectite in marl seams from the Turonian chalks of northern France based upon their behaviour when heated. Lindgreen et al. (2002, 2008) have utilized detailed structural analysis of the smectite assemblages from the Chalk of Denmark and the North Sea to postulate various sequences of modification related to diagenesis. It is possible that all these origins for the Chalk smectite assemblage have played a role, but it does not answer the question of how the clays from these different sources have ended up as this apparently monotonous and widespread clay assemblage that has characterized the Chalk over the 30 –40 Ma years during its deposition and the subsequent 60 million years of diagenesis.

This paper outlines a novel method by which this variation in clay mineralogy can be studied within an independent diagenetic framework by making use of recent advances in the understanding of the cementation and lithification of the Chalk. A similar approach has been used with carbonate nodules (e.g. Raiswell, 1971, 1976; Dickson & Barber, 1976; Hendry et al., 2006) and siliceous nodules (e.g. Jeans 1978; Jeans et al., 1977, 1997) to unravel the sequence of mineral changes that occur during the diagenesis of fine-grained sediments. Sufficient is now known about the cementation and lithification (see Jeans, 1980, p. 82 for use of these terms) in the Cenomanian Chalk of eastern England to treat it as a giant nodule which has developed initially through microbial influence and overpressuring followed by a loss of hydrostatic pressure causing pressure dissolution, the widespread precipitation of a late diagenetic cement, lithification and brittle fracturing. This late calcite cement was responsible for the regional hardening of the Chalk. A subsequent paper in this issue will discuss how the variations and patterns in clay mineral assemblages may be better understood by relating them to the different types and timings of lithification.

The exceptionally fine-grained nature of the Chalk has restricted the direct study of the geochemistry of the cements involved in its lithification in spite of advances in instrumentation (e.g. Maliva & Dickson, 1997). The reason is that sufficiently large crystals of calcite cement are not generally available to determine the history of cementation. Recently this was achieved in the Cenomanian Chalk of eastern England by investigating the calcite cement infilling vugs within the shell cavity of terebratulid brachiopods (Hu et al., 2012). Figure 1 shows such an example. It could be argued that these calcite-filled vugs are the result of allochthonous porefluids that have infiltrated the Chalk at some stage during its diagenesis and have managed to penetrate the well preserved brachiopods with their tightly fitting valves. However, evidence outlined below demonstrates that the calcite infillings are the result of precipitation from the Chalk’s autochthonous pore waters. The pattern of trace elements in the calcite cement in the vugs can be matched with the bulk geochemistry of different types of lithification in the main mass of the Chalk. This will be discussed in more detail later in this paper.

Locations mentioned in the text are shown in Fig. 2. Stratigraphical terms used in the description of the Cenomanian Chalk of eastern England are shown in Fig. 3. Sample horizons are shown in Fig. 4 (Speeton) and Fig. 5 (Hunstanton and Stenigot).

Geochemistry of the Brachiopod Calcite Cement

The calcite-filled vugs preserved in these terebratulid brachiopods are probably associated with individuals that had been buried alive within the chalk sediment. In such circumstances their valves would have been tightly shut and it was only after death when the musculature holding the valves together started to decompose that some sediment found its way into the body cavity leaving a considerable pore-fluid filled void into which grew large crystals of calcite cement. Cathodoluminescent imagery, staining, electron microprobe and stable isotope analysis of polished sections (methods of analysis are given in Hu et al., 2012) show that the calcite crystals preserve a record of the geochemistry of the cements which is related to the early diagenesis of the chalk sediment. Figure 6a,b shows the cathodoluminescent imagery from brachiopod T1 (also see Fig. 1).

Two well defined patterns of calcite cementation based on their geochemistry are recognized. The suboxic series is associated with brachiopods from the coloured chalks at Speeton (Fig. 7) which contain a fine-grained hematite pigment (its precursor is assumed to be Fe(OH)3). This displays a pattern (Fig. 8) starting with Mg-rich, followed by Mn-rich and finally Fe-rich cements with increasingly high δ13C values. The other cement pattern, the anoxic series, is found in brachiopods associated with the development of glauconite-stained submarine hardgrounds from chalks that originally contained the precursor Fe(OH)3 but no longer do so. The trace element pattern is distinctive (Fig. 9) starting with a Mg-rich phase (zone A) followed by zones (B-F) enriched in Mn showing antipathetic variations in Fe and Mn, then zones (F to H) dominated equally by Fe and Mn – the youngest zones (I, J) have very low concentrations of trace elements. Carbon isotope values become increasingly negative as cementation progressed. Figure 10 summarizes the relationship between these two patterns and how they are related to the early lithification of the chalk sediment (Hu et al., 2012).

Chalk Lithification and its Relationship to the Brachiopod Cement Stratigraphy

In order to use the brachiopod cement stratigraphy to establish the diagenetic development of the smectite clay assemblage it is necessary to show that the various phases of lithification recognized by Jeans (1980) in the Cenomanian Chalk in eastern England and the Campanian Chalk at Peacehaven in Sussex can be related to what has been observed in the brachiopods. Only then is there a basis to obtain samples for clay mineral analysis that will reflect this diagenetic framework.

Four different types of early lithification have been recognized by their field occurrence and Fe geochemistry of their bulk calcite in the Cenomanian Chalk of eastern England (Jeans, 1980). A late diagenetic lithification was also identified in this Chalk where it had not been affected by early lithification – this was related to regional pressure dissolution. Additional trace element (Fe, Mn, Mg, Sr) and stable isotope (δ18O and δ13C) analysis of the bulk calcite of chalks affected by these different types of lithification have been carried out using the methods described in Jeans et al. (1991, 2012) in order to identify (a) the geochemical fingerprint of their cements above the background composition of the original bioclastic material, and (b) similarities with the pattern of trace elements found in the calcite filled vugs of the brachiopods.

Large ammonites at Speeton (Jeans, 1980, Type 4 lithification)

The preferential calcite cementation associated with large ammonites preserved in the nodular chalks is considered to represent the earliest lithification as it is related to what must have been their rotting remains within their aragonite shell buried in the chalk sediment. There is, however, no systematic difference in the stable isotope values between the ammonites and their nodular chalk surrounds (Jeans et al., 2012, table 4). Table 1 contrasts the trace element concentrations of the bulk calcite associated with a number of large ammonites with that of the surrounding nodular chalk. The ammonite calcite displays enhanced values of Mg and Mn relative to the nodules, and also of Fe relative to the matrix of their nodular surrounds. Comparison to the suboxic pattern of cementation indicates that the Mg-rich, Mn-rich and Fe-rich phases of brachiopod cement are represented.

Nodular chalks at Speeton (Jeans, 1980, Type 5 lithification)

Nodular chalks are a major component of the Red Chalk and Ferriby formations at Speeton (Fig. 4). Jeans et al. (2012, table 5) have shown that there are no systematic differences in the stable isotope values between the bulk calcite from the nodules and matrices making up the nodular chalks. Tables 2 and 3 contrast the trace element contents of the bulk calcite in two series of nodular chalk from different members of the Ferriby Formation at Speeton. The Fe content of the nodule calcite is always enhanced relative to the matrix, whereas the Mg and Mn contents show no consistent pattern. Comparison to the suboxic pattern suggests that this early cementation represents only the late Fe-rich phase in the brachiopod cement and was not affected by the earlier Mg- and Mn-enriched cements.

Late lithification associated with pressure solution cements at Speeton

At Speeton, regional cementation due to pressure solution has been shown to postdate the completion of suboxic cementation (Hu et al., 2012). Table 4 shows the δ18O and δ13C values of the residual calcite from a series of pressure dissolution marl seams and the bulk calcite of the adjacent beds of chalk in the Louth Member (Ferriby Formation). The marl seams not only have considerably higher acid insoluble residues but they have higher δ18O and δ13C values. The values from the marl seams are considered to represent the original chalk bioclasts without addition of any calcite cement; however, the original bulk δ13C value has been modified by the selective pressure dissolution of the finest calcite fraction (<2 μm) consisting of coccolith crystals which usually have lower δ13C values than bioclastic material (Jeans et al., 1991, fig. 8). The lower δ13C values in the chalk beds reflect the inheritance of the coccolith HCO3 from the marl seams, whereas the lower δ18O values suggest an enhanced temperature of cementation. This late-stage cementation will therefore display a δ13C trend towards more negative values.

Paradoxica Bed hardground at Hunstanton (Jeans, 1980, Type 1 lithification)

The development of this hardground lithification (Fig. 11) reflects a regional event, in this instance the whole area of the East Midland Shelf from north Norfolk to south Yorkshire (Fig. 3). The hardground is related to a long established surface of nil or negative deposition marking the top of the Belchford Member of the Ferriby Formation that affected the underlying chalks containing Fe(OH)3, the precursor to the hematite pigment responsible for the red or pink colouration of much of the Cenomanian Chalk in eastern England (Fig. 7). The sediment underlying the surface was extensively colonized and bioturbated by an infauna, the Fe(OH)3 having been lost and all the surfaces of the resulting Paradoxica Bed hardground that were in contact with ambient seawater are stained with glauconite (Fig. 12). The development of anoxic conditions, cementation and dissolution of the Fe(OH)3 took place while the sediment was penetrated by an open reticulate system of branching burrows occupied by callianassid crustaceans (Jeans, 1980, fig. 9). The trace element chemistry of the bulk calcite from different levels within the Paradoxica Bed at Hunstanton and Stenigot is shown in Table 5. The enrichment in Mg towards the top surface of the bed at Hunstanton is noteworthy, probably reflecting the increasing amount of the early Mg-enriched cement that is seen in brachiopods T12–T16 (Fig. 9).

Incipient hardgrounds, Campanian Chalk, Peacehaven, Sussex

Another variety of early regional lithification which is related to anoxic calcite cementation occurs at a much higher stratigraphical level in the Chalk of southern England. It may be much more widespread both stratigraphically and regionally, but this is uncertain as it is not readily identifiable in the field. This is known from the sequence of variably cemented white chalk with flints and thin marl seams of Campanian age (stratigraphical details are shown in Jeans et al., 2012, fig. 13) exposed at the Peacehaven Steps Cliff section on the Sussex coast (Fig. 13). There is no evidence of hardground development or of glauconite staining. Rusty patches and stains to fossils suggest the former presence of iron sulfides. Stable isotope analysis of the bulk calcite of 65 samples collected at 25 cm interval shows covariation between the δ18O and δ13C values (Fig. 14), suggesting that two end-member chalks of different stable isotope composition were involved, one with relatively high δ18O and δ13C values, the other with lower values. Further investigation (Jeans et al., 2012) has confirmed this. Very little difference was present in the δ18O and δ13C values of the separated sand (>63 μm) and silt fractions (2–63 μm) whereas the <2 μm fraction δ18O values ranged from –2.0 to –5.4‰ and δ13C values from 1.4 to –8.3‰. The <2 μm fraction of the topmost thirteen samples (PHS53-PHS65) from the section were also analysed but displayed δ18O (–1.7 to –2.0‰) and δ13C (1.6 to 2.3‰) values no different from the bulk. Figure 15 is a δ18O/δ13C cross plot on which all the analyses from the bulk chalk and the <2 μm fractions have been plotted. It demonstrates that the isotopic covariance originally noted is the result of varying mixtures of normal chalk (δ18O, –1.5 to –2.5‰; δ13C, 2.5‰) with a fine-grained cement with approximate values of δ18O –8‰ and δ13C –8‰.

Conclusions

The identification of different phases of lithification within the Chalk and the geochemistry of their calcite cements provides a basis by which various stages in the development of the smectite clay assemblage can be investigated. This is summarized in Fig. 16.

Suboxic cementation

At Speeton, where chalk deposition was essentially continuous and reactions between ambient seawater and the contents of the sediment must have been minimal, five different stages of diagenesis can be used as a time framework to investigate the variation in the smectite assemblage. The earliest stage in the evolution of the smectite assemblage may be preserved in the chalk associated with large ammonites where the full pattern of suboxic cementation is retained. The next stage is that associated with the nodules in nodular chalks which display only the final Fe-rich phase of cementation. A later stage in the development of this assemblage at Speeton may be reflected in the sediment little affected by early cement but influenced by the movement of compaction pore waters migrating upwards. A further stage could be preserved in the chalk unaffected by early lithification but which underwent an extended phase of overpressuring before being cemented and lithified by pressure dissolution. An even later phase may be recorded in the differences that are present between the pressure dissolution marl seams and the adjacent chalk undergoing regional calcite cementation.

Anoxic cementation

The Paradoxica Bed hardground and its crustacean burrow system is likely to reveal more about the original clay assemblage at the time of deposition and how this has been modified by the long-term interaction of ambient seawater with the anoxic pore fluids and unstable siliceous and carbonate debris within the sediment. The staining by glauconite of all the surfaces of this hardground in contact with the Chalk sea gives an indication of the extent of the transformations and neoformations that may have occurred in the development of the Chalk’s smectite assemblage.

The Campanian sequence of chalk, flint and thin marls at Peacehaven displays cryptic anoxic cementation that must have developed some way below the water/sediment interface. The variable degrees of cementation and the variation in the heating behaviour of the smectite assemblage (Jeans et al., 2014) could be related to the development of smectite under changing cation populations associated with periods of anoxic cementation.

We wish to thank the following: Vivien Brown and Philip Stickler respectively for help in preparing the manuscript and figures; Michael Hall and James Rolfe for stable isotope analysis; Chiara Petroni for electron microprobe analysis; Rory Mortimore for much stratigraphical guidance and help in sample collecting at Peacehaven as well as for the aerial photograph; Tony Dickson for his unrivalled carbonate expertise and unstinting help; Tony Fallick, Alex Brasier and Mark Wilkinson for their helpful comments and suggestions;

the Chinese Science Council and the Department of Earth Sciences respectively for financial support and hospitality during X.F. Hu’s visit (2009–10) to Cambridge.