Examining Sedimentary Rocks with Cathodoluminescence Imaging

Cathodoluminescence spectroscopy is a robust technique for the micro-characterization of rocks and minerals, harmonious with other electron microscopy-based tools, such as backscattered electron imaging and energy/wavelength dispersive x-ray spectroscopy.

This article demonstrates how the methodology can be implemented to examine quartz sandstone, a common category of sedimentary rock, which has garnered both fundamental interest as well as practical interest, in the form of fossil fuel exploration and extraction.

In particular, panchromatic (unfiltered), color-filtered, and hyperspectral cathodoluminescence imaging are used to uncover relevant textures and contrasts at the microscale, which can assist in determining geological chronology, and can also bolster porosity/permeability studies of such rocks.

Introduction

Sedimentary rocks are composed of a specific category of rocks, which are configured by the weathering of the Earth’s crust, the accretion of shells secreted by organisms on the seabed, and the precipitation of minerals from water. They are conventionally formed via the lithification of granular material, upon which pressure and temperature is exerted as the rock is embedded deeper in the Earth’s crust (if the deformations and alterations in the material are too severe, the rock is categorized as metamorphic). The grain size can differ, from under 4 μm for clay materials, up to over 25 cm for certain rudite materials.

As a result of their granular structure, sedimentary rocks are typically rather permeable to fluids such as water. Minerals can precipitate from this infused water to constitute a cement that firmly secures the individual grains in the rock. Sedimentary rocks offer a precious source of iron and natural fertilizer.

Moreover, they represent the primary source of fossil fuels (oil, natural gas, and coal) and therefore embody tremendous practical worth. In this regard, the porous matrix in the rock dictates the oil/gas holding function, as well as the transport characteristics which are relevant for the extraction of these resources [1].

Sandstone is a specific category of sedimentary rock which has robust applicability within the petroleum sector, as it represents one of the primary rock groups in which fossil fuels can be derived. In sandstone, at least 50% of the grains are required to have a size between 60 and 200 μm. Sandstones regularly possess a high quartz content, as quartz is more impervious to weathering in relation to other familiar minerals such as feldspar and mica.

Likewise, cement can be present in sandstone which can robustly impact on its porosity and permeability. The microscale/nanoscale topology and texture can be utilized to deduce the geological history of the rock. Moreover, they directly modify the total chemical, mechanical, and fluid transport characteristics of the rock, which makes examining the rocks at those length scales of particular interest.

During the last decades, scanning electron microscope (SEM)-based cathodoluminescence (CL) spectroscopy has become a manifestly efficacious methodology for the micro-characterization of sedimentary rocks [2-4]. More specifically, it dispenses data that is complementary to SEM based tools, including energy and wavelength dispersive x-ray spectroscopy (EDS, WDS), and electron backscattering imaging (BSE, EBSD).

Conventionally, the CL yield exceeds the minimum criteria for rapid- and, in some instances, even video-rate scanning, enabling the rapid examination of reasonably substantial areas. In the case of sandstone, CL imaging can be implemented to quantitatively map the quartz profile of the sample which allows, for example, the meticulous segmentation of granular material and cemented material.

By integrating such outcomes with in-depth investigation of the detected textures, such as (healed) fractures and grain contacts, the geological history, coupled with the porosity of the rock, can be examined in a detailed manner. Due to the first rate spatial resolution, particularly delicate details in the structure can be exposed. This can be utilized to adjust information that derives from less refined imaging methodologies, such as 3D micro-computed x-ray tomography (μCT) or optical microscopy studies.

These results can represent valuable input for progressive mechanical and fluid dynamics simulations that are used to ascertain the macroscopic characteristics of the rock. While such studies are of fundamental interest, the petroleum industry also employs them for the specific examination of reservoir rocks.

Here, CL measurements are carried out on quartz sandstone samples. Rapid ‘large-area’ measurements with a photomultiplier tube (PMT) have been carried out to determine the overall CL structure of rock and more infinitesimal hyperspectral measurements with a CCD spectrometer, to achieve a more thorough insight into particular sections. CL images are compared with the SEM images, the observed textures and contrasts are described and the wavelength dependence of the CL emission for these rock samples are studied.

Experimental System and Sample Preparation

The experiments were executed in a FEI XL-30 SFEG scanning electron microscope (SEM), equipped with a prototype of the SPARC CL-system. The CL is gathered by an aluminum paraboloid mirror, which is correctly aligned with the use of piezoelectric actuators. The CL detection was undertaken using an uncooled analog PMT (λ0 = 230 – 870 nm spectral range where λ0 is the free-space wavelength).

This was utilized for panchromatic imaging (no color filtering), and wavelength-filtered imaging, in which a band pass filter was located in front of the PMT. A fiber-coupled Czerny-Turner spectrometer was used with a cooled back-illuminated silicon camera (λ0 = 350 – 1000 nm spectral range) for hyperspectral imaging. A more comprehensive description of the SPARC CL network can be found in [5].

Experiments were performed on two related bulk quartz sandstone samples, which both possessed a polished surface in order to reveal the structure more effectively. It was noted that the surface preparation can have a significant impact on the experimental end product and should be approached with caution [2]. The samples were coated with 30 nm of amorphous carbon for proper electrical conduction. Figure 1 provides a photograph of sample 1.

Photograph of sandstone sample 1. This sample is 1 cm in diameter. Sample 2 is 2 cm in diameter.

Figure 1. Photograph of sandstone sample 1. This sample is 1 cm in diameter. Sample 2 is 2 cm in diameter.

Cathodoluminescence Imaging of Sandstones

Fast large-area CL scanning

Thin-sections of polished sandstone (45 x 25 mm is a standard size for such a section) or surface-polished sections of rock which have been extracted from larger drill cores, such as the sample exhibited in Figure 1 (ordinarily 1 – 3 cm in diameter), generate a substantial surface area for SEM research. To acquire an adequate summary of such a ’vast‘ sample, the scanning of relatively large areas is of particular interest. This is facilitated by a selection of experimental choices.

Firstly, a robust CL signal is favorable for rapid inspection, and the CL output from the sample is boosted by a high acceleration voltage and current. Secondly, the detector needs to satisfy specific read-out time speed criteria, and should assimilate a ‘large’ field of view, which is regulated by the detector size and the optics. Thirdly, the light has to be gathered and directed to the CL detector in an efficient manner. This is transacted using the SPARC’s motorized mirror collection system, which permits exemplary reproducible optical alignment.

Effective signal collection is double advantageous: numerous minerals, including quartz, can be structurally affected by the electron beam if the dose is excessive, upsetting the emitted signals and bringing about possible data misinterpretations [4,6]. By discerning the signal with optimal efficiency, the electron dose per unit area can be minimized, which enhances methodological reliability.

(a) SEM image of an area on sample 2. 17 megapixel panchromatic PMT image, collected with a dwell time of (b) 800 ns and (c) 10 μs of the same area on the sandstone. The color scaling is adjusted to the dynamic range of the image for proper visibility of all the features.

Figure 2. (a) SEM image of an area on sample 2. 17 megapixel panchromatic PMT image, collected with a dwell time of (b) 800 ns and (c) 10 μs of the same area on the sandstone. The color scaling is adjusted to the dynamic range of the image for proper visibility of all the features.

To act in accordance with the aforementioned demands, the PMT detector was implemented in panchromatic mode, i.e. spectrally unfiltered, at 15 kV acceleration voltage and a 5 nA probe current. For sample 1 a 400 pA current and 10 kV voltage were used, because the conductive layer did not function as effectively as for sample 1. The PMT amplification was programmed to equal the dynamic intensity range of the sample.

Figure 2(a) exhibits a SEM image of an area in the sandstone. Singular crystal grains can be observed from the surface topology, but as the material is the same (quartz), there exists a low contrast between grain and grain. Moreover, such grains can have a complex internal structure which cannot be visualized with SEM imaging. Figure 2 (b&c) demonstrate examples of panchromatic 4096 × 4096 (17 megapixels) PMT CL scans of equivalent area for differing pixel dwell times. These measurements necessitate the electron beam to be raster scanned over the sample, and for each position the CL intensity is calculated by the PMT.

Clearly, the CL imaging provides a more divergent contrast in comparison to the SEM image. The fundamental physical reasons for this contrast will be described in more detail below. The pixel density in these CL images is adequately high to reconcile fine spatial details in the CL structure. The signal levels are also sufficiently high to permit pixel dwell times of 800 ns. At 800 ns dwell time a 17 megapixel image can be captured in 13 s but a 512 × 512 image could be captured in ~0.2 seconds, allowing live CL imaging at an admissible resolution.

In a number of sample areas, the use of such rapid dwell times engenders streaking in the image. This is a consequence of the fact that the radiative lifetime of certain excited transitions surpasses the scan time per pixel, causing residual CL emission after the beam has moved.

Some rocks, such as carbonates, can possess rich extrinsic color centers, with exceedingly long radiative lifetimes, restricting the scan speeds to ms timescales [2]. Fortunately, in these sandstones such slow color centers are infrequent, so overall this fast scanning mode operates well.

By utilizing a 10 μs integration this streaking effect is totally eradicated, and an enhanced signal-to-noise ratio is also secured, generating smoother images. Only the electron beam is scanned in these single scan images. However, in theory the SEM stage can also be scanned, allowing tiling and stitching for a completely automated sample inspection workflow.

The effect of acceleration voltage

Next, the effect of acceleration voltage on the CL imaging is examined. Figure 3(a&b) exhibit images from the same area gathered at 15 and 5 kV respectively. At 15 kV the CL intensity is stronger due to the more energetic electron beam. This can be advantageous for rapid scanning, as previously described.

However, in concordance with EDS and WDS, this increase in signal typically transpires to the detriment of spatial resolution, as the penetration into the material and the consequent interaction volume is also larger (electrons can penetrate more than 2 μm of material compared to 300 nm for 5 kV, as approximated using the Casino Monte Carlo program supposing a density of SiO2) [7].

Therefore, the images at 5 kV have more noise but show more precise spatial properties, which could be of value for high-resolution studies. This is plainly evidenced by the images in Figures 3(c&d). A further example of this effect is shown in Figure 5(c&d). In sedimentary rocks (including mudstone and shale) the feature sizes are so minute that a lowered voltage may be required to record all the crucial spatial properties, but in sandstone most of the properties are of sufficient size to be reconciled with increased voltages.

It was noted that the electron interaction volume does not merely govern the spatial resolution and CL output. The translucency and texture of the material are also influential, as (wavelength-dependent) absorption and scattering in the material, in addition to reflections on material interfaces, can impede light outcoupling and detection [8].

Moreover, the depth-dependent structure of the actual material may become a factor (e.g. the material is stratified or zoned in the vertical direction). Quartz is ordinarily reasonably transparent, so there is a strong possibility of achieving light transport through microns of materials, without major propagation losses.

Panchromatic 1 megapixel PMT images taken at (a) 15 and (b) 5 kV on the same area using a 10 μs dwell time. The images in (c) and (d) show a blow up of the area enclosed by the white box in (b) for 15 and 5 kV respectively, revealing the difference in spatial resolution. Measurements taken on sample 2.

Figure 3. Panchromatic 1 megapixel PMT images taken at (a) 15 and (b) 5 kV on the same area using a 10 μs dwell time. The images in (c) and (d) show a blow up of the area enclosed by the white box in (b) for 15 and 5 kV respectively, revealing the difference in spatial resolution. Measurements taken on sample 2.

Cathodoluminescence contrast in sandstone

The perceived CL contrasts are now examined in greater detail. Figures 4(a,b) again reveal an SEM image and panchromatic CL image gathered from the same sample area. Despite the CL image being panchromatic, it nonetheless offers a notably higher contrast than the SEM image. For example, in the CL image it is demonstrably clear that the top right grain was fractured and the fissures were (partly) recovered by another material with a similar SEM contrast, but a weaker panchromatic CL signal.

This also applies to the bottom left grain, in which a section of the material on the border is dark in CL, and the adjoining surface of the material has an intermediate CL-intensity. To find out more about the cracked area, hyperspectral imaging has been undertaken, in which the electron beam is scanned and a comprehensive spectrum for each excitation position (80 ms dwell time) is assembled, thus generating a 3D datacube.

Figures 4(c,d) show the CL intensity distribution at 425 and 650 nm wavelength incorporated into a 10 nm spectral bandwidth. Clearly, the material in the fissure is dark at λ0 = 425 nm in relation to the neighboring material from the grain, but more luminous at 650 nm. These contrasts can be highlighted by producing a (false) color RGB image from the datacube.

(a) SEM image of an area on sample 1 (b) Panchromatic PMT image taken from the same area with a 500 μs dwell time per pixel (image resolution: 512 × 512). The color scaling is adjusted to the dynamic range of the image for proper visibility of all the features. Wavelength slice through hyperspectral datacube at (c) λ0 = 425 nm and (d) 650 nm showing a 2D CL intensity map for that wavelength (integration bandwidth is 10 nm). The scanned area corresponds to the white box in (b). (e) False color RGB image visualizing the CL contrast between the material in the crack and the surrounding grain. (f) SEM image corresponding to the CL maps in (c-e). (g) CL spectrum from the from the grain (1) and crack (2). We have included a spectrum from a synthetic 300 nm thick thermal SiO2 layer on silicon for reference. The spectra have been normalized to 1 for direct comparison of the spectral shape.

Figure 4. (a) SEM image of an area on sample 1 (b) Panchromatic PMT image taken from the same area with a 500 μs dwell time per pixel (image resolution: 512 × 512). The color scaling is adjusted to the dynamic range of the image for proper visibility of all the features. Wavelength slice through hyperspectral datacube at (c) λ0 = 425 nm and (d) 650 nm showing a 2D CL intensity map for that wavelength (integration bandwidth is 10 nm). The scanned area corresponds to the white box in (b). (e) False color RGB image visualizing the CL contrast between the material in the crack and the surrounding grain. (f) SEM image corresponding to the CL maps in (c-e). (g) CL spectrum from the from the grain (1) and crack (2). We have included a spectrum from a synthetic 300 nm thick thermal SiO2 layer on silicon for reference. The spectra have been normalized to 1 for direct comparison of the spectral shape.

The spectra are segregated in 3 sections, which are binned in such a way that the overall intensity in these spectral regions describes an RGB code. In this instance B = 400 – 483 nm, G = 484 – 567 nm, and R = 568 – 650 nm were chosen, which resembles genuine RGB color space quite adequately, and encompasses the pertinent spectral contributions in these quartz samples.

On the other hand, the CL spectrum corresponding to a particular position can be mapped. In Figure 4(g) the CL spectrum for excitation on the grain (position 1) and in the crack (position 2) are plotted. Both at position 1 and 2 there is a contribution at λ0 = 650 nm but at 1 a considerable peak at 425 nm is also evident.

The perceived contrasts are linked to the fact that the materials in the image have a divergent physical structure/phase as a consequence of their formation processes. This substantially impacts on the absolute CL emission intensity, in addition to the color of emission. Both the grain and the material in the crack is quartz so there is only low level elemental and density contrast. Consequently, the SE/EDS/BSE contrast in such instances is often low, meaning the extra CL contrast becomes very precious.

The observed spectral peaks are typical for quartz, exhibiting broad peaks at free-space wavelengths 425 and 650 nm [2,6,9]. They are ordinarily related to innate deficiencies in the SiO2 matrix [2,3]. The peak at λ0 = 425 nm is mainly located in plutonic and volcanic quartz which is produced at reasonably high temperatures. Contrastingly, the peak at λ0 = 650 nm coinciding with a non-bridging oxygen hole center (NBOHC) defect, can be found in most varieties of quartz [2,6]. The CL spectra suggest that the material at position 1 correlates with a detrital plutonic quartz grain with cracks that are filled with a low temperature authigenic SiO2 precipitate (position 2) [2,11].

This is consistent with the PMT image in (b) which offers an extended contextual summary of the local environment scrutinized in the spectral image. As well as quartz of geological origin, these spectral properties can also be located in synthetic quartz such as a thermal oxide on silicon (see gray curve in Figure 4(g)). This material is also generated under high temperatures, much like igneous rocks. However, the spectral peaks are narrower than in the geological specimens, which is likely linked to a lowering in homogeneous peak broadening as a result of the higher material quality and purity.

It has been demonstrated that extrinsic CL from titanium impurities in the quartz can also have a significant CL signature that engenders a robust CL feature at λ0 = 458 nm [9,10]. Aluminum is another regular impurity in quartz which can bring about CL emission bands at λ0 = 390 nm and λ0 = 500 nm [4,6].

Although it cannot be said for certain, it seems that these impurities have a minor effect on the CL emission in Delmic’s case, as has also been found in other instances [4]. Elemental x-ray mapping or mass spectrometry can be utilized to corroborate the incidence and even establish the concentration of such impurities.

(a) PMT image (5 kV) which is used to identify a ROI (indicated by the white box) on sample 2 . (b) SEM image of the ROI from (a). False color RGB images derived from the CL datacube collected on the area shown in (b) with (c) 5 kV and (d) 15 kV acceleration voltage. Spectra taken from the positions indicated in (d) for (e) 5 kV and (f) 15 kV. The spectra have been normalized to the maximum of the blue peak in spectrum 1 in (f).

Figure 5. (a) PMT image (5 kV) which is used to identify a ROI (indicated by the white box) on sample 2 . (b) SEM image of the ROI from (a). False color RGB images derived from the CL datacube collected on the area shown in (b) with (c) 5 kV and (d) 15 kV acceleration voltage. Spectra taken from the positions indicated in (d) for (e) 5 kV and (f) 15 kV. The spectra have been normalized to the maximum of the blue peak in spectrum 1 in (f).

Figure 5 describes a similar process to Figure 4, but in this example the acceleration voltage is also altered. In 5(a) the PMT image from which an area of interest (ROI) is chosen for hyperspectral imaging is demonstrated. Figure 5 (c,d) reveals the RGB CL images for 5 and 15 kV acceleration voltage.

The CL measurements again uncover areas with substantial ‘blue’ and ‘red’ emission. It is also indicated that the colors are not completely equivalent for the two voltages, and that the image for 15 kV is more nebulous and uniform; arising from the difference in excitation volume.

In (e) and (f) spectra is shown for the two distinctly colored regions as designated in (d), for 5 and 15 kV respectively. In this instance all spectra has been normalized to the maximum of spectrum 1 in (f) (blue curve) as to permit a more quantitative differentiation. Observing the intensities, the overall CL yield at 5 kV is more or less five times lower when compared 15 kV. In accordance with the color coding in (c,d) it is discovered that the blue peak is dominant at position 1 whereas the red peak is dominant at position 2.

Moreover, the peak ratios are different for 5 and 15 kV. This could be a by-product of the difference in how the material transitions are generated by the electron beam. The excitation density, coupled with the energy distribution of electrons in the interaction volume, is non-identical for these beam parameters, perhaps impacting on peak amplitudes. In addition, depth-dependent light outcoupling/absorption effects and depth-dependent alterations in the material structure/composition could become influential.

(a) Panchromatic PMT image taken on a single grain on sample 1. Wavelength-filtered (40 nm bandpass) PMT images at (b) λ0 = 450 nm and (c) 650 nm taken with a 500 μs dwell time. Wavelength slice through hyperspectral datacube at (d) λ0 = 425 nm and (e) 650 nm. The scanned area corresponds to the white box in (a). (f) RGB image visualizing the CL contrast between the microcracks and the surrounding grain. (g) SEM image corresponding to the white box in (a). (h) CL spectrum from the grain (1) and from the microcrack (2).

Figure 6. (a) Panchromatic PMT image taken on a single grain on sample 1. Wavelength-filtered (40 nm bandpass) PMT images at (b) λ0 = 450 nm and (c) 650 nm taken with a 500 μs dwell time. Wavelength slice through hyperspectral datacube at (d) λ0 = 425 nm and (e) 650 nm. The scanned area corresponds to the white box in (a). (f) RGB image visualizing the CL contrast between the microcracks and the surrounding grain. (g) SEM image corresponding to the white box in (a). (h) CL spectrum from the grain (1) and from the microcrack (2).

A meticulous experimental study in which acceleration voltage and beam current are exhaustively varied might yield greater insight, but is beyond this paper's remit. In the measurements in Figures 4 and 5, the cracks were somewhat large (~5 μm). The grain examined in Figure 6 reveals smaller microcracks and spiders which are regular in plutonic quartz grains [2,9]. Like the larger fissures, these also show up as dark in the panchromatic CL.

By utilizing color filtered PMT imaging it becomes obvious that these cracks are also infilled with authigenic quartz. This is reinforced by the hyperspectral experiment which exhibits very similar behavior to what was ascertained in Figures 4 and 5. In some instances the perceived behavior is different from Figures 4-6.

 (a) Panchromatic PMT image taken on sample 1. (b) Wavelength-filtered (40 nm bandpass) PMT image at λ0 = 650 nm. Wavelength slice through hyperspectral datacube at (d) λ0 = 425 nm and (e) 650 nm. The scanned area corresponds to the white box in (a). (e) RGB image visualizing the CL contrast between the material in the crack and the surrounding grain. (f) SEM image corresponding to the white box in (b). (g) CL spectrum from the bright surrounding matrix (1) and from the darker grain (2).

Figure 7. (a) Panchromatic PMT image taken on sample 1. (b) Wavelength-filtered (40 nm bandpass) PMT image at λ0 = 650 nm. Wavelength slice through hyperspectral datacube at (d) λ0 = 425 nm and (e) 650 nm. The scanned area corresponds to the white box in (a). (e) RGB image visualizing the CL contrast between the material in the crack and the surrounding grain. (f) SEM image corresponding to the white box in (b). (g) CL spectrum from the bright surrounding matrix (1) and from the darker grain (2).

In Figure 7 the robust blue/UV contribution is missing but there is nonetheless a contrast in intensity between smaller crystals and the neighboring network. Spectral examination demonstrates that the color of emission is alike, and that only the intensity differs. It is understood that glassy volcanic matrix quartz exhibits red CL emission because it has a higher number of defects than plutonic quartz [2] but the two quartz types examined here could also be divergent forms of precipitated SiO2, similar to what was seen in the healed cracks.

Spectrally filtered PMT imaging

In Figures 6 and 7 band pass filters are already made use of in front of the PMT detector to designate specific spectral characteristics from the sample. Figure 8 demonstrates another sample of images captured at the same area with band pass filters centered at λ0 = 450 nm (b) and λ0 = 600 nm (c) (filters have a 40 nm bandwidth).

Single-wavelength PMT measurements taken with a (a) 450 nm and a (b) 600 nm band pass filter with a 40 nm bandwidth. (c) SEM image corresponding to the region inspected with the PMT. (d) Red-blue color image in which the PMT images in (a) and (b) are placed in the blue and red color channel, respectively. For these measurements we used a 40 μs dwell time. Measurements taken on sample 2.

Figure 8. Single-wavelength PMT measurements taken with a (a) 450 nm and a (b) 600 nm band pass filter with a 40 nm bandwidth. (c) SEM image corresponding to the region inspected with the PMT. (d) Red-blue color image in which the PMT images in (a) and (b) are placed in the blue and red color channel, respectively. For these measurements we used a 40 μs dwell time. Measurements taken on sample 2.

Evidently, the images differ greatly; sections that are dark for 450 nm are bright for 600 nm and vice versa, similar to what was observed in Figures 4-6, on a more magnified scale. Such color-filtered images can be assimilated into a multicolor image as is demonstrated in Figure 8(d) (in this case the images in (a) and (b) are plotted onto the blue and red RGB color channels, respectively).

Such false coloring can be utilized for any synthesis of color filters, and assists in visualizing CL properties. This technique is broadly applicable and can be employed to pinpoint particular spectrally well-defined intrinsic and extrinsic color centers. In the context of these sandstones this technique would be best placed to undertake rapid segmentation of substantial sections into cemented and plutonic detrital components, for example.

Conclusions and Outlook

To summarize, the value of CL imaging in the context of quartz sandstones has been shown. Complementary imaging modalities in the CL network can be integrated to constitute an effectual process for in-depth research. The (spectrally-filtered) PMT imaging permits fast inspection for ROI determination and large area imaging for representing the total spatial structures and textures of the rock at the micro/nanoscale.

The more comprehensive hyperspectral research with the spectrograph can be implemented to uncover the structure and material characteristics of these rocks at a more profound, quantitative level.

The results as they are presented plainly show that cemented and granular detrital quartz can be effectively located and segmented in CL imaging. Quantitative techniques including point counting, or even full-scale CL imaging, can be utilized to approximate the volumetric ratio between intergranular porosity, overgrowths, and detrital quartz.

This ratio can afterward be implemented to appraise the initial pore volume, which can regulate the extent of compaction in sandstones, for example [2,12,13]. Scrutinizing porosity, cementation, compaction, and (micro)crack generation with CL is pertinent for petroleum geology, as these yield quantitative indicators on the permeability and storage capacity of reservoir rocks.

By incorporating CL imaging with other analytical SEM-based methodologies, such as BSE, EBSD, WDS, and EDS totally comprehensive automated sedimentary rock studies can be undertaken. Moreover, the high-resolution 2D data that is extracted in CL can in theory be integrated with focused ion-beam milling or with information acquired from μCT to gain awareness of the 3D rock structure. These prospects are very stimulating, and could conceivably bring about even more comprehensive and automated methods for studying sedimentary rock processes and properties.

References

[1] S. N. Ehrenberg and P. H. Nadeau, Sandstone vs. Carbonate petroleum reservoirs: A global perspective on porosity-depth and porosity-permeability relationships, AAPG Bull. 89, 435-445 (2005).

[2] S. Boggs Jr. and D. Krinsley, Application of Cathodoluminescence Imaging to the Study of Sedimentary Rocks, Cambridge University Press (2006).

[3] M. Pagel, V. Barbin, P. Blanc, and D. Ohnenstetter, Cathodoluminescence in Geosciences, Springer (2000).

[4] D. K. Richter, Th. Götte, J. Götze, and R. D. Neuser, Progress in application of cathodoluminescence (CL) in sedimentary petrology, Mineralogy and Petrology 79, 127-166 (2003).

[5] T. Coenen, S. V. den Hoedt, and A. Polman, The SPARC cathodoluminescence system: a platform for nanoscale optics and materials science, Micros. Today (2016).

[6] M. A. Stevens Kalceff and M. R. Phillips, Cathodoluminescence microcharacterization of the defect structure of quartz, Phys. Rev. B. 52, 3122-3134 (1995).

[7] D. Drouin, A. R. Couture, D. Joly, X. Tastet, V. Aimez and R. Gauvin, CASINO V2.42 - A Fast and Easy-to-use Modeling Tool for Scanning Electron Microscopy and Microanalysis Users, Scanning 29, 92-101 (2007).

[8] K. L. Milliken, SEM-Based Cathodoluminescence Imaging for Discriminating Quartz Types in Mudrocks, URTeC (2013).

[9] http://www.csiro.au/luminescence/default.aspx

[10] W. P. Leeman, C. M. MacRae, N. C. Wilson, A. Torpy, C. A. Lee, J. J. Student, J. B. Thomas, and E. P. Vicenzi, A Study of Cathodoluminescence and Trace Element Compositional Zoning in Natural Quartz from Volcanic Rocks: Mapping Titanium Content in Quartz, Microsc. Microanal. 18, 1322-1341 (2012).

[11] K. L. Milliken, The widespread occurrence of healed microfractures in siliciclastic rocks: Evidence from scanned cathodoluminescence imaging, American Rock Mechanics Association (1994).

[12] D. W. Houseknecht, Use of cathodoluminescence petrography for understanding compaction, quartz cementation, and porosity in sandstones. Luminescence Microscopy and Spectroscopy: Quantitative and Qualitative Applications, SEPM Short Course 25, 59–66 (1991).

[13] K. L. Milliken, R. M. Reed, and S. E. Laubach, Quantifying compaction and cementation in deformation bands in porous sandstones, AAPG Memoir 85, 237-249 (2005).

This information has been sourced, reviewed and adapted from materials provided by Delmic B.V.

For more information on this source, please visit Delmic B.V.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this article?

Leave your feedback
Submit