- Email: [email protected]
Interstratified illite-smectite phases: formation mechanisms and practical applications
Available online at www.sciencedirect.com
ScienceDirect Russian Geology and Geophysics 59 (2018) 1120–1128 www.elsevier.com/locate/rgg
Interstratified illite–smectite phases: formation mechanisms and practical applications G.A. Krinari †, M.G. Khramchenkov * Kazan Federal University, Institute of Geology and Petroleum Technologies, ul. Kremlevskaya 4/5, Kazan, 420008, Russia Received 23 June 2017; accepted 2 November 2017
Abstract Secondary micas after smectite, including mix illite–smectite phases, can form in sediments by three mechanisms, each being specific to particular environments. As the process develops, the newly formed phases undergo structure ordering. Two mechanisms involve transformation of 2:1 mixed-layer structures, and the third is the growth of screw dislocations, with formation of ordered mix phases having a Reichweite parameter of R = 1 or R = 2. We propose methods for identifying such phases when they are present in small amounts or when their XRD patterns lack well-pronounced superperiodic reflections, as well as mathematical formalism for illitization modeling. The theoretical issues are illustrated with field examples, and the illitization mechanisms are discussed in terms of their possible practical applications. © 2018, V.S. Sobolev IGM, Siberian Branch of the RAS. Published by Elsevier B.V. All rights reserved. Keywords: XRD analysis; structure; interstratified illite–smectite phases
Introduction Conversion of smectite to illite (illitization) is a global process common to all sedimentary strata. It is assumed to be basically a single event induced by pressure and temperature increase upon progressive burial of sediments. It occurs by the dissolution–precipitation mechanism (Drits and Sakharov, 1976; Srodon′ et al., 2000; Drits et al., 2007), whereas the 2:1 layer structure remains invariable in the case of hydrothermal processes (Frank-Kamenetsky et al., 1983). Illitization includes the formation of interstraified illite–smectite (I/S): first disordered mix phases, at R = 0, then ordered mix phases at R = 1 and then R = 2, where R is the Reichweite parameter (German for range or reach). According to transmission electron microscopy (TEM), however, particles of I/S phases with different R values may coexist in one sample (Dong et al., 1997). Later other illitization mechanisms were suggested, such as direct conversion of smectite into illite, skipping he intermediate I/S phases (Krinari and Khramchenkov, 2005; Solotchina, 2009), and the dislocation growth mechanism for illite crystals (Krinari and Khramchenkov, 2008). Furthermore, illite particles show both normal and lognormal distributions of v
†
Deceased. * Corresponding author. E-mail address: [email protected] (M.G. Khramchenkov)
measured thicknesses instead of normal distribution only, thus indicating more than one illitization mechanisms (Dudek et al., 2002). The inconsistency stems from the wrong assumption that the smectite–illite transition inferred for mature sediments with high P-T parameters would be universal and applicable to low-grade sedimentary rocks. The structure of fine illite–smectite phases should be analyzed with special X-ray diffraction methods.
Experimental methods for studying the structure of interstratified illite–smectite phases The experiments were applied to the ≤2.5 µm fraction extracted from dry or originally water-bearing samples of reservoirs D1 and D0 at the Romashkino oil field in West Siberia, as well as from Neogene and Upper Permian clay samples from Tatarstan. Both basal (00l) and other diffraction spectra were collected from oriented specimens. The (00l) reflections were recorded in air-dried material saturated with ethylene glycol, on the 1/Å linear scale, in the range from 0.002 to 0.405 1/Å (x axis), at a step of 0.0008 1/Å. Thus we culled out samples containing kaolinite together with mix phases with a high proportion of smectite, which produce separate reflections in glycolated clay.
1068-7971/$ - see front matter D 201 8, V.S. So bolev IGM, Siberian Branch of the RAS. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.rgg.201 + 8.08.006
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
Checking the assumption that secondary illite would retain the primary turbostratic smectite structure till its complete recrystallization to 1Md polytype requires collecting Bragg’s reflections (Krinari and Khramchenkov, 2008; Zviagina and Krinari, 1989). This approach provides separate recording of reflections from different layers of stacking arrangement which belong to certain crystallographic zones localized at some distance Bhk relative to the 1/Å axis (zones 11L–02L and 13L–20L are the most informative). The spectra were collected on a special automatic double-crystal diffractometer at the optimal values B11 = 0.2225 1/Å and B13 = 0.3720 1/Å, and at the spacing dhkl = 1/(X2 + Bhk2)1/2, where X is the current spectrum coordinate on the x axis. Investigation into the structure of the clay component in sediments may have various purposes. The probabilistic components of interstratification in each mix phase are determined by fitting based on the Markov chain theory (Sakharov et al., 1999; Solotchina, 2009). We focus on the very presence of mix phases which form in sediments at certain stages of their history, possibly as minor components of nonequilibrium systems. The problem is solved using difference between the spectra of air-dried and glycolated samples normalized to the reflection of a phase free from labile components (Krinari et al., 2014): the spectrum of the glycolated sample is subtracted from that of the air-dried sample. The procedure can be performed with the standard software for XRD analysis (e.g., the Topas software). The resulting difference spectra contain only the contributions from mix phases with labile (expandable) interlayers. They appear in the experimental curve as local maximums dmax and minimums dmin, and their position is almost independent of the structure of primary phases and the instrument function. The calculated models were used to plot the parameters of a structure with one (1H2O) or two (2H2O) water layers between 2:1 layers as a function of mica (pM) contents. The dmin and dmax values and the zero-line heights Lz can be used to choose theoretical models for the found mix phases with regard to the presence of 1H2O or 2H2O layers. The straight line falling to the x axis from the point dmin on the curve dmin = f(pM) should cross the curve Lz = f(pM) at the height close to Lz in the theoretical spectrum for structures with R ≥ 1. The point with the coordinates dmin and dmax for structures with R = 0 is located between the curves Lz = f(pM) for 1H2O and 2H2O (Krinari et al., 2014). Figure 1 presents modeling examples for samples with R = 1 and R = 2; the same curves for phases with R = 0 and R = 3 were reported earlier (Krinari et al., 2014).
Illitization mechanisms in different physicochemical conditions of sediment maturation A system of layers possesses excess surface energy in the cases of interlayer composition change and volume reduction (pressure increase), at any illitization mechanism, including the formation of mix phases. Therefore, the Gibbs–Duhem equation can be used (Deryagin et al., 1987):
1121
V dP = −A⋅h⋅d Π (h) + ∑ ni⋅dµi + A⋅dσ,
(1)
i
where µi is the chemical potential of the ith component in the system; ni is the number of moles of the ith component; V is the average volume of the particle subject to transformation; A is the surface of secondary mica particles; h is the thickness of a expandable interlayer in mix phases; σ is the surface energy density; P is the pressure, Π (h) is the wedging pressure in a system of 2:1 layers. After dividing the left- and right-hand sides of (1) by V (V/A = r, where r is the thickness of illite particles) and subsequent integration, (1) becomes P = −ξ⋅Π (h) + RT∑ (Ci − Ci(0)) + i
αGb 2 8π (1 − v)
2
r ln ρ ,
(2)
where Ci(0) and Ci are the initial and current contents of the ith component; G is the shear modulus; b is the Burgers vector; v is the Poisson ratio; α is the surface density of dislocations; ρ is reverse to α (Hirth and Lothe, 1982). The parameter ξ = (1 / r) ⋅ h refers to changes in the contribution of wedging pressure when smectite layers become partially replaced by illite ones in mix particles. Equation (2) can be used to estimate the energy released in each mechanism effectuated in nature. The maximum energy contribution of wedging pressure is ≈ 2–3 J/cm3 (Churaev, 1990). This is obviously too small for phase transformation but sufficient for providing structure-dependent rotation of 2:1 layers at the account of interlayer composition change, whereby the neighbor planes of basal oxygens will have trigonal antiprsimatic coordination. The energy contribution of this process estimated to be 8–15 J/cm3 (Khramchenkov, 2008) is close to the internal energy of illite (≈12–25 J/cm3) and ensures direct illitization without the formation of mix phases, while the total smectite content in the system is reducing (Krinari and Khramchenkov, 2005). This mechanism can work in the presence of K1+ and Al3+ ions which may form by biochemical decomposition of K-feldspar upon oxidation of hydrocarbons in oil reservoirs (Krinari et al., 2005). The reaction zone is located within the domain of Eh polarity change in unconsolidated sediments where dehydration of smectite layers is physically inevitable (Krinari and Khramchenkov, 2005). Ordering in 2:1 layers during direct illitization controls the rate of Eh polarity change but not the P-T parameters of the system. Therefore, the 11L–02L spectra may record reflections of any mica polytype, including 2M2 which is common to high-pressure settings, but the 11–02 2D diffraction band predominates (Fig. 2). As the burial depth of sediments increases, heating becomes a more important mechanism than biochemical ordering. The least stable polytype 2M2 disappears first, but the system corresponds to a mixture of 2M1 and 1M with their peak position and relative reflection intensities close to the computed theoretical values. Therefore, the 2:1 structure does not change much at this stage (Fig. 3a).
1122
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
Fig. 1. Variations of the computed parameters of mixed-layer structures with R = 1 and R = 2 at 1H2O (1W) and 2H2O (2W) as a function of mica (pM) percentage. See text for explanation. Samples mentioned in text are abbreviated as OL 01, oil-bearing siltstone from Urengoi area, well 6263, core depth 2776 m, bed BU9; OL 03, water-bearing siltstone, Urengoi area, well 6279, depth 2821 m, bed BU9; Mi 46, oil-bearing siltstone, Kogalym area, bed BC111, well 1128, depth 2862 m; N 915, oil-bearing siltstone, Nivgal area, well 2061, bed BC8, depth 2974 m.
As the P-T parameters increase smoothly and pore fluid becomes progressively more saline, all 2:1 structures, including 2M1 and 1M mica polytypes, tend to lower entropy of the solids with increasing 3D ordering, which is not directly related to the total system energy as its volume may remain invariable. According to Belov’s theorem for lattice sums, the originally different structures are compatible if they share reciprocal lattice nodes (Drits and Sakharov, 1976). Some planes in illite polytypes 1M and 2M1 may coincide (Velde and Meunier, 2008) due to redistribution of cis- and trans-vacancies that change the interlayer shift by changing the angle β (Zviagina et al., 2007). Smectite-to-illite conversion with partial layer replacement is attendant with reduction of the unit cell size, whereby the trans-vacant __ layers 13L of 2M1 may coincide with the cis-vacant 13(L-2) layers of 1Mcv, which corresponds to the rotation of the 2M1 cell through 60° to 1M. Then some trans- and cis-vacant layers of 13L in 2M1 __ and 13(L-2) in 1Mtv, respectively, may produce 3D interstratification. This increases the intensity of the 13L layer reflections of 2M1 in the 13L–20L zone and __ causes their shift (Fig. 3b), but does not affect much 13L maximums (Krinari and Khamchenkov, 2008). The neighbor peaks of 1M and 2M1 merge in the patterns of deeper or less permeable samples (Fig. 3c).
Dislocation growth becomes the principal illitization mechanism when pressure increases and the rotation of layers exhausts the energy source. In this mechanism, polytypes 1M and 2M1 cannot make a common structure and produce separate reflections, while remnant mix 3D phases give weak reflections (Fig. 3d). The 2:1 mica layers can twist if enough K1+ and Al3+ cations are available, besides the energy potential, but twisting may continue even if one of them lacks, due to a single expandable interlayer. This produces homogeneous mix structures with R = 1 or R = 2 in the a0b0 plane, with persistent 3D ordering of the dislocation twists. This ordering holds in mix phases with R = 1 or R = 2, if the expandable interlayers are homogenized by the presence of one or two H2O layers (12.6 Å or 14.6 Å thick stacks, respectively). They are expected to appear as 3D diffraction reflections when the relative percentages of components approach a high-order structure. The 13L–20L spectrum of sample Mi 46 (Fig. 3d) contains reflections from 1M and 2M1, i.e., dislocations reduce the effect of 3D interstratification. There are three more peaks having no relation to any reflection from the considered mica polytypes. Structures corresponding to one or another mica polytype arise if screw dislocations are extended along the c* axis in the presence of 1H2O or 2H2O layers between the 2:1
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
1123
Fig. 2. XRD patterns of zone 11L–02L, five-fold spectra collection, sample O 80, Vetluga River, T1 to J3 transition. Lines show measured reflections (Å) and peak positions for models of mica polytypes 1M, 2M1 and 2M2. Abbreviations stand for mineral names: M, microcline; Q, quartz; D, dolomite; P, apatite; G, goethite; Py, pyrite; A, albite; O, orthoclase; Ca, calcite; Am, amphibole; Cr, cristobalite. The two curves are a superposition of two XRD patterns with different identified polytypes 2M1di, 2M2di, 2M1tri, 1Mtv, 1Mcv, which correspond to 1/Å from 0.05 to 0.20 (upper curve) and from 0.20 to 0.40 (lower curve). Inset shows the pattern below 1/Å = 0.05. The subscripts refer to: di, dioctahedral; tri, trioctahedral; tv, trans-vacancy; cv, cis-vacancy. Bar above a numeral denotes an oppositely oriented plane, according to a crystallographic rule.
layers oriented invariably to each other. A similar effect was observed in high-order kaolinite saturated with dimethylsulfoxide, in which the 3D structure was controlled more by twisting continuity than by interaction of layers with organic molecules (Galimova et al., 1990). Peak 1 with d = 2.252 Å (Fig. 3) is a diffraction from planes 135 of the polytype 2M1, where a 2H2O layer is located behind every second layer (R = 1), as one may predict using the previously applied formalism (Krinari and Khramchenkov, 2008). Node 135 of the reciprocal lattice is shifted from 0.05006 to 0.04065 1/Å along the c* axis, which corresponds
to the d001 spacing increase from 19.98 to 24.60 Å. The same calculations for peak 2 at d = 2.497 Å lead to reflection 133 of 2M1, with a 1H2O layer between R = 2 mica layers, while d001 increases to 32.72 Å. Peak 2 may also contain a contribution from a structure with 2H2O at R = 2. Peak 3 most likely represents the 3D structure of dioctahedral vermiculite with d001 = 12.6 Å, which has its 2:1 layers twisted in the same_ way as in 1M and produces the strongest reflection 133 at these wavelengths. These phases, possibly also with 3D ordering, were described earlier by Sakharov et al. (1999).
1124
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
Fig. 3. XRD patterns of zone 13L–20L in different samples: a, Lb 10, Kama River, P2kz3; b, Lm 23, Kama mouth, P2tat1; c, Mu 15, Kholmy area, well 31015, depth 1806 m, reservoir D0; d, Mi 46 (Fig. 1). See text for explanation.
The volume of particles that contribute to 3D diffraction in TEM images is much greater than in the case of XRD basal reflections. For this reason, the 1/Å scale patterns better resolve ordered phases with R = 1 and R = 2 than those of 00L reflections which miss superperiod reflections for sample Mi 46. This is because the basal reflections from mica and chlorite, zeroed down by subtraction of spectra, are orders of magnitude stronger than the superperiod reflections, which are our main interest in this study. If the difference spectrum from ordered structures with R = 1 or R = 2 lacks reflections of other phases, it should bear local peaks from air-dried samples and local troughs of glycolated samples, corresponding to superperiod reflections. Rakhmatullina and Krinari (2013) discussed an example of a 00L pattern of Na-rectorite that contained a series of superperiod reflections when the illite component reached a concentration of pM = 0.56 (sample OL 01). Figure 4 shows local minimums from a glycolated rectorite (002, 003) and a local maximum 002 from an air-dried rectorite sample. The spectrum also contains local peaks and troughs from phases with R = 1 and pM much above 50% (pM = 0.78; 0.90), as well as phases with R = 0 and pM = 0.83. A compositionally similar sample of water-bearing siltstone OL 03 lacks rectorite reflections, while the difference spectrum reveals phases with R = 1 at pM = 0.86 and pM = 0.95, which confirms the possibility for coexistence of structurally similar phases with different pM values. In the same way, the difference spectrum of sample Mi 46 includes almost fully ordered mix structures,
including those revealed by scanning of the 13L–20L zone (Fig. 3d). There are local peaks and troughs of superperiod reflections of different phases with R = 1 and R = 2, which contain either 1H2O or 2H2O layers in interlayers. Furthermore, there are phases with R = 1 and R = 2 and pM much greater than for ordered structures (Fig. 5). The suggested mechanism imposes no limitations on relative percentages of components for phases with R = 1 and R = 2. K1+ ions can be captured in expandable interlayers also after twisting, which increases pM and the total number of phases with different pM values. Indeed, the analyzed samples showed no phases with R = 1 and pM < 0.5 or R = 2 and pM < 0.66. This fact, along with low concentration, most likely impedes the identification of phases with R = 1 and R = 2 in basal reflection images (Srodon′ , 1980), while the superperiod reflections attenuate rapidly as pM increases and their position changes. Dislocation growth (Hirth and Lothe, 1982), which may release mechanic stress, occurs by partial dissolution and concurrent precipitation, and the process may involve any layer and layer structure. Let the particle size dependence of surface energy density in (1) be v
dσ = (dσ / dr) dr.
(3)
The term A (dσ / dr) expressed via the dislocation energy density is
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
1125
Fig. 4. Difference spectrum: sample OL 01 with rectorite (Fig. 1). Abbreviations: e.g., ehtylene glycol; exp, experimental spectrum.
A
r dσ AαGb 2 = ln . dr 4π (1 − v) ρ
Discussion, main conclusions, and practical applications (4)
These equations can be used to find the energy contribution of dislocation growth to illitization, with G = 1.5 × 109 Pa and v = 0.2, which is from 10 to 16 J/cm3, depending on the Burgers vector. Note that the energy contributions of the second and third mechanisms are very similar. Therefore, structures produced by both mechanisms consuming the excess energy may coexist in the system subject to transformations. The energy found by (2) is 10 J/cm3 for the mix stack I : S : I : S, with a characteristic total thickness of 2.5 nm, and 16 J/cm3 for the stack I : I : S : I : I : S with a total thickness of 3.5 nm. With reference to equation (2), one can estimate the thickness rc of an ordered mix phase with R = 1 or R = 2, which corresponds to the minimum energy of the whole system in a unit volume: ), β = A / (12πh 2) + α, 7A / (4√ 3πβ rc = √
(5)
where α is the specific system energy acquired by forming mica particles at the account of K substitution for interlayer cations; A is the Hamaker constant; h is the thickness of the water interlayer. Calculations following Deryagin et al. (1987) lead to rs = 2.2 nm if the presence of dislocations is neglected and to two values of rs = 2.2 nm and 3.2 nm for R = 1 and R = 2, respectively, if the energy contribution of dislocations is taken into account. The difference is significant but not crucial for coexistence of both structures in a single sample (Fig. 5). The difference spectra often include diffractions from phases with R = 0 which makes basis for the curve, as well as superposed local maximums and minimums of phases with R ≥ 1.
Earlier Huang et al., (1993) suggested using the model of illite-to-smectite conversion as a geothermometer. However, illitization may occur by several mechanisms and the behavior of smectite contents (pS) as a function of temperature or depth may differ in specific sections and change successively with increasing burial depth. See, for instance, pS = f (T) curves for two areas in the North Sea (Fig. 6) borrowed from (Srodon′ and Eberl, 1984), with three segments. If illitization is a direct low-temperature process (mechanism 1), the curves pS = f (T) only shift along the x axis, the amount of shift being controlled by biological processes rather than by temperature. The next segment is relatively straight and corresponds to smectite-toillite conversion by increasing illite component (mechanism 2). Another smooth bending and slope change of the pS = f(T) curve records the onset of dislocation growth (mechanism 3) as the sediments reach greater depths. The interstratified smectite–illite structure can be used as a geothermometer provided that the true temperature of some section part is known somehow a priori. Detecting zones of higher temperatures due to biochemical oil oxidation or local water inputs from deeper strata into reservoir rocks appears to be of greater practical value. Variance in local depth-dependent pS values observed in explicitly documented sections (Lanson et al., 2009) may be due to variations of wedging pressure caused by pore pressure changes which affect the rotation ability of clay layers, i.e., by porosity changes. The structure of ordered mixed-layer illite–smectite should depend on temperature. As a general tendency, R = 2 structures are found at greater core depths, while the total number of R = 1 phases increases. No more than two R = 1 phases were commonly found in test wells or in originally water-bearv
1126
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
Fig. 5. Difference spectrum: sample Mi 46 (Figs. 1, 3). a.d., Air-dried.
v
Fig. 6. Smectite-to-illite conversion in shales from different sedimentary basins, plotted as a function of temperature, after (Srodon′ and Eberl, 1984). Dashed lines delineate zones with one predominant illitization mechanism (1, 2, or 3).
ing sediments where two-phase flow was impossible. On the contrary, R = 1 and R = 2 structures coexisted in reservoir or producing rocks, which suggests the effect of percolation (Hunt and Ewing, 2009). In the case of two-phase flow, the medium parameters may be different for water and oil migrating along disconnected local paths. This tendency, if confirmed for a greater number of sections, may be used for reference in concurrent development of several simultaneously drilled reservoirs. For example, the difference spectra of sample N 915 (Fig. 7) show two R = 2 and one R = 1 phases at different numbers of H2O layers and pM values. Additionally, there are broader peaks of R = 0 phases at pM = 74 and pM = 95%. The latter may correspond to an almost monomineralic mica (but not illite), with scarce expandable interlayers, which formed late during the smectite-to-illite conversion.
Iillitization mechanisms 2 and 3 can be easily effectuated and applied in industrial technologies. There are methods for obtaining composite materials based on mechanically dispersed micas with a specific surface area ten times smaller than in the product composed of mix nanoparticles with R = 1 or R = 2. One can select a monomer which would interact with the surface of nanoparticles and penetrate into expandable interlayers between 2:1 layers. The mechanic properties of this composite material, after polymerization, should provide tight bonding of twisted silicate layers which would require much greater force to break down than the separation of organic molecular chains from the mineral surface. In addition to natural or manmade direct smectite-to-illite conversion leading to the formation of ordered secondary mica with a mosaic structure, there should be a reverse process
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
1127
Fig. 7. Difference spectrum: sample N 915 (Fig. 1).
driven by chemical or biochemical mechanisms producing high mechanical stress. This process in sediments with low open porosity may cause deformation of fittings and secondary cracking (Krinari et al., 2014). In clastic reservoirs, it leads to breakdown of secondary mica particles into isolated nanoparticles with a high surface charge. This may happen during injection of fresh water into oil reservoirs and create a return electroosmotic flow which reduces the formation permeability (Krinari and Khramchenkov, 2011). Reservoirs may become impermeable as the pore water salinity reduces to 90 g/L (a density of 1.09 g/L) while the nanoparticles reach a concentration of 0.1 vol.% of the rock skeleton (which causes almost no porosity decrease) and a number of N ≈ 7 × 1018/m3, with a basal surface area of S = 44 × 1022 nm2 and a noncompensated charge of Q = 2.8 × 105 C/m3. Such obstruction of permeability was observed in the Romashkino field (Krinari and Khramchenkov, 2011), with the reaction zone always about 10 m ahead of the flood front (Krinari et al., 2013). Water breakthrough at local reservoir sites can be mitigated using various chemical agents (Muslimov, 2003), but all of them change irreversibly the porosity of the reservoir which fails to regain its initial state. In the case of mica nanoparticles and related inhomogeneity of pores, the permeability depends more on pore fluid salinity, which is manageable. The electroosmotic screens produced by a freshwater flow, compress it vertically and then split into several arms while bypassed oil may preserve in the reservoir. These screens can be moved using flow management technologies (Muslimov, 2003), and their performance can be improved notably if the reservoir can return to its original permeability. The electroosmotic screens can block water breakthrough in any clastic reservoir, whether they contain secondary micas or not.
The study was supported by grant 15-11-10015 from the Russian Science Foundation.
References Churaev, N.V., 1990. Physicochemical Background of Mass and Heat Transfer in Porous Materials [in Russian]. Khimiya, Moscow. Deryagin, B.V., Churaev, N.V., Muller, V.M., 1987. Surface Forces [in Russian]. Nauka, Moscow. Dong, H., Peacor, D., Freed, R., 1997. Phase relations among smectite, R1 illite–smectite, and illite. Am. Mineral. 82, 379–391. Drits, V.A., Sakharov, B.A., 1976. XRD Analysis of Mixed-Layer Minerals (GIN Transactions, Issue 295) [in Russian]. Nauka, Moscow. Drits, V.A., Lindgreen, H., Sakharov, B.A., Jakobsen, H.J., Fallick, A.E., Salyn, A.L., Dainyak, L.G., Zviagina, B.B., Barfod, D.N., 2007. Formation and transformation of mixed-layer minerals by Tertiary intrusives in Cretaceous mudstones, West Greenland. Clays Clay Miner. 55, 260–283. Dudek, T., Srodon′ , J., Eberl, D.D., Elsass, F., Uhlik, P., 2002. Thickness distribution of illite crystals in shales. XRD, HRTEM measurements. Clays Clay Miner. 50, 85–91. Frank-Kamenetsky, V.A., Kotov, N.V., Goilo, E.A., 1983. Transformation of layered silicates at high P-T parameters [in Russian]. Nauka, Leningrad. Galimova, S.E., Krinari, G.A., Morozov, V.P., 1990. Intercalation and real interlayer structure of kaolinite crystals: A kinetic model, in: Spectroscopy, Crystal Chemistry, and Real Structure of Minerals and Their Analogs [in Russian]. Kazan University, Kazan, pp. 30–51. Hirth, J.P., Lothe, J., 1982. Theory of Dislocations. Krieger Publishing Company, Malabar. Huang, W., Longo, J., Peyear, D., 1993. An experimentally derived kinetic model for smectite-to-illite conversion and its use as a geothermometer. Clays Clay Miner. 41, 162–177. Hunt, A., Ewing, R., 2009. Percolation Theory for Flow in Porous Media, Lection Notes Physics 771. Springer, Berlin–Heidelberg. Khramchenkov, M.G., 2008. Fan-shaped model of clay swelling process, in: Thermo-Hydromechanical and Chemical Coupling in Geomaterials and Applications. Polytech-Lille, France, pp. 297–304. v
1128
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
Krinari, G.A., Khramchenkov, M.G., 2005. Low-temperature illitization of smectite as a biomineral process. Dokl. Earth Sci. 403A (6), 926–930. Krinari, G.A., Khramchenkov, M.G., 2008. Three-dimensional structure of secondary micas in sedimentary rocks: features and mechanisms of formation. Dokl. Earth Sci. 423A (9), 1435–1439. Krinari, G.A., Khramchenkov, M.G., 2011. Inverse transformation of secondary micas in sedimentary rocks: mechanisms and applications. Dokl. Earth Sci. 436 (5), 674–679. Krinari, G.A., Khramchenkov, M.G., Rakhmatulina, Yu.Sh., 2013. Mechanisms of inverse transformations of secondary micas: evidence from changes in the illite–smectite structure. Dokl. Earth Sci. 452 (2), 1014–1019. Krinari, G.A., Khramchenkov, M.G., Rakhmatulina, Yu.Sh., 2014. Changes in the structures of mixed-layer illite–smectite during flooding of terrigenous oil reservoirs. Russian Geology and Geophysics (Geologiya i Geofizika) 55 (7), 915–928 (1153–1167). Krinari, G.A., Shinkarev, A.A., Giniyatullin, K.G., 2005. Abiogenic and biogenic degradation of minerals: difference, mechanisms, and practical applications. Zapiski RMO, No. 1, 18–32. Lanson, B., Sakharov, B.A., Claret, F., Drits, V.A., 2009. Diagenetic smectite-to-illite transition in clay-rich sediments: a reappraisal of X-ray diffraction results using the multi-specimen method. Am. J. Sci. 309, 476–516. Muslimov, R.Kh., 2003. Modern Methods for Management of Oil Reservoirs with the Use of Flooding [in Russian]. Kazan University, Kazan.
Rakhmatullina, Yu.Sh., Krinari, G.A., 2013. Formation mechanisms of secondary micas: Effect on the character of illite–smectite phases in flooding of oil reservoirs. Georesursy 53 (3), 27–30. Sakharov, B.A., Lindgreen, H., Salyn, A., Drits, V.A., 1999. Determination of illite–smectite structures using multispecimen X-ray profile fitting. Clays Clay Miner. 47 (5), 555–566. Solotchina, E.P., 2009. Structure Typomorphism of Clay Minerals in Sedimentary Sections and Weathering Profiles [in Russian]. Geo Publishers, Novosibirsk. Srodon′ , J., 1980. Precise identification of illite/smectite interstratifications by X-Ray powder diffraction. Clays Clay Miner. 28, 401–411. Srodon′ , J., Eberl, D., 1984. Illite, in: Bailey, S.W. (Ed.), Micas. Rev. Mineral., Mineral. Soc. Am. 13, 495–544. Srodon′ , J., Eberl, D.D., Drits, V.A., 2000. Evolution of fundamental particle-size during illitization of smectite and implications for reaction mechanism. Clays Clay Miner. 48 (4), 446–458. Velde, B., Meunier, A., 2008. The Origin of Clay Minerals in Soils and Weathered Rocks. Springer-Verlag, Berlin–Heidelberg. Zviagina, B.B., Krinari, G.A., 1989. Geometrical features and principles of diffraction imaging of ordered-structure samples. Kristallografiya 34 (2), 288–291. Zviagina, B.B., Sakharov, B.A., Drits, V.A., 2007. X-ray diffraction criteria for the identification of trans- and cis-vacant varieties of dioctahedral micas. Clays Clay Miner. 55 (5), 467–480. v
v
v
Editorial responsibility: N.V. Sennikov
ScienceDirect Russian Geology and Geophysics 59 (2018) 1120–1128 www.elsevier.com/locate/rgg
Interstratified illite–smectite phases: formation mechanisms and practical applications G.A. Krinari †, M.G. Khramchenkov * Kazan Federal University, Institute of Geology and Petroleum Technologies, ul. Kremlevskaya 4/5, Kazan, 420008, Russia Received 23 June 2017; accepted 2 November 2017
Abstract Secondary micas after smectite, including mix illite–smectite phases, can form in sediments by three mechanisms, each being specific to particular environments. As the process develops, the newly formed phases undergo structure ordering. Two mechanisms involve transformation of 2:1 mixed-layer structures, and the third is the growth of screw dislocations, with formation of ordered mix phases having a Reichweite parameter of R = 1 or R = 2. We propose methods for identifying such phases when they are present in small amounts or when their XRD patterns lack well-pronounced superperiodic reflections, as well as mathematical formalism for illitization modeling. The theoretical issues are illustrated with field examples, and the illitization mechanisms are discussed in terms of their possible practical applications. © 2018, V.S. Sobolev IGM, Siberian Branch of the RAS. Published by Elsevier B.V. All rights reserved. Keywords: XRD analysis; structure; interstratified illite–smectite phases
Introduction Conversion of smectite to illite (illitization) is a global process common to all sedimentary strata. It is assumed to be basically a single event induced by pressure and temperature increase upon progressive burial of sediments. It occurs by the dissolution–precipitation mechanism (Drits and Sakharov, 1976; Srodon′ et al., 2000; Drits et al., 2007), whereas the 2:1 layer structure remains invariable in the case of hydrothermal processes (Frank-Kamenetsky et al., 1983). Illitization includes the formation of interstraified illite–smectite (I/S): first disordered mix phases, at R = 0, then ordered mix phases at R = 1 and then R = 2, where R is the Reichweite parameter (German for range or reach). According to transmission electron microscopy (TEM), however, particles of I/S phases with different R values may coexist in one sample (Dong et al., 1997). Later other illitization mechanisms were suggested, such as direct conversion of smectite into illite, skipping he intermediate I/S phases (Krinari and Khramchenkov, 2005; Solotchina, 2009), and the dislocation growth mechanism for illite crystals (Krinari and Khramchenkov, 2008). Furthermore, illite particles show both normal and lognormal distributions of v
†
Deceased. * Corresponding author. E-mail address: [email protected] (M.G. Khramchenkov)
measured thicknesses instead of normal distribution only, thus indicating more than one illitization mechanisms (Dudek et al., 2002). The inconsistency stems from the wrong assumption that the smectite–illite transition inferred for mature sediments with high P-T parameters would be universal and applicable to low-grade sedimentary rocks. The structure of fine illite–smectite phases should be analyzed with special X-ray diffraction methods.
Experimental methods for studying the structure of interstratified illite–smectite phases The experiments were applied to the ≤2.5 µm fraction extracted from dry or originally water-bearing samples of reservoirs D1 and D0 at the Romashkino oil field in West Siberia, as well as from Neogene and Upper Permian clay samples from Tatarstan. Both basal (00l) and other diffraction spectra were collected from oriented specimens. The (00l) reflections were recorded in air-dried material saturated with ethylene glycol, on the 1/Å linear scale, in the range from 0.002 to 0.405 1/Å (x axis), at a step of 0.0008 1/Å. Thus we culled out samples containing kaolinite together with mix phases with a high proportion of smectite, which produce separate reflections in glycolated clay.
1068-7971/$ - see front matter D 201 8, V.S. So bolev IGM, Siberian Branch of the RAS. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.rgg.201 + 8.08.006
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
Checking the assumption that secondary illite would retain the primary turbostratic smectite structure till its complete recrystallization to 1Md polytype requires collecting Bragg’s reflections (Krinari and Khramchenkov, 2008; Zviagina and Krinari, 1989). This approach provides separate recording of reflections from different layers of stacking arrangement which belong to certain crystallographic zones localized at some distance Bhk relative to the 1/Å axis (zones 11L–02L and 13L–20L are the most informative). The spectra were collected on a special automatic double-crystal diffractometer at the optimal values B11 = 0.2225 1/Å and B13 = 0.3720 1/Å, and at the spacing dhkl = 1/(X2 + Bhk2)1/2, where X is the current spectrum coordinate on the x axis. Investigation into the structure of the clay component in sediments may have various purposes. The probabilistic components of interstratification in each mix phase are determined by fitting based on the Markov chain theory (Sakharov et al., 1999; Solotchina, 2009). We focus on the very presence of mix phases which form in sediments at certain stages of their history, possibly as minor components of nonequilibrium systems. The problem is solved using difference between the spectra of air-dried and glycolated samples normalized to the reflection of a phase free from labile components (Krinari et al., 2014): the spectrum of the glycolated sample is subtracted from that of the air-dried sample. The procedure can be performed with the standard software for XRD analysis (e.g., the Topas software). The resulting difference spectra contain only the contributions from mix phases with labile (expandable) interlayers. They appear in the experimental curve as local maximums dmax and minimums dmin, and their position is almost independent of the structure of primary phases and the instrument function. The calculated models were used to plot the parameters of a structure with one (1H2O) or two (2H2O) water layers between 2:1 layers as a function of mica (pM) contents. The dmin and dmax values and the zero-line heights Lz can be used to choose theoretical models for the found mix phases with regard to the presence of 1H2O or 2H2O layers. The straight line falling to the x axis from the point dmin on the curve dmin = f(pM) should cross the curve Lz = f(pM) at the height close to Lz in the theoretical spectrum for structures with R ≥ 1. The point with the coordinates dmin and dmax for structures with R = 0 is located between the curves Lz = f(pM) for 1H2O and 2H2O (Krinari et al., 2014). Figure 1 presents modeling examples for samples with R = 1 and R = 2; the same curves for phases with R = 0 and R = 3 were reported earlier (Krinari et al., 2014).
Illitization mechanisms in different physicochemical conditions of sediment maturation A system of layers possesses excess surface energy in the cases of interlayer composition change and volume reduction (pressure increase), at any illitization mechanism, including the formation of mix phases. Therefore, the Gibbs–Duhem equation can be used (Deryagin et al., 1987):
1121
V dP = −A⋅h⋅d Π (h) + ∑ ni⋅dµi + A⋅dσ,
(1)
i
where µi is the chemical potential of the ith component in the system; ni is the number of moles of the ith component; V is the average volume of the particle subject to transformation; A is the surface of secondary mica particles; h is the thickness of a expandable interlayer in mix phases; σ is the surface energy density; P is the pressure, Π (h) is the wedging pressure in a system of 2:1 layers. After dividing the left- and right-hand sides of (1) by V (V/A = r, where r is the thickness of illite particles) and subsequent integration, (1) becomes P = −ξ⋅Π (h) + RT∑ (Ci − Ci(0)) + i
αGb 2 8π (1 − v)
2
r ln ρ ,
(2)
where Ci(0) and Ci are the initial and current contents of the ith component; G is the shear modulus; b is the Burgers vector; v is the Poisson ratio; α is the surface density of dislocations; ρ is reverse to α (Hirth and Lothe, 1982). The parameter ξ = (1 / r) ⋅ h refers to changes in the contribution of wedging pressure when smectite layers become partially replaced by illite ones in mix particles. Equation (2) can be used to estimate the energy released in each mechanism effectuated in nature. The maximum energy contribution of wedging pressure is ≈ 2–3 J/cm3 (Churaev, 1990). This is obviously too small for phase transformation but sufficient for providing structure-dependent rotation of 2:1 layers at the account of interlayer composition change, whereby the neighbor planes of basal oxygens will have trigonal antiprsimatic coordination. The energy contribution of this process estimated to be 8–15 J/cm3 (Khramchenkov, 2008) is close to the internal energy of illite (≈12–25 J/cm3) and ensures direct illitization without the formation of mix phases, while the total smectite content in the system is reducing (Krinari and Khramchenkov, 2005). This mechanism can work in the presence of K1+ and Al3+ ions which may form by biochemical decomposition of K-feldspar upon oxidation of hydrocarbons in oil reservoirs (Krinari et al., 2005). The reaction zone is located within the domain of Eh polarity change in unconsolidated sediments where dehydration of smectite layers is physically inevitable (Krinari and Khramchenkov, 2005). Ordering in 2:1 layers during direct illitization controls the rate of Eh polarity change but not the P-T parameters of the system. Therefore, the 11L–02L spectra may record reflections of any mica polytype, including 2M2 which is common to high-pressure settings, but the 11–02 2D diffraction band predominates (Fig. 2). As the burial depth of sediments increases, heating becomes a more important mechanism than biochemical ordering. The least stable polytype 2M2 disappears first, but the system corresponds to a mixture of 2M1 and 1M with their peak position and relative reflection intensities close to the computed theoretical values. Therefore, the 2:1 structure does not change much at this stage (Fig. 3a).
1122
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
Fig. 1. Variations of the computed parameters of mixed-layer structures with R = 1 and R = 2 at 1H2O (1W) and 2H2O (2W) as a function of mica (pM) percentage. See text for explanation. Samples mentioned in text are abbreviated as OL 01, oil-bearing siltstone from Urengoi area, well 6263, core depth 2776 m, bed BU9; OL 03, water-bearing siltstone, Urengoi area, well 6279, depth 2821 m, bed BU9; Mi 46, oil-bearing siltstone, Kogalym area, bed BC111, well 1128, depth 2862 m; N 915, oil-bearing siltstone, Nivgal area, well 2061, bed BC8, depth 2974 m.
As the P-T parameters increase smoothly and pore fluid becomes progressively more saline, all 2:1 structures, including 2M1 and 1M mica polytypes, tend to lower entropy of the solids with increasing 3D ordering, which is not directly related to the total system energy as its volume may remain invariable. According to Belov’s theorem for lattice sums, the originally different structures are compatible if they share reciprocal lattice nodes (Drits and Sakharov, 1976). Some planes in illite polytypes 1M and 2M1 may coincide (Velde and Meunier, 2008) due to redistribution of cis- and trans-vacancies that change the interlayer shift by changing the angle β (Zviagina et al., 2007). Smectite-to-illite conversion with partial layer replacement is attendant with reduction of the unit cell size, whereby the trans-vacant __ layers 13L of 2M1 may coincide with the cis-vacant 13(L-2) layers of 1Mcv, which corresponds to the rotation of the 2M1 cell through 60° to 1M. Then some trans- and cis-vacant layers of 13L in 2M1 __ and 13(L-2) in 1Mtv, respectively, may produce 3D interstratification. This increases the intensity of the 13L layer reflections of 2M1 in the 13L–20L zone and __ causes their shift (Fig. 3b), but does not affect much 13L maximums (Krinari and Khamchenkov, 2008). The neighbor peaks of 1M and 2M1 merge in the patterns of deeper or less permeable samples (Fig. 3c).
Dislocation growth becomes the principal illitization mechanism when pressure increases and the rotation of layers exhausts the energy source. In this mechanism, polytypes 1M and 2M1 cannot make a common structure and produce separate reflections, while remnant mix 3D phases give weak reflections (Fig. 3d). The 2:1 mica layers can twist if enough K1+ and Al3+ cations are available, besides the energy potential, but twisting may continue even if one of them lacks, due to a single expandable interlayer. This produces homogeneous mix structures with R = 1 or R = 2 in the a0b0 plane, with persistent 3D ordering of the dislocation twists. This ordering holds in mix phases with R = 1 or R = 2, if the expandable interlayers are homogenized by the presence of one or two H2O layers (12.6 Å or 14.6 Å thick stacks, respectively). They are expected to appear as 3D diffraction reflections when the relative percentages of components approach a high-order structure. The 13L–20L spectrum of sample Mi 46 (Fig. 3d) contains reflections from 1M and 2M1, i.e., dislocations reduce the effect of 3D interstratification. There are three more peaks having no relation to any reflection from the considered mica polytypes. Structures corresponding to one or another mica polytype arise if screw dislocations are extended along the c* axis in the presence of 1H2O or 2H2O layers between the 2:1
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
1123
Fig. 2. XRD patterns of zone 11L–02L, five-fold spectra collection, sample O 80, Vetluga River, T1 to J3 transition. Lines show measured reflections (Å) and peak positions for models of mica polytypes 1M, 2M1 and 2M2. Abbreviations stand for mineral names: M, microcline; Q, quartz; D, dolomite; P, apatite; G, goethite; Py, pyrite; A, albite; O, orthoclase; Ca, calcite; Am, amphibole; Cr, cristobalite. The two curves are a superposition of two XRD patterns with different identified polytypes 2M1di, 2M2di, 2M1tri, 1Mtv, 1Mcv, which correspond to 1/Å from 0.05 to 0.20 (upper curve) and from 0.20 to 0.40 (lower curve). Inset shows the pattern below 1/Å = 0.05. The subscripts refer to: di, dioctahedral; tri, trioctahedral; tv, trans-vacancy; cv, cis-vacancy. Bar above a numeral denotes an oppositely oriented plane, according to a crystallographic rule.
layers oriented invariably to each other. A similar effect was observed in high-order kaolinite saturated with dimethylsulfoxide, in which the 3D structure was controlled more by twisting continuity than by interaction of layers with organic molecules (Galimova et al., 1990). Peak 1 with d = 2.252 Å (Fig. 3) is a diffraction from planes 135 of the polytype 2M1, where a 2H2O layer is located behind every second layer (R = 1), as one may predict using the previously applied formalism (Krinari and Khramchenkov, 2008). Node 135 of the reciprocal lattice is shifted from 0.05006 to 0.04065 1/Å along the c* axis, which corresponds
to the d001 spacing increase from 19.98 to 24.60 Å. The same calculations for peak 2 at d = 2.497 Å lead to reflection 133 of 2M1, with a 1H2O layer between R = 2 mica layers, while d001 increases to 32.72 Å. Peak 2 may also contain a contribution from a structure with 2H2O at R = 2. Peak 3 most likely represents the 3D structure of dioctahedral vermiculite with d001 = 12.6 Å, which has its 2:1 layers twisted in the same_ way as in 1M and produces the strongest reflection 133 at these wavelengths. These phases, possibly also with 3D ordering, were described earlier by Sakharov et al. (1999).
1124
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
Fig. 3. XRD patterns of zone 13L–20L in different samples: a, Lb 10, Kama River, P2kz3; b, Lm 23, Kama mouth, P2tat1; c, Mu 15, Kholmy area, well 31015, depth 1806 m, reservoir D0; d, Mi 46 (Fig. 1). See text for explanation.
The volume of particles that contribute to 3D diffraction in TEM images is much greater than in the case of XRD basal reflections. For this reason, the 1/Å scale patterns better resolve ordered phases with R = 1 and R = 2 than those of 00L reflections which miss superperiod reflections for sample Mi 46. This is because the basal reflections from mica and chlorite, zeroed down by subtraction of spectra, are orders of magnitude stronger than the superperiod reflections, which are our main interest in this study. If the difference spectrum from ordered structures with R = 1 or R = 2 lacks reflections of other phases, it should bear local peaks from air-dried samples and local troughs of glycolated samples, corresponding to superperiod reflections. Rakhmatullina and Krinari (2013) discussed an example of a 00L pattern of Na-rectorite that contained a series of superperiod reflections when the illite component reached a concentration of pM = 0.56 (sample OL 01). Figure 4 shows local minimums from a glycolated rectorite (002, 003) and a local maximum 002 from an air-dried rectorite sample. The spectrum also contains local peaks and troughs from phases with R = 1 and pM much above 50% (pM = 0.78; 0.90), as well as phases with R = 0 and pM = 0.83. A compositionally similar sample of water-bearing siltstone OL 03 lacks rectorite reflections, while the difference spectrum reveals phases with R = 1 at pM = 0.86 and pM = 0.95, which confirms the possibility for coexistence of structurally similar phases with different pM values. In the same way, the difference spectrum of sample Mi 46 includes almost fully ordered mix structures,
including those revealed by scanning of the 13L–20L zone (Fig. 3d). There are local peaks and troughs of superperiod reflections of different phases with R = 1 and R = 2, which contain either 1H2O or 2H2O layers in interlayers. Furthermore, there are phases with R = 1 and R = 2 and pM much greater than for ordered structures (Fig. 5). The suggested mechanism imposes no limitations on relative percentages of components for phases with R = 1 and R = 2. K1+ ions can be captured in expandable interlayers also after twisting, which increases pM and the total number of phases with different pM values. Indeed, the analyzed samples showed no phases with R = 1 and pM < 0.5 or R = 2 and pM < 0.66. This fact, along with low concentration, most likely impedes the identification of phases with R = 1 and R = 2 in basal reflection images (Srodon′ , 1980), while the superperiod reflections attenuate rapidly as pM increases and their position changes. Dislocation growth (Hirth and Lothe, 1982), which may release mechanic stress, occurs by partial dissolution and concurrent precipitation, and the process may involve any layer and layer structure. Let the particle size dependence of surface energy density in (1) be v
dσ = (dσ / dr) dr.
(3)
The term A (dσ / dr) expressed via the dislocation energy density is
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
1125
Fig. 4. Difference spectrum: sample OL 01 with rectorite (Fig. 1). Abbreviations: e.g., ehtylene glycol; exp, experimental spectrum.
A
r dσ AαGb 2 = ln . dr 4π (1 − v) ρ
Discussion, main conclusions, and practical applications (4)
These equations can be used to find the energy contribution of dislocation growth to illitization, with G = 1.5 × 109 Pa and v = 0.2, which is from 10 to 16 J/cm3, depending on the Burgers vector. Note that the energy contributions of the second and third mechanisms are very similar. Therefore, structures produced by both mechanisms consuming the excess energy may coexist in the system subject to transformations. The energy found by (2) is 10 J/cm3 for the mix stack I : S : I : S, with a characteristic total thickness of 2.5 nm, and 16 J/cm3 for the stack I : I : S : I : I : S with a total thickness of 3.5 nm. With reference to equation (2), one can estimate the thickness rc of an ordered mix phase with R = 1 or R = 2, which corresponds to the minimum energy of the whole system in a unit volume: ), β = A / (12πh 2) + α, 7A / (4√ 3πβ rc = √
(5)
where α is the specific system energy acquired by forming mica particles at the account of K substitution for interlayer cations; A is the Hamaker constant; h is the thickness of the water interlayer. Calculations following Deryagin et al. (1987) lead to rs = 2.2 nm if the presence of dislocations is neglected and to two values of rs = 2.2 nm and 3.2 nm for R = 1 and R = 2, respectively, if the energy contribution of dislocations is taken into account. The difference is significant but not crucial for coexistence of both structures in a single sample (Fig. 5). The difference spectra often include diffractions from phases with R = 0 which makes basis for the curve, as well as superposed local maximums and minimums of phases with R ≥ 1.
Earlier Huang et al., (1993) suggested using the model of illite-to-smectite conversion as a geothermometer. However, illitization may occur by several mechanisms and the behavior of smectite contents (pS) as a function of temperature or depth may differ in specific sections and change successively with increasing burial depth. See, for instance, pS = f (T) curves for two areas in the North Sea (Fig. 6) borrowed from (Srodon′ and Eberl, 1984), with three segments. If illitization is a direct low-temperature process (mechanism 1), the curves pS = f (T) only shift along the x axis, the amount of shift being controlled by biological processes rather than by temperature. The next segment is relatively straight and corresponds to smectite-toillite conversion by increasing illite component (mechanism 2). Another smooth bending and slope change of the pS = f(T) curve records the onset of dislocation growth (mechanism 3) as the sediments reach greater depths. The interstratified smectite–illite structure can be used as a geothermometer provided that the true temperature of some section part is known somehow a priori. Detecting zones of higher temperatures due to biochemical oil oxidation or local water inputs from deeper strata into reservoir rocks appears to be of greater practical value. Variance in local depth-dependent pS values observed in explicitly documented sections (Lanson et al., 2009) may be due to variations of wedging pressure caused by pore pressure changes which affect the rotation ability of clay layers, i.e., by porosity changes. The structure of ordered mixed-layer illite–smectite should depend on temperature. As a general tendency, R = 2 structures are found at greater core depths, while the total number of R = 1 phases increases. No more than two R = 1 phases were commonly found in test wells or in originally water-bearv
1126
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
Fig. 5. Difference spectrum: sample Mi 46 (Figs. 1, 3). a.d., Air-dried.
v
Fig. 6. Smectite-to-illite conversion in shales from different sedimentary basins, plotted as a function of temperature, after (Srodon′ and Eberl, 1984). Dashed lines delineate zones with one predominant illitization mechanism (1, 2, or 3).
ing sediments where two-phase flow was impossible. On the contrary, R = 1 and R = 2 structures coexisted in reservoir or producing rocks, which suggests the effect of percolation (Hunt and Ewing, 2009). In the case of two-phase flow, the medium parameters may be different for water and oil migrating along disconnected local paths. This tendency, if confirmed for a greater number of sections, may be used for reference in concurrent development of several simultaneously drilled reservoirs. For example, the difference spectra of sample N 915 (Fig. 7) show two R = 2 and one R = 1 phases at different numbers of H2O layers and pM values. Additionally, there are broader peaks of R = 0 phases at pM = 74 and pM = 95%. The latter may correspond to an almost monomineralic mica (but not illite), with scarce expandable interlayers, which formed late during the smectite-to-illite conversion.
Iillitization mechanisms 2 and 3 can be easily effectuated and applied in industrial technologies. There are methods for obtaining composite materials based on mechanically dispersed micas with a specific surface area ten times smaller than in the product composed of mix nanoparticles with R = 1 or R = 2. One can select a monomer which would interact with the surface of nanoparticles and penetrate into expandable interlayers between 2:1 layers. The mechanic properties of this composite material, after polymerization, should provide tight bonding of twisted silicate layers which would require much greater force to break down than the separation of organic molecular chains from the mineral surface. In addition to natural or manmade direct smectite-to-illite conversion leading to the formation of ordered secondary mica with a mosaic structure, there should be a reverse process
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
1127
Fig. 7. Difference spectrum: sample N 915 (Fig. 1).
driven by chemical or biochemical mechanisms producing high mechanical stress. This process in sediments with low open porosity may cause deformation of fittings and secondary cracking (Krinari et al., 2014). In clastic reservoirs, it leads to breakdown of secondary mica particles into isolated nanoparticles with a high surface charge. This may happen during injection of fresh water into oil reservoirs and create a return electroosmotic flow which reduces the formation permeability (Krinari and Khramchenkov, 2011). Reservoirs may become impermeable as the pore water salinity reduces to 90 g/L (a density of 1.09 g/L) while the nanoparticles reach a concentration of 0.1 vol.% of the rock skeleton (which causes almost no porosity decrease) and a number of N ≈ 7 × 1018/m3, with a basal surface area of S = 44 × 1022 nm2 and a noncompensated charge of Q = 2.8 × 105 C/m3. Such obstruction of permeability was observed in the Romashkino field (Krinari and Khramchenkov, 2011), with the reaction zone always about 10 m ahead of the flood front (Krinari et al., 2013). Water breakthrough at local reservoir sites can be mitigated using various chemical agents (Muslimov, 2003), but all of them change irreversibly the porosity of the reservoir which fails to regain its initial state. In the case of mica nanoparticles and related inhomogeneity of pores, the permeability depends more on pore fluid salinity, which is manageable. The electroosmotic screens produced by a freshwater flow, compress it vertically and then split into several arms while bypassed oil may preserve in the reservoir. These screens can be moved using flow management technologies (Muslimov, 2003), and their performance can be improved notably if the reservoir can return to its original permeability. The electroosmotic screens can block water breakthrough in any clastic reservoir, whether they contain secondary micas or not.
The study was supported by grant 15-11-10015 from the Russian Science Foundation.
References Churaev, N.V., 1990. Physicochemical Background of Mass and Heat Transfer in Porous Materials [in Russian]. Khimiya, Moscow. Deryagin, B.V., Churaev, N.V., Muller, V.M., 1987. Surface Forces [in Russian]. Nauka, Moscow. Dong, H., Peacor, D., Freed, R., 1997. Phase relations among smectite, R1 illite–smectite, and illite. Am. Mineral. 82, 379–391. Drits, V.A., Sakharov, B.A., 1976. XRD Analysis of Mixed-Layer Minerals (GIN Transactions, Issue 295) [in Russian]. Nauka, Moscow. Drits, V.A., Lindgreen, H., Sakharov, B.A., Jakobsen, H.J., Fallick, A.E., Salyn, A.L., Dainyak, L.G., Zviagina, B.B., Barfod, D.N., 2007. Formation and transformation of mixed-layer minerals by Tertiary intrusives in Cretaceous mudstones, West Greenland. Clays Clay Miner. 55, 260–283. Dudek, T., Srodon′ , J., Eberl, D.D., Elsass, F., Uhlik, P., 2002. Thickness distribution of illite crystals in shales. XRD, HRTEM measurements. Clays Clay Miner. 50, 85–91. Frank-Kamenetsky, V.A., Kotov, N.V., Goilo, E.A., 1983. Transformation of layered silicates at high P-T parameters [in Russian]. Nauka, Leningrad. Galimova, S.E., Krinari, G.A., Morozov, V.P., 1990. Intercalation and real interlayer structure of kaolinite crystals: A kinetic model, in: Spectroscopy, Crystal Chemistry, and Real Structure of Minerals and Their Analogs [in Russian]. Kazan University, Kazan, pp. 30–51. Hirth, J.P., Lothe, J., 1982. Theory of Dislocations. Krieger Publishing Company, Malabar. Huang, W., Longo, J., Peyear, D., 1993. An experimentally derived kinetic model for smectite-to-illite conversion and its use as a geothermometer. Clays Clay Miner. 41, 162–177. Hunt, A., Ewing, R., 2009. Percolation Theory for Flow in Porous Media, Lection Notes Physics 771. Springer, Berlin–Heidelberg. Khramchenkov, M.G., 2008. Fan-shaped model of clay swelling process, in: Thermo-Hydromechanical and Chemical Coupling in Geomaterials and Applications. Polytech-Lille, France, pp. 297–304. v
1128
G.A. Krinari and M.G. Khramchenkov / Russian Geology and Geophysics 59 (2018) 1120–1128
Krinari, G.A., Khramchenkov, M.G., 2005. Low-temperature illitization of smectite as a biomineral process. Dokl. Earth Sci. 403A (6), 926–930. Krinari, G.A., Khramchenkov, M.G., 2008. Three-dimensional structure of secondary micas in sedimentary rocks: features and mechanisms of formation. Dokl. Earth Sci. 423A (9), 1435–1439. Krinari, G.A., Khramchenkov, M.G., 2011. Inverse transformation of secondary micas in sedimentary rocks: mechanisms and applications. Dokl. Earth Sci. 436 (5), 674–679. Krinari, G.A., Khramchenkov, M.G., Rakhmatulina, Yu.Sh., 2013. Mechanisms of inverse transformations of secondary micas: evidence from changes in the illite–smectite structure. Dokl. Earth Sci. 452 (2), 1014–1019. Krinari, G.A., Khramchenkov, M.G., Rakhmatulina, Yu.Sh., 2014. Changes in the structures of mixed-layer illite–smectite during flooding of terrigenous oil reservoirs. Russian Geology and Geophysics (Geologiya i Geofizika) 55 (7), 915–928 (1153–1167). Krinari, G.A., Shinkarev, A.A., Giniyatullin, K.G., 2005. Abiogenic and biogenic degradation of minerals: difference, mechanisms, and practical applications. Zapiski RMO, No. 1, 18–32. Lanson, B., Sakharov, B.A., Claret, F., Drits, V.A., 2009. Diagenetic smectite-to-illite transition in clay-rich sediments: a reappraisal of X-ray diffraction results using the multi-specimen method. Am. J. Sci. 309, 476–516. Muslimov, R.Kh., 2003. Modern Methods for Management of Oil Reservoirs with the Use of Flooding [in Russian]. Kazan University, Kazan.
Rakhmatullina, Yu.Sh., Krinari, G.A., 2013. Formation mechanisms of secondary micas: Effect on the character of illite–smectite phases in flooding of oil reservoirs. Georesursy 53 (3), 27–30. Sakharov, B.A., Lindgreen, H., Salyn, A., Drits, V.A., 1999. Determination of illite–smectite structures using multispecimen X-ray profile fitting. Clays Clay Miner. 47 (5), 555–566. Solotchina, E.P., 2009. Structure Typomorphism of Clay Minerals in Sedimentary Sections and Weathering Profiles [in Russian]. Geo Publishers, Novosibirsk. Srodon′ , J., 1980. Precise identification of illite/smectite interstratifications by X-Ray powder diffraction. Clays Clay Miner. 28, 401–411. Srodon′ , J., Eberl, D., 1984. Illite, in: Bailey, S.W. (Ed.), Micas. Rev. Mineral., Mineral. Soc. Am. 13, 495–544. Srodon′ , J., Eberl, D.D., Drits, V.A., 2000. Evolution of fundamental particle-size during illitization of smectite and implications for reaction mechanism. Clays Clay Miner. 48 (4), 446–458. Velde, B., Meunier, A., 2008. The Origin of Clay Minerals in Soils and Weathered Rocks. Springer-Verlag, Berlin–Heidelberg. Zviagina, B.B., Krinari, G.A., 1989. Geometrical features and principles of diffraction imaging of ordered-structure samples. Kristallografiya 34 (2), 288–291. Zviagina, B.B., Sakharov, B.A., Drits, V.A., 2007. X-ray diffraction criteria for the identification of trans- and cis-vacant varieties of dioctahedral micas. Clays Clay Miner. 55 (5), 467–480. v
v
v
Editorial responsibility: N.V. Sennikov