- © The Mineralogical Society
Scanning electronic microscopy (SEM) and transmission electronic microscopy (TEM) investigations of chlorite grain coatings from four different sandstone reservoirs indicate a progressive change in both texture and arrangement of chlorite particles from the contact with the detrital substratum to the centre of the pore. Such spatial distribution results from growth by geometrical selection. Geometrical selection of chlorite crystals proceeded during a single event of continuous growth which began before the consolidation of the sandy sediments, lasted through part of the subsequent mechanical compaction and ceased before the occurrence of quartz cement. Nanopetrographic investigations near the detrital quartz-chlorite coating interface demonstrate that inhibition of the quartz cement is due to the limitation of the epitaxial growth of quartz to the interparticular space at the base of the chlorite coating and not an absence of nucleation. It is suggested that these results can be applied to most of the sandstones which contain Fe-rich chlorite grain coatings.
Authigenic chlorites are well known for their major influence on the reservoir quality of deeply buried sandstones. Numerous published works have shown that diagenetic chlorites, most often the Fe-rich varieties, tend to inhibit the formation of quartz overgrowths in the primary pores of sandstones (Pittman & Lumsden, 1968; Dutton, 1977; Thomson, 1979; Tillman & Almon, 1979; Larese et al., 1984, Dixon et al., 1989, among others) and consequently have a determinant role in the preservation of unusually high primary intergranular porosity in deeply buried reservoirs (Ehrenberg, 1993). Even if a relation between the occurrence of chlorite and the preservation of porosity in deeply buried sandstone reservoirs is now undoubtedly established, the mechanisms by which chlorite contributes to the preservation of porosity are still debated.
Despite the fact that some authors proposed chemical considerations (i.e. undersaturated solution chemistry vs. quartz) as a potential cause of the absence of quartz overgrowth in chloritized sandstones, many have shown that the development of quartz overgrowths in the pore space of chloritized sandstones was inhibited because of physical considerations. The mechanism by which chlorite blocks quartz growth is unclear but it has commonly been assumed that the pore-lining chlorites were formed prior to the quartz cementation stage (Hillier, 1994) and have isolated detrital surfaces from contact with pore-water, thus preventing crystallization of authigenic quartz. Several hypotheses were proposed to explain the absence or the scarcity of secondary quartz in the chloritized sandstones: (1) quartz overgrowths did not occur because pressure solution was retarded (Pittman & Lumsden, 1968); (2) the conditions were favourable for the crystallization of quartz but chlorite coatings on the surface of clastic grains significantly inhibited the growth of secondary quartz crystals (Heald & Larese, 1974; Imam, 1986); (3) chlorite coatings inhibited the growth of quartz because the surfaces of the clastic quartz grains were isolated from pore-water and in this way the nucleation of secondary quartz was not possible (Ehrenberg, 1993).
In this paper, samples of chloritized sandstones from four reservoirs having suffered different conditions of burial diagenesis have been investigated in order to ascertain how the growth of chlorite crystallites has proceeded on the detrital grain substratum and by which mechanism the crystallites of chlorite inhibited the subsequent development of quartz overgrowths. The conclusions of this work were supported by multi-scale petrographic data including transmission electron microscope (TEM) analyses of the relations between quartz and chlorites at nanoscopic scale and the microtexture of the chlorite particles.
MATERIALS AND METHODS
Samples of chloritized medium to coarse-grained sandstones have been collected in drill cores from six wells located in four reservoirs whose main characteristics are summarized in Table 1⇓. These sandstone reservoirs are located in Oman (Malih X1), Gabon (M’Bya), Norway (Haltenbanken) and Syria (Kahar). They differ in age of sedimentation, depositional environment, burial history (maximal burial depth), bulk mineralogy (Fig. 1⇓) and nature of the resident pore-fluid. Despite their different geological settings, all these reservoirs have been subjected to an extensive formation of chlorite coatings around the clastic grains during the diagenesis. In all the studied samples, quartz cementation is absent or very poorly developed even at microscopic scales.
Thin sections of each sample of chloritized sandstones were made and examined using an optical polarizing microscope (Olympus BH2).
Scanning electron microscope (SEM) observations were made on small and freshly fractured core samples coated with gold using a JEOL 6320F equipped with an energy dispersive spectrometer. The SEM observations, in particular, were used for the textural characterization of chlorite particles (size and morphology) and for petrographic relations at the micrometric scale.
Transmission electron microscope (TEM) observations were made on 17 samples from all the reservoirs in order to observe at a nanometric scale, the mutual relations between chlorite and quartz (clastic and secondary) near the interface of the chlorite coating and clastic quartz substratum centred on zones that are rich in chlorite-coated grains. The nanopetrographic study by TEM of the chloritized sandstone requires a specific preparation of the sample. Indeed, the examination by TEM of authigenic clays in situ on their clastic substratum is very difficult to perform because pore-filling clay minerals and resin are generally destroyed before adequate thinning of the selected site by ion milling. In order to avoid this problem, core samples were impregnated with a mixture of resin with micrometric grains of corundum and standard thin sections were prepared with lakeside resin. After petrographic study, the zones of interest determined were encircled and isolated from the thin section by erosion. Standard single-hole copper grids were glued on the remaining disks, and the disk-grid assemblies were detached from the glass slide substrate by gentle local heating. Repeated washing in absolute ethanol removed excess Lakeside resin prior to ion thinning with a Gatan 600 Duomill machine (5 kV, 0.25 mA per gun, 12°, and finish tension of 3.5 kV over 3 min) working at room temperature. A thin (150–200 Å) carbon coating completed the sample preparation. The TEM observations were performed with a JEOL 2000 FX electron microscope working at 200 kV with a resolution of 2.85 Å.
When examined under a transmitted-light optical microscope, the samples from the different reservoirs show very similar features of chloritization which can be generalized as follows. Most of the chlorite cement appears as grain coating independent of the grain-size of the clastic mineral (Fig. 2a,b⇓). The thickness of the chlorite grain coating varies from one site to another; however, it is always thicker (5–10 μm) at the wall of the pore spaces where particles present a typical radial arrangement than at the boundary between adjacent grains (<2 μm). Minor chlorite-rich pellets with ooidal coating of concentrically laminated chlorite (Fig. 2c⇓) accompany locally the grain-coating features and suggest a detrital origin of at least a part of the clay material. In all samples, quartz cements are very scarce and they occur only as disseminated quartz overgrowths (Fig. 2d⇓) locally developed in continuity with the detrital quartz grain substratum in zones where the coating of chlorite is ‘optically’ discontinuous.
In summary, the petrographic observations suggest that the chlorite coatings formed by crystals of Fe-rich chlorite, largely predominate over other diagenetic mineralizations. These chlorite coatings are also present at the punching between detrital quartz grain and quartz overgrowths (Fig. 2b⇑). The relevant scale for all observations of grain-coating chlorites is that of the pore. At this scale, two different types of sites may be distinguished on the basis of both texture and organization of chlorite coatings around the clastic grains (Fig. 3⇓): (1) A sites corresponding to the contacts between adjacent grains and (2) B sites corresponding to the walls of the primary pores.
All the observations performed at smaller scales with SEM and TEM and presented in the following sections will refer to these two types of sites.
Arrangement and microtextural and chemical properties of the chlorite particles in the coatings
Observations by SEM confirm that, in all the studied sandstones, chlorite coatings carpet the clastic minerals. The fact that, in certain places, chlorite rims form hollow structures (Fig. 4a⇓) which outline the original grain boundary of dissolved grains (feldspars?) confirms the early origin of the chloritization process.
Both arrangement and microtexture of the particles of chlorite in the grain coatings change dramatically according to the type of site considered (Fig. 4b,c⇑). In the A site, the particles of chlorite formed at the contact surface between adjacent grains of clastic quartz have very homogeneous textural characters. They appear as xenomorph flakes with an average diameter of <1 μm and are arranged parallel to the surface of the detrital grains (Fig. 4d⇑). Due to their very small thickness, these chlorite particles are often invisible in optical microscopy.
In the B site, the thickness of the clay coating averages 5 to 10 μm and the chlorite particles constituting these clay coatings have contrasted textural characters (Fig. 4e⇑).
Both the morphology and the size of chlorite particles differ strongly from detrital grain edges to the centre of the pore. The SEM observations of a few chlorite coatings which have been pulled off during the sample preparation allow us to visualize cross-sections perpendicular to the coating surface (Fig. 4f⇑). In these cross-sections, it can be seen that at the base of the clay coating (i.e. at the contact with the clastic substratum), the chlorite particles are very fine grained and show arrangement and textural characters similar to those observed in the A site (i.e. xenomorph flakes with an average diameter <1 μm which are arranged parallel to the surface of the mineral substratum).
Near the centre of the pore, the chlorite crystals are coarse grained, oriented perpendicular to grain surfaces (radial orientation), and composed of euhedral subhexagonal plates with a face-to-edge or a face-to-face morphology (Fig. 4c⇑). It should be noted that the above characters corresponding to those classically reported for diagenetic chlorites in literature are in fact the characters of the surface of the outer side of the chlorite coatings and not representative of the whole clay coating. The average diameter and thickness of the plates of chlorite tend to be relatively constant in each reservoir but variable from one reservoir to another (Table 2⇓). The coarsest authigenic chlorite particles have been observed in the sandstones of the Kahar reservoir (diameter of 10 μm for a thickness of 1 μm) whereas the smallest ones were observed in the sandstones of the M’Bya reservoir (diameter 3–4 μm for a thickness of 0.1 μm).
In summary, SEM observations indicate that fine-sized particles of chlorites with very similar texture and arrangement were formed in the A site and at the base of the chlorite coating in the B site. On the other hand, coarse-grained euhedral crystals of chlorite have grown toward the centre of the primary pores only in B sites. However, at the SEM scale the relations between the two types of chlorite particles determined in the B site, on both sides of the chlorite coatings, remains unclear (Fig. 4f⇑). They will be easily characterized at smaller scale by use of transmission electron microscopy.
The TEM observations indicate that, in the A site, the chlorite coatings are composed of very thin particles of chlorite which constitute some packets with a preferred orientation tangential to oblique with respect to the edge of the detrital quartz substratum (Fig. 5a⇓). The size and the width of the chlorite packets ranges between 100 and 900 nm and between 10 and 30 nm respectively (Fig. 5b⇓).
A TEM observation of a cross-section through the detrital quartz substratum and its overlying chlorite grain coating in a B site is presented in Fig. 5c⇑. This image does not permit examination pf the full section of chlorite grain coating. However, it does illustrate the progressive change in both arrangement and texture of the chlorite particles between the contact with the substratum and >1 μm further toward the centre of the pore. It can be seen that most of the chlorite particles observed near the surface of the detrital grains are parallel or slightly oblique to the detrital grain surface and that their size does not exceed a few hundred of manometers. It can also be seen that, with increasing distance from the detrital substratum, the chlorite particles progressively coarsen and become oriented perpendicularly to the surface of the detrital grains. In such a profile, it is particularly interesting to note that at any distance of the detrital grain surface, the thickest particles are always perpendicular to the detrital grain surface (Fig. 5d⇑).
The AEM analyses of chlorite particles from the A and B sites of all the studied samples agree with those of Fe-rich trioctahedral chlorites (Table 3⇓). The main compositional variation from one reservoir to another is expressed by a change in Fe/Mg ratio (Billault, 2002). For each studied reservoir, the AEM analyses indicate no compositional variation between chlorite particles from either the A site or the B site (Fig. 6⇓).
Relations between chlorite particles and quartz overgrowths
Quartz overgrowths are rare in the sandstones cemented by chlorite and they occurred in zones where the coating of chlorite appears discontinuous at the scale of the optical microscopy. Features of crystallization of chlorite on a secondary quartz substratum have never been observed. Inversely, SEM observations indicate that when secondary quartz is associated with chlorite, the chlorite grain coating preceded the precipitation of the quartz overgrowth which presents many features of growth stopping at the contact of the authigenic crystals of chlorite which constitute the outer side of the clay coatings (Fig. 7⇓). An example of stopping of the growth feature of secondary quartz crystals at the contact with chlorites particles is presented in Fig. 4f⇑: the faces of a quartz overgrowth have ceased to grow or changed growth direction at the contact point with the euhedral plates of chlorites which constitute the outer part of the grain-coating material.
Additional data concerning the relations between chlorite particles and quartz overgrowths are provided by the TEM analysis of the interface between detrital grains and chlorite coating. The surfaces of the detrital quartz grains display several very small knobs (100 to 300 nm in diameter) which correspond to quartz overgrowths developed in the porous space between the detrital quartz and the tangling up of the overlying fine-grained chlorite particles (Fig. 5d⇑). The size and the shape of these quartz overgrowths are delimited by the geometry of the interparticular porous space at the contact with the detrital quartz. The development of these very small quartz overgrowths at the base of the chlorite grain coating led to partial insertion in the quartz grains of the chlorite particles which are oriented oblique or perpendicular to the grain surface (Fig. 5d⇑).
All the data obtained on the chlorite grain coatings of sandstones from the four reservoirs described in this paper indicate that at the scale of the pore, they present similar petrographic relationships and textural characters despite their different geological settings (age of sedimentation, depositional environment, burial history, bulk mineralogy and nature of the resident pore fluid). This suggests the same process of formation, several aspects of which have already been discussed in the literature. As specified by many authors, the petrographic relationships established at the scale of the pore indicate an early occurrence of chlorite grain coatings during the diagenesis and the inhibition of the subsequent quartz cementation within the pore space (Hayes, 1970; Ehrenberg, 1993; Grigsby, 2001). The determination of the origin of the material which composes the grain coating chlorites is not the objective of the present paper, though this point has been discussed by several authors who concluded that Fe-chlorite in sandstones probably derives from the diagenetic transformation of syndepositional Fe-rich clays (Odin, 1988; Ehrenberg, 1993; Ryan & Reynolds, 1996).
The multiscale petrographic and textural study presented in the above sections provides original data which can be used to constrain the timing and mechanism of chlorite growth in the grain coatings and the mechanism by which the particles of chlorite inhibit the extensive quartz cementation of the pore space.
Timing and mechanism of chlorite growth
Several petrographic relationships are relevant to interpretion of the timing of the chlorite coating development. In each studied sandstone, chlorite is present mostly as grain coatings whose characteristics change according to the fact that they occur at the contact between adjacent grains or at the wall of the primary pore space. The nanopetrographic study of both types of crystallization sites indicates that they essentially differ in thickness, arrangement and textural properties of chlorite particles. The fact that a thin film of chlorite particles oriented parallel to the sand grain surface can be observed by TEM at the contact between the adjacent grains that had come in contact during the compaction, suggests that grain-coating chloritization began before the complete consolidation of the sand by mechanical compaction (i.e. very shallow burial conditions).
The SEM and TEM investigations of cross-sections through the thick chlorite grain coatings crystallized at the wall of the pore space reveal an asymmetric organization which consists of strong differences in arrangement, morphology and size of the chlorite particles from the detrital grain edge to the centre of the pore. When observed by SEM, the chlorite coatings crystallized at the walls of the primary pore space appear as a composite of two different types of chlorite coatings: (1) an inner coating with the same characteristics as the chlorite coating observed at the contact between adjacent grains (film of very small particles of chlorites, <1 μm thick, oriented oblique or parallel to the grain surface); (2) an outer rim, 4 to 10 μm thick, composed of radially oriented euhedral crystals of chlorite.
The average thickness of the inner chlorite coating remains constant in all the studied samples whereas the thickness of the outer layer which depends directly on the crystal size of the authigenic chlorites varies from one reservoir to another. This variation of crystal size is not correlated to the maximum temperature reached by the sandstones during their diagenesis (Table 1⇑).
The composite structure of the whole chlorite grain coatings crystallized at the walls of the primary pore space could be interpreted as the superimposition of two distinctive growth processes during the burial history of these sandstones: (1) an early stage during which chlorites crystallized in replacement of Fe-rich clay precursors as proposed by Ehrenberg (1993); and (2) a later stage during which radially oriented authigenic chlorites crystallized upon the previous chlorite coating. The first stage would occur at shallow burial depth in a rock that was still not totally consolidated (i.e. before the sand grains had come in contact and had acquired their present relative position by mechanical compaction). The second stage would occur at greater burial depth and results in the growth of authigenic coarse-grained euhedral chlorites by redistribution of the chloritic material of the first chlorite coating through Ostwald process only at the wall of the remaining pore space after the total consolidation of the sandstone (Jahren, 1991; Grigsby, 2001).
The nanopetrographic relationships established by TEM in both types of chlorite grain coatings call into question the hypothesis of two distinctive events of chlorite crystallization. The fact that the inner chlorite coating observed at the wall of the pore space is similar to that observed at the contact between the adjacent sand grains suggests that it did not suffer the dissolution expected in the case of recrystallization of euhedral chlorite by an Ostwald process, and consequently invalidates the growth of a second generation of chlorite by cation redistribution.
The TEM observations of cross-sections through the chlorite coatings formed at the wall of the pore space demonstrate that, in fact, the changes in arrangement and texture of the chlorite particles from the contact with the substratum to the centre of the pore are progressive. The very small chlorite particles which are parallel or slightly oblique to the substratum surface at the contact with the detrital grains tend to become progressively oriented perpendicularly to the surface of the detrital grains and coarsen with increasing distance from the detrital substratum. This type of spatial evolution of the textural properties of crystal aggregates (Fig. 5c,d⇑) agrees broadly with the principle of growth by geometric selection. Geometric selection is a growth process rather common in nature which results in the formation of a series of mineral aggregates on a planar surface, as druses for example. This growth process has been discussed extensively by Grigor’ev (1965). It is based on the fact that during common growth a characteristic struggle for space occurs between the mineral individuals which are randomly oriented on a planar surface. As the crystals come into contact during growth, a geometric selection occurs: the crystals oriented favourably gain growth space whereas for others less favourably oriented it is inhibited and ceases. In such a process, only crystals steeply inclined to the growth surface can continue to grow and finally only crystals whose direction of maximum growth rate is strictly perpendicular to the growth surface will be able to grow. Nothing can inhibit the growth of crystals which are oriented in this direction because there is no competition between these crystals. A consequence of this process is a rapid decrease in the number of crystals growing, coupled with an increase in crystal size, as the distance from the surface of initial growth increases and then the number of coarse-grained individual crystals oriented perpendicular to the surface of the substratum stabilizes. It should be noted that the differences in textural properties of the crystals thus formed represent various stages of a continuous crystallization process and not the superimposition of several specific processes.
All the above considerations lead us to interpret the formation of the chlorite grain coatings studied in the four sandstone reservoirs as a continuous growth process controlled by geometrical selection as illustrated in Fig. 8⇓. On the basis of petrographic considerations, it is proposed that this growth process operated at a range of shallow burial depth in which the transition from unconsolidated to consolidated sandy sediments occurred. It began before the compaction of the sand grains, probably by replacement of an Fe-rich clay precursor and stopped before the development of quartz cementation. This interpretation is compatible with data from the literature on the temperature and burial conditions for the first occurrence of chlorite and of quartz cement in sandstones. On the basis of isotope data, Grigsby (2001) proposed temperatures as low as 20–40°C for the beginning of chlorite growth in sandstones of the lower Vickburg formation in South Texas. Burial depths between 1800 and 2500 m and temperatures between 70 and 90°C have been estimated for the first occurrence of quartz cements in the Tertiary sandstone of the Gulf Coast (Land et al., 1987; Grigsby, 2001) and in the sandstones of the North Sea (Ehrenberg, 1990; Bjorlykke et al., 1992; Robinson & Gluyas, 1992; Walderhaug, 1994; Hogg et al., 1995; Girard et al., 2001).
In summary, we propose that the chlorite grain coatings of the sandstone reservoirs studied result in a single crystallization stage during which the spatial organization and the textural properties of the chlorite crystals were controlled by geometrical selection in any crystallization site. The changes of properties existing between the chlorite grain coatings in different sites are due to the fact that, because of the rearrangement of the clastic grains during the sediment compaction, chlorite growth has been terminated at different stages of the geometrical selection.
The textural properties of the euhedral crystals of chlorite constituting the outer rim of the coatings represent the latest stage of the chlorite growth process. These properties cannot be related to the maximum temperatures reached during the diagen-esis of the sandstones because euhedral chlorites ceased to grow at relatively low temperature and at shallow burial depth before quartz overgrowth began to crystallize.
Mechanism of quartz cement inhibition
In a detailed study of deeply buried chloritized sandstones from the Norwegian continental shelf (including the case of the Haltenbanken reservoir), Ehrenberg (1993) discussed the mechanism of quartz cement inhibition and proposed the following interpretation: Chlorite coatings inhibit quartz cement growth by isolating detrital quartz surfaces from the pore-water and thus preventing necleation of authigenic quartz. However, quartz nucleation is never blocked completely and a few quartz overgrowths having the same crystallographic orientation could result from the epitaxial growth of secondary quartz on the face of a monocrystalline sand grain. From the different types of chlorites observed in sandstones, authigenic grain coatings of radially oriented chlorite are expected to be responsible only for inhibiting, not preventing the growth of quartz cement.
The results of this study agree with the overall mechanism proposed above and provide new information which allow us to improve it significantly. Observations by TEM of the interface between the chlorite coating and the detrital quartz surface (Fig. 5c,d⇑) clearly indicate that nucleation sites were still abundant at the moment of the crystallization of secondary quartz and that growth has been effective in the very small interstitial porous space which existed between the quartz substratum and the tangling up of overlying fine-grained chlorite particles. This means that: (1) detrital quartz surfaces were not physically isolated from the pore fluids by crystals of the chlorite coating; and (2) inhibition of quartz cementation is due to the drastic limitation of the epitaxial growth of quartz into the interparticle space at the base of the chlorite coating.
The major consequence of this phenomenon is that almost all the quartz nuclei gave way to numerous but isolated quartz nanocrystals which cemented the base of the chlorite coating. In the absence of discontinuous chlorite coating, rare coarse quartz overgrowth can be developed if and only if a nucleation site is connected to the centre of the pore by a microporous space whose geometry (tortuosity) and orientation are compatible with the direction of maximum rate of growth of the quartz crystal. This means that the most effective chlorite for inhibition of quartz cementation was not the radially oriented euhedral crystals but the rather tangentially oriented very small anhedral crystallites which comprise the inner part of the chlorite coating.
In this study, it was demonstrated that texture and arrangement of chlorite particles within the chlorite grain coatings and nanopetrographic relationships at the scale of the pore space give new insights on the timing of chloritization in the sandstone reservoirs, on the growth process of the crystals forming coating and on the mechanism by which chlorite particles inhibited the subsequent quartz cementation of the pore space.
In the four reservoirs studied, the chlorite coatings result from a continuous growth process which followed the principle of geometric selection which is well known to be the growth process of drusy minerals. As a consequence, chlorite coatings result from a continuous growth the crystallites of which may contain information on the dynamics and timing of the crystallization process. Chlorite coatings began to grow before the consolidation of the sandy sediment and the differences in thickness and organization presently observed in many samples can be explained by the local change in pore geometry in response to the mechanical compaction and its potential influence on the growth rate of chlorites.
The relationships between chlorite and secondary quartz at the nanometric scale demonstrate that detrital quartz surfaces were not physically isolated from the pore-fluids by the crystals of the chlorite coating at the moment of quartz crystallization. The inhibition of quartz cementation is essentially due to the drastic limitation of the epitaxial growth of quartz to the interparticular space at the base of the chlorite coating and not to a lack of nucleation sites as previously proposed in the literature.
The fact that very similar results have been obtained in four sandstone reservoirs from very different geological settings (age of sedimentation, depositional environment, burial history, bulk mineralogy and nature of the resident pore fluid) suggests that our proposed mechanism for the formation of chlorite-grain coatings can be generalized to most of the sandstones which contain Fe-rich chlorite-grain coatings.
Grain coating composed of chlorite particles crystallized during an early stage of diagenesis during which the conditions of burial depth and temperature upgraded progressively from <1000 m to 2000–2500 m and from 20–40°C to 70–80°C respectively.
As the spatial distribution of the chlorite particles results in a geometric selection process, they can be discriminated on the basis of their relative age. This offers new perspectives on the validity of the different models, proposed in the literature, for the formation of diagenetic chlorites arising from selective HRTEM and AEM investigations on particles formed at the different stages of the growth process of the chlorite coatings.
The authors are grateful to D. Pacquet and E. De Laulanie for their help with preparation of samples for TEM analyses, to S. Nitsche for his early help with the TEM and ion-thinning work and to TotalFinaElf for the financial support. The previous version of this paper benefited from careful and constructive reviews by J.M. Adams, A.M. Karpoff, F. Nieto Garcia and an anonymous reviewer.
- Received September 2, 2002.
- Revision received February 10, 2003.