Metripol birefringence imaging of unconsolidated glaciotectonized and ice keel scoured sediments: identification of unistrial plasmic fabric

Please download to get full document.

View again

of 15
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Metripol birefringence imaging of unconsolidated glaciotectonized and ice keel scoured sediments: identification of unistrial plasmic fabric
  Metripol birefringence imaging of unconsolidated glaciotectonized andice keel scoured sediments: identification of unistrial plasmic fabric LORNA D. LINCH AND JAAP J. M. VAN DER MEERLinch, L. D. & van der Meer, J. J. M. 2012: Metripol birefringence imaging of unconsolidated glaciotectonizedand ice keel scoured sediments: identification of unistrial plasmic fabric.  Boreas , 10.1111/j.1502-3885.2012.00290.x. ISSN 0300-9483.In unconsolidated sediments subject to strain, clays and silts are realigned into particular optical birefringentarrangements (plasmic fabrics), which provide information on the style and intensity of sediment deformation. Arelatively new, non-destructive, optical microscopy technique for automatically recording and quantifying bire-fringence (previously commercialized under the name ‘Metripol’) is pioneered in this study as a valuable andinnovative micromorphological tool with which to examine deformation in unconsolidated sediments. Metripolis applied to unistrial plasmic fabric in glaciotectonized and ice keel scoured sediment from the Netherlands andformer Glacial Lake Agassiz (Manitoba, Canada) respectively. Colour-coded images are produced in whichcolour represents relative optical retardation and thus optical anisotropy through the quantity |sin d | and opticalorientation of anisotropy through the angle  Ø   (also indicated by linear azimuths). In this study Metripol typicallydemonstrates that the better developed the unistrial plasmic fabric is, the higher the |sin d | values, the larger theareas of high |sin d | values, and the longer and more densely populated the azimuths. In addition, some unistrialplasmic fabrics under Metripol demonstrate lower |sin d | than previous examples and the surrounding sediment,despite being ‘perceived’ as demonstrating higher birefringence under a standard petrographic microscope. Thisis particularly true in clay-rich sediments and has implications for the way we currently describe and interpretunistrial plasmic fabrics in unconsolidated sediment. Finally, the identification and quantification of additionalstructures that would otherwise have gone undetected using a standard petrographic microscope (e.g. linear andcircular structures that are likely to represent discrete shears and skelsepic plasmic fabric, respectively) highlightthe potential for Metripol to gather information on the deformation history of unconsolidated sediments that isunavailable to standard techniques. Lorna D. Linch (e-mail: and Jaap J. M. van der Meer (e-mail:, Centre for Micromorphology, School of Geography, Queen Mary University of London, Mile End Road, London E1 4NS,UK; received 7th March 2012, accepted 30th July 2012. Anisotropic materials demonstrate a variable refrac-tive index related to the direction of vibration of lightin cases where anisotropy arises mainly from thedirectionality of atomic or molecular arrangements, orfrom strain (Glazer  et al  . 1996). Plane-polarized lightthat passes through anisotropic material is split intotwo rays travelling subject to different relative indices n n and  n 1 , which can be described by an ellipsoidalcross-section of the refractive index surface known asthe optical indicatrix (Fig. 1). The difference in lengthof the two ellipse axes is the linear birefringence forlight directed along the third axes (Glazer  et al  . 1996).Between the two light paths there is a phase differ-ence, so that when the two rays recombine the finalphase difference  d   between them is a measure of the optical anisotropy of the birefringent material(Pajdzik & Glazer 2006). The optical anisotropy canthen be represented by the phase shift  d   of the light,given by δ   = − ( ) 2  1 πλ   n n t n , (1)where  l   is the wavelength of light,  n 1 - n n is the bire-fringence of the sample, and  t  is the thickness of thesample (Glazer  et al  . 1996).Optical retardation, represented by ( n 1 - n n ) t , can beobserved by placing the specimen between crossedpolarizers in a microscope, the latter of which ensurethat the light entering the microscope is not transmittedexcept when a birefringent material is encountered,whereupon striking colours are observed and subse-quent rotation of the specimen shows the extinction of the transmitted light waves every 90° (Glazer  et al  .1996). Most transparent solids (e.g. crystals, polymers,glasses and biological material) not of cubic symme-try are optically anisotropic and demonstrate opticalretardation and birefringence. Unconsolidated sedi-ment that has undergone stress may also show pat-terns of birefringence through the reorientation of claydomains. These patterns are referred to as plasmicfabrics (see Glossary) and are a well-established meansof determining the deformation history of sedimentduring micromorphological analysis (e.g. Jim 1990; vander Meer 1993; Lachniet  et al  . 1999, 2001; Menzies2000; Phillips & Auton 2000; Hiemstra 2001; Carr2004; Tarplee 2006; Kilfeather  et al  . 2010; Linch 2010;Linch  et al  . 2012).Measurement of optical retardation, and thus bire-fringence, has been a standard tool in the study of anisotropic properties for nearly two centuries (Pajdzik& Glazer 2006). The most well-established method to bs_bs_banner DOI 10.1111/j.1502-3885.2012.00290.x © 2012 The AuthorsBoreas © 2012 The Boreas Collegium  measure optical retardation is to insert compensatingcrystalline plates and rotate the sample into a numberof positions to effectively cancel the birefringence(Glazer  et al  . 1996). In samples with numerous orien-tations this method becomes extremely time consuming(Metripol Birefringence Microscopy Technique 2003).Other disadvantages of this method include (i) the dif-ficulty in detecting birefringence if there is not muchbirefringence in the sample, (ii) inaccurate sampleorientation with respect to the polarization direction of the light (making determination of birefringence innon-homogenous samples difficult), (iii) the inability torotate and compare different regions of a sample simul-taneously, (iv) limited accuracy through manual meas-urements, and (v) image interference as a result of coloured patterns (Glazer  et al  . 1996; Metripol Bire-fringence Microscopy Technique 2003; Pajdzik &Glazer 2006). As an alternative, Wood & Glazer (1980)describe a different method of accurately measuringrelative optical retardation using a modulation tech-nique resulting in a sinusoidal (in time) wave whoseamplitude is modulated by the term sine delta (sin d  ).This system is able to sense both optical anisotropythrough the quantity of sin d   and the optical orientationof this anisotropy through the angle  Ø  . It also has thepotential to obtain a signal proportional to opticalchanges across the specimen, although such measure-ments prove time consuming (Glazer  et al  . 1996).Consequently, Glazer  et al  . (1996) further developedthe modulation technique described by Wood & Glazer(1980), and in collaboration with manufacturers atOxford Cryosystems (UK), designed a new opticalmicroscopy technique for automatically recordingand quantifying birefringence, commercialized underthe name ‘Metripol’ (UK patent application numberBG9604785.7). The Metripol system uses a combina-tion of monochromatic light, a plane polarizer capableof being rotated to fixed angles  a  from a referenceposition, a circular-polarizing analyser, a CCD camera,an electronic controller and computer software forpolarizer control, image collection and analysis. Mono-chromatic light passes through a rotating polarizer andthen through the birefringent sample. The light fromthe sample then passes through a 1/4 l   plate and ana-lyser (arranged together to form a circular analyser) toprovide an image, which is captured by the CCDcamera (Pajdzik & Glazer 2006) (Fig. 2). The signal istransferred to a computer where it is processed by spe-cially written software that separates three types of image components that would normally be superim-posed in conventional polarizing microscopy (Pajdzik& Glazer 2006). These image components are catego-rized as follows: (i) the optical anisotropy, namelyquantitative information on the modulus of sine delta(|sin d  |) at any point within the image, where delta  d   isthe phase difference introduced by the birefringentsample i.e. |sin d  | is a measure of birefringence; (ii) theoptical orientation ( Ø  ) of anisotropy of one of the axesof a section of the optical indicatrix measured from apredetermined direction i.e. birefringence orientationknown as the azimuthal angle (the value of the azi-muthal angle  Ø   is measured anticlockwise from thehorizontal direction within the image captured by theCCD camera (Pajdzik & Glazer 2006)); and (iii) a rep-resentation of light transmission through the samples( I  0 ) i.e. transparency (Glazer  et al  . 1996; Lewis &Glazer 1996; Metripol Team 2002; Metripol Birefrin-gence Microscopy Technique 2003; Pajdzik & Glazer2006). The values of each component are assigned false,coded colours by an integrated software suite separatedinto three images corresponding to the above catego-ries. In this way, the three components can easily bequantified and monitored spatially with the resolutionof a traditional light microscope (Lewis & Glazer Fig. 1.  The optical indicatrix. Refractive indices are  n 1 and  n n , andthe slow direction (the largest refractive index,  n 1 ) is oriented at anangle  Ø   from the horizontal (modified from Glazer  et al  . 1996). Fig. 2.  Schematic diagram of the Metripol (modified from Glazer et al  . 1996). 2  Lorna D. Linch and Jaap J. M. van der Meer  BOREAS  1996). In the present system 100 colours are used inimages for the |sin d  | terms, thus giving a resolution of 1 nm†, while for the orientation  Ø   images 180 coloursare used, giving a resolution of 1° (Glazer  et al  . 1996).Glazer  et al  . (1996) found that the use of colour has asignificant advantage over grey-scale, where the slowchange of intensity makes it difficult for the human eyeto see small changes from pixel to pixel. In the colouredplates there is no black or white, as these colours arereserved for particular circumstances: (i) pixels areautomatically set to black in areas where light intensitybecomes too low, typically when the area being exam-ined is strongly absorbing; (ii) pixels are automaticallyset to white when |sin d  | becomes close to zero and thusthe orientation  Ø   becomes indeterminate. This happenswhenever an area is isotropic or when retardation is amultiple of   l  /2 (Glazer  et al  . 1996).The obvious advantages to Metripol are that: thethreequantitiesof|sin d  |, Ø  and I  0 canbemeasuredfromall parts of the sample image simultaneously; the tech-nique is extremely sensitive; and it does not depend onthe orientation of the sample to a particular angle withrespect to the polarization direction (Pajdzik & Glazer2006). Metripol is already being successfully applied toqualitatively and quantitatively analyse, for instance,strain in industrial diamonds (Howell  et al  . 2006), andbiological birefringent specimens, for example collagenand hydroxyapatite distribution in bone (Glazer  et al  .1996; Lewis & Glazer 1996; Metripol Team 2002;Metripol Birefringence Microscopy Technique 2003).Untilnowhowever,MetripolhasnotbeenusedinEarthSciences. The aim of this study is to explore and pioneerMetripol as a useful and innovative, qualitative andquantitative technique in sedimentology. The emphasiswill be in particular on the identification and analysis of unistrial plasmic fabric, which typically forms simple,easily identified linear patterns of birefringence under astandard petrographic microscope (see Glossary). Toachieve the overall aim, the following objectives areaddressed: (i) to use Metripol to confirm and quantifythe presence of microscopically identified unistrialplasmic fabric in sheared, glaciotectonized sedimentfrom the Netherlands (van der Meer 1993), which pro-vides one of the best known examples of unistrialplasmic fabric from which to gather initial informationunder Metripol; and (ii) to use Metripol to confirm andquantifythepresenceofmicroscopicallyidentifiedunis-trialplasmicfabricinsheared,icebergscouredsedimentfrom former Glacial Lake Agassiz (GLA), Manitoba,Canada (Linch 2010; Linch  et al  . 2012) to test Metripolagainst objective (i). Methods Micromorphological samples Thin section number R745, retrieved from Lunteren inthe central Netherlands, was chosen for Metripolanalysis. The sediment comprises mainly subglaciallydeformed till interspersed with rhythmites. The rhyth-mites demonstrate well-developed unistrial plasmicfabric (van der Meer 1993) (Fig. 3). Thus, thin sectionR745 provides one of the best, well-established andpublished examples of plasmic fabric from which togather initial data from Metripol. Subsequently, seventhin sections were selected from a collection related toa study of the micromorphological characteristics of iceberg scoured sediment (Linch 2010; Linch  et al  .2012). Two samples come from the non-iceberg scouredUpper Sherack Formation (USF), and five come fromthe iceberg scoured Brenna Formation (BF) from GLA(Clayton  et al  . 1965; Dredge 1982; Woodworth-Lynas& Landva 1988; Woodworth-Lynas & Guigné 1990;Woodworth-Lynas 1992; Linch 2010; Linch  et al  . Fig. 3.  Thin section R745, Lunteren, theNetherlands. Boxes: 1  =  well-developedunistrial plasmic fabric; 2  =  well-developedunistrial plasmic fabric (lower line) andmoderately developed plasmic fabric(upper line); 3  =  moderately developedunistrial plasmic fabric; 4 and 5  =  weaklydeveloped unistrial plasmic fabric; 6  =  anarea free of plasmic fabric; 7  =  patchybirefringence of surrounding material;8  =  continuous birefringence of surround-ing material; 9 and 10  =  quartz grains;11  =  Fe staining (modified from van derMeer 1993). This figure is available incolour at Metripol birefringence imaging of unconsolidated sediments  3 BOREAS  2012). Thin sections were specifically selected fromGLA sediment for two reasons: (i) sediment is predomi-nantly clay, which implies that plasmic fabric had thebest chance to develop, for example during icebergscour; and (ii) the thin sections were found to exhibitclear examples of unistrial plasmic fabric (under astandard petrographic microscope) (Linch 2010; Linch et al  . 2012).In order to ascertain how unistrial plasmic fabricappears under Metripol, it is important to examine it atvariable perceived degrees of development, which is thestandard practice in micromorphological analysis.Linch (2010) and Linch  et al  . (2012) categorize plasmicfabric as (i) well-developed, (ii) moderately developed,or(iii)weaklydeveloped,basedonitsappearanceundera standard petrographic microscope. Well-developedunistrial plasmic fabric characteristically appears as abright, sharp, yellow to orange line of variable thick-ness, reaching an angle of extinction as the stage isturnedthrough90°.Incontrast,weaklydevelopedunis-trial plasmic fabric appears as a dull and faded, indis-tinct, yellow to orange line of variable thickness, whilemoderately developed unistrial plasmic fabric isdescribed as being somewhere between well-developedand weakly developed unistrial plasmic fabric.Thin section R745 from Lunteren, as seen undercross-polarizers, demonstrates the complexities of plasmic fabric development (Fig. 3). The yellow-golden, curved, linear elements are the unistrial plasmicfabric. Microscopic analysis concentrates on theplasmic fabric pattern and its strength, and thus it isnoted that the thickness of the pattern and strength of the birefringence vary over the image (1–2), despite thefact that it is surrounded by sediment that itself showsa variable strength of birefringence (3–4–5). However,it should be noted that when the stage is rotated theapparent pattern and strength will change. Further-more, the wider host sediment shows a different patternof plasmic fabric, with its own variable strength of birefringence (from 6 non-existent, to 7 patchy and 8continuous), while quartz grains show different colours(9–10), and some areas are obscured by iron (Fe) stain-ing (11). The advantage of the human eye is that it candiscern the pattern, while its limitation is that it canonly see what is in the current image. The advantage of the Metripol system is that it ‘sees’ birefringencestrength in a multitude of images, while its weakness isthat it cannot select the pattern. These differences inobservation mean that Metripol may not ‘see’ a patternthat is dominant to the human eye if the surroundingbirefringence is high. Finally, it must be noted that aMetripol image is much smaller than the microscopicimage in Fig. 3.The current method in the discrimination of plasmicfabrics of variable (perceived) degrees of developmentremains basic, with an obvious disadvantage that itrelies on subjective observations. The ability to quan-tify birefringent properties of unistrial plasmic fabricwithin sediment remains the main objective for apply-ing Metripol. To do so however, it is also important toapply Metripol to areas of sediment where there isapparently no microscopically detectable plasmicfabric. All thin sections in this study, as well as specificlocations selected for examination within those thinsections, were chosen on the basis that they demon-strate previously established unistrial plasmic fabric of variable (perceived) development, and/or areas whereplasmic fabric is (perceived to be) absent (Table 1). Ineach thin section sample, at least two examples of unis-trial plasmic fabric or areas free of plasmic fabric wereselected for analysis. Unistrial plasmic fabrics wereselected using a standard microscope from thin sectionR745 (van der Meer 1993), including (i) well-developed,(ii) moderately developed and (iii) weakly developedfabrics, which were then analysed under Metripol.Voids, clasts and Fe stains, as identified under a stand-ard microscope, were also analysed under Metripol,thus providing reference points. Once unistrial plasmicfabric of variable (perceived) development and areas(perceived to be) free of plasmic fabric were recognizedin thin section R745 under Metripol, the same stepswere taken to recognize unistrial plasmic fabric of vari-able (perceived) development and areas (perceived tobe) free of plasmic fabric in thin sections selected fromGLA (Table 1). Metripol  Metripol was calibrated following the steps in the‘Metripol Birefringence System Operation and Instruc-tion Guide’ (Oxford Cryosystems © 1999–2002). Thesystem was operated using a 550 nm filter, and magni-fication remained constant at 10 ¥ , which allowed amaximum field of view of 1360  m m width for oneMetripol image. Images were consistently measured ina series of 10 automatic rotational steps (Oxford Cryo-systems © 1999–2002). For the purpose of this studyattention was focused on |sin d  | images, although someinformation on orientation  Ø   was gained from apply-ing azimuth lines (linear indicators that lie parallel tothe slow axis ( n 1 ) in the optical indicatrix). It should berealized however, that although |sin d  | values obtainedat any point in the image are a measure of birefrin-gence, in the absence of absolute birefringent measure-ments it is not possible to gain information on thenumber of periods (known as the orders) that the sinefunction has passed through (Glazer  et al  . 1996). Con-sequently, it is not possible to determine directly theactual value of the phase difference and thus the opticalretardance and birefringence of the sample usingMetripol alone (Pajdzik & Glazer 2006). This alsomeans that each time |sin d  | passes through zero, thecomputed value of the azimuthal angle  Ø   changesthrough 90° (Pajdzik & Glazer 2006). To correct this, 4  Lorna D. Linch and Jaap J. M. van der Meer  BOREAS  the order within which the retardation values lie can bedetermined by using a standard compensating platewith white light or by carrying out a multi-wavelengthmeasurement (Glazer  et al  . 1996; Pajdzik & Glazer2006). In this study, an Ehringhaus rotary compensatorwith calcite combination plates was used to determinethe order of retardation for each feature analysed ineach sample. Observations using this technique suggestthat we can be confident that all features and surround-ing material analysed in this study in all thin sectionsshow average retardation values in the first order of thesine function. In turn, this indicates that in samplesused in this study, the |sin d  | coloured scale bar (0–1) onimages corresponds directly to an average total bire-fringence (Table 2), and that the computed value of theazimuthal angle  Ø   does not change through 90 ° acrossimages. Azimuths were added to |sin d  | images to helpdelineate patterns of average birefringence and facili-tate identification of unistrial plasmic fabric. Azimuthswere kept consistent at 150  m m long (the value of whichis proportional to |sin d  |), and their average density waskept consistent over an area of 15  m m 2 . These param-eters were considered the most useful in order to delin-eate birefringent patterns without azimuths being toolong/short or too dense/sparse that they might maskand/or fail to delineate birefringent patterns. Someimages were stitched together using standard C  software to increase image dimensions. Grain size and carbonate content Clay platelets in strained, unconsolidated sediment arerealigned to form oriented domains with specific bire-fringent patterns (plasmic fabric). Fine-grained sedi-ment has low permeability and thus, when stressed,high pore-water pressures are more likely to developwithin it, which facilitates sediment deformation (Benn& Evans 1998) and the potential for plasmic fabric.Therefore, it is important to know grain size when com-paring the abundance and development of plasmicfabric in different samples. Grain sizes of the rhyth-mites in thin section R745 comprise  ~ 40–50% clay,50–60% silt and 5–10% sand (van der Meer  et al  . 1985). Table 1.  Thin section numbers, locations and lithofacies; and abundance and development of unistrial plasmic fabric within those thin sections.Key:  = High abundance, well-developed; ••• =  High abundance, moderately developed;  =  High abundance, weakly developed;  =  Moderate abundance, well-developed; •• =  Moderate abundance, moderately developed;  =  Moderate abundance, weakly devel-oped;  =  Low abundance, well-developed; • =  Low abundance, moderately developed;  =  Low abundance, weakly developed (after Linch et al  ., 2012) . Micromorphological thinsection numberFigure Location and lithofacies Abundance anddevelopmentWell-developed unistrial plasmic fabricR745 4C/D Lunteren   2B1aa 5A Glacial Lake Agassiz:Brenna Formation  2B2aa 5B Glacial Lake Agassiz:Brenna Formation  Moderately developed unistrial plasmic fabricR745 4D/E Lunteren •• 4B2aa2 5C Glacial Lake Agassiz:Brenna Formation  •• 1B2aa 5D Glacial Lake Agassiz:Brenna Formation  •• Weakly developed unistrial plasmic fabricR745 4F/G Lunteren   3L3aa 5E Glacial Lake Agassiz:Upper Sherack Formation  2M4aa 5F Glacial Lake Agassiz:Brenna Formation  Areas free of plasmic fabricR745 4H Lunteren – 1M1aa 5G Glacial Lake Agassiz:Upper Sherack Formation –  Table 2.  Metripol |sin d  | colour scale in approximate percentages . Colour scale |sin d  | numericalscale (0–1)|sin d  |percentages (%)Dark red 0.9–1 90–100Red 0.8–0.9 80–90Orange 0.7–0.8 70–80Yellow 0.65–0.7 65–70Green 0.4–0.65 40–65Light blue 0.3–0.4 30–40Dark blue 0.1–0.3 10–30Violet 0.1 10 Metripol birefringence imaging of unconsolidated sediments  5 BOREAS
Related Documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks