Quartz preferred orientation in naturally deformed mylonitic rocks (Montalto shear zone–Italy): a comparison of results by different techniques, their advantages and limitations

Abstract

In the geologic record, the quartz c-axis patterns are widely adopted in the investigation of crystallographic preferred orientations (CPO) of naturally deformed rocks. To this aim, in the present work, four different methods for measuring quartz c-axis orientations in naturally sheared rocks were applied and compared: the classical universal stage technique, the computer-integrated polarization microscopy method (CIP), the time-of-flight (TOF) neutron diffraction analysis , and the electron backscatter diffraction (EBSD). Microstructural analysis and CPO patterns of quartz, together with the ones obtained for feldspars and micas in mylonitic granitoid rocks, have been then considered to solve structural and geological questions related to the Montalto crustal scale shear zone (Calabria, southern Italy). Results obtained by applying the different techniques are discussed, and the advantages as well as limitations of each method are highlighted. Importantly, our findings suggest that patterns obtained by means of different techniques are quite similar. In particular, for such mylonites, a subsimple shear (40% simple shear vs 60% pure shear) by shape analysis of porphyroclasts was inferred. A general tendency of an asymmetric c-maximum near to the Z direction (normal to foliation) suggesting dominant basal slip, consistent with fabric patterns related to dynamically recrystallization under greenschist facies, is recognized. Rhombohedral slip was likely active as documented by pole figures of positive and negative rhombs (TOF), which reveal also potential mechanical Dauphiné twinning. Results showed that the most complete CPO characterization on deformed rocks is given by the TOF (from which also other quartz crystallographic axes can be obtained as well as various mineral phases may be investigated). However, this use is restricted by the fact that (a) there are very few TOF facilities around the world and (b) there is loss of any domainal reference, since TOF is a bulk type analysis. EBSD is a widely used technique, which allows an excellent microstructural control of the user covering a good amount of investigated grains. CIP and US are not expensive techniques with respect the other kind of investigations and even if they might be considered obsolete and/or time-consuming, they have the advantage to provide a fine and grain by grain “first round” inspection on the investigated rock fabric.

Appendices

Appendix 1

For XRF analyses, glass disks were prepared of Lithium-tetra-borate and sample powder (<125 µm) which are mixed to a ratio of 6:1. A wave length-dispersive instrument (Panalytical Axios) was used for major element determination except hydrogen and carbon which together were determined as Loss on Ignition at 1150 °C (detection limit: 0.6 wt%). Accuracy is documented by repetitive analyses of international and in-house reference rocks and is better than 2% for elements Ca to Fe, better than 4% for elements from Al to K, and better than 6% for Na and Mg. Precision in the form of standard deviation is in the range of ±2% (rel.) for all major elements, except for loss on ignition, which is in the range of ±5% (rel.). Quantitative mineral analyses were done on a Cameca SX100 electron microprobe with 15 kV accelerating voltage and 20 nÅ beam current calibrated with natural mineral standards. Geochemical investigations were carried out at the Institut für Endlagerforschung, Clausthal University.

Appendix 2

For the CIP analysis, thin sections of approximately 25 μm thickness were prepared to ensure that quartz appears with a first-order gray color (Heilbronner 2010). This technique allows one to represent the c-axis position of each pixel with a characteristic color (Heilbronner and Barrett 2013), and pole figures were calculated from the azimuth and inclination images. Pictures were captured on a Zeiss polarization microscope using a Zeiss AxiocamMRm monochromatic camera and a narrow band interference filter transmitting at 660 ± 9 nm (near infrared). By applying the CIP method, on each sample we have selected quartz domains following two fundamental criteria: (1) the grain size of quartz within the domain is representative of the entire thin section; (2) regions in the vicinity of large porphyroclasts which have a strong influence on the strain distribution in their surroundings were avoided. To prepare the CIP input, the freeware Image SXM software (Barrett 2008, http://www.liv.ac.uk/~sdb/ImageSXM/) was used. More details on the CIP method can be found elsewhere (Heilbronner 2000a, http://www.unibas.ch/earth/micro/). This filter renders interference colors as follows: first-order yellow as white, first-order red as gray and first-order blue as black. By inserting the filter, the interference colors (3 channels) are mapped uniquely into a gray value image (1 channel). The number of possible c-axis orientations for any given gray value is reduced with respect to the classical crossed polarizers conditions, but still there is no unique representation. If the quartz domain is acquired under so-called circular polarization conditions with crossed polarizers and two quarter wave length plate (1/4 lambda plates) inserted, these conditions yield an inclination image, where the grains appear dark if the c-axes are normal to the plane of the section, and white if the c-axes are parallel to it. Circular polarization is not sensitive to the azimuth of the c-axis. Grains which appear nearly white indicate that its c-axis lies very close to the plane of the section whereas black grains have a vertical c-axis. To map the two-dimensional orientation space, i.e., to uniquely color-code c-axis orientations, azimuth and inclination images are calculated and treated as two channels of a color image. Two-dimensional color look-up tables (CLUTs) are used to assign unique colors to any given pixel depending on the azimuth and inclination values of the c-axis at that point.

The standard input for CIP calculations consists of 18 rotation images, 2 tilt images and a circular polarization image (Heilbronner 2000b): the rotation images are a series of images with incremental relative rotation of the thin section with respect to the polarizers and the lambda plate (typically using a fixed interval of 10°), the tilt images are captured with the thin section tilted about the N–S and E–W axes. For the circular polarization image, crossed polarizers and two quarter-lambda plates (above and below the thin section) are used. Placing the input images in a stack (a multi-image sequence) and using macros, the images are registered (Heilbronner 2000b).

Additional images may be taken with parallel polarizers to obtain an image of dust and scratches for masking purposes, or one without the thin section in place in order to be able to correct uneven lighting.

Appendix 3

Nomenclature used in rigid grain net (RGN) analysis

W _m :

Mean kinematic vorticity number

$B^{*}$ :

Shape factor

θ :

Angle between clast long axis and macroscopic foliation

R :

Porphyroclast aspect ratio (long axis/short axis)

a :

Major axis of the porphyroclast

b :

Minor axis of the porphyroclast

Equations used in the RGN calculations

$$ \theta = 1/2\sin^{ - 1} W_{\text{m}} /B^{*} \left\{ {\left( {1 - W_{\text{m}}^{2} } \right)^{{{1 \mathord{\left/ {\vphantom {1 2}} \right. \kern-0pt} 2}}} - \left( {B^{*} - W_{\text{m}}^{2} } \right)^{{{1 \mathord{\left/ {\vphantom {1 2}} \right. \kern-0pt} 2}}} } \right\} $$

$$ B^{*} = (a^{2} - b^{2} )/(a^{2} + b^{2} ). $$