Ian’s methods

Collecting rock samples

When I’m out collecting samples, I bear in mind the fact that I’ll be aiming to make approx. 40 x 20 mm thin-sections mounted on 75 x 25 mm microscope slides and look for suitably sized pieces of the rock that interest me.
When I find what I’m looking for, I use a very useful app for the Apple iPhone and iPad called Outdoors GPS that provides zoomable 1:25k Ordinance Survey maps to mark the exact sample location with a flag and note, and then I label the sample using permanent marker.

A screenshot of Outdoors GPS on iPad

Screenshot of ‘Outdoors GPS’ on iPad.

Cutting a sample to size

Back in the workshop, I decide on the samples that I am going to thin and then I set about the process of shaping them into suitably sized 5mm thick rectangular blocks. This is a noisy, messy job that I prefer to do outdoors where the spray of the water that cools the diamond disc, and the dirty water that runs off the cutting table is easy to clean up. I use a Clarke 33mm stone/tile cutter that is fairly low powered but it has, to date, coped with all the rock that I’ve pushed in its direction.
The video illustrates the shaping process but what you don’t see are the essential protective glasses and the preferred ear plugs.
My first step is to cut out a block from the rough sample that is no larger than the maximum cut of the disc i.e. 33mm. In doing this, I’ll be creating one relatively flat face on the block.
The second step is run this flat face along a suitably sized wooden guide to thin the block to about 5mm.
Finally, I shape the block into a 40 x 25 mm rectangle using a combination of guided and freehand cuts.

Video: Shaping a sample of rock into a rectangular block

When the sample blocks are cut and dry, I paint a small patch of white acrylic paint onto each one on the side that won’t be glued to the microscope slide. Then I label each sample in waterproof ink with an identifier that refers back to its location marker on the O.S. map.

Levelling and smoothing the side to be glued

The next step is to level and smooth the surface of the sample that is going to be glued to the microscope slide. I use an Ameritool lap wheel that takes 8″ diamond discs. I bought it via eBay and I’m happy with it because it’s easy to use, clean, reliable and provides an adequate finish for this job. Being set up for the American market, it requires a transformer; I bought a Tacima 500VA that was a little more expensive than most of its competitors but it is neat, powerful and it doesn’t produce an irritating hum. It was also necessary to super-glue a couple of spacer washers to the finger-tightened nut that holds the disc in place on the lapping wheel because the cheaper model that I bought arrived set up for use with a thick platen to support silicon carbide discs.
The video below shows how I go about the process of levelling and smoothing one face of the sample. I tend to start with an 150-grade disc and when all the major scratches are out, move on through 350 and 600-grade discs, to finish the machine smoothing with a 1200-grade disc.

Video: Using a lap wheel to smooth the side to be glued

I move the block from side to side across the radius of the disc so that the abrasive coating wears evenly. I also rotate the block so that it is evenly smoothed as the trailing edge of the block tends to be abraded more quickly than the forward edge.
When I can see that that the ground surface is flat and free from scratches, I do a final bit of smoothing by hand using 600 grade silicon carbide paper laid on a 10x10cm square of float glass.

Gluing the sample to the microscope slide

I prepare the microscope slide by abrading on the lap wheel with a 1000-grade disc and cleaning and degreasing it thoroughly with Isoclene a cheap and handy isopropyl alcohol-based cleaning fluid. I also degrease the surface of the sample using the same fluid.
I’ve tried different glues and currently use Permabond UV675, a high strength, single part UV curable adhesive that is optically clear, and resistance to yellowing. The fact that it has a low viscosity and doesn’t require mixing means that it is far less fuss and results in fewer bubbles than the Araldite 2020 two-part epoxy glue I used to use. I’ve constructed a simple light box containing a Prolite 25 Watt blacklight energy saving CFL bulb. I slide in the mounted samples, cover the opening to prevent UV light spill, switch on and the Permabond is cured in seconds. The Araldite 2020 required carefully measured amounts of resin and hardener and took three days to cure completely.

Resources for gluing

Resources for gluing the sample to the microscope slide

I spread the glue with a cocktail stick

I spread the glue with a cocktail stick

Bubbles in glue

Allow any bubbles to surface and pop

Next, with the sample lying glue-face up, I lower a microscope slide onto it. I use one end of the sample as a fulcrum and carefully lower the slide until its own weight pushes out the air and makes full contact with the glue.

First stage in lowering the slide onto the glued sample

Using the edge of the block as a fulcrum

Second stage in lowering the slide onto the glued sample

Beginning to lower the slide onto the glued surface

Third stage in lowering the slide onto the glued sample

The slide falls under its own weight

Fourth stage in lowering the slide onto the glued surface

Making bubble free contact

Then I apply a couple of micro clamps that hold sample and glass together while I use a UV torch to create an initial bond before removing the clamps and placing the mounted sample into the UV lightbox for a minute or two.

A glued sample held in place with a couple of clamps

A sample held in place with a couple of clamps prior to initial curing with a UV torch

UV lightbox made in MDF

My UV lightbox made in MDF

Trimming the slide-mounted sample to 1mm thick

Once the sample is glued firmly to the microscope slide, it is ready for cutting so that just a a Imm thick slice remains on the slide. I have a Vitrex 17mm tile cutter just for this job that has a narrower disc than the Clarke machine. I’ve permanently fixed the position of the machine’s cutting guide and made the wooden jig shown below to position the mounted sample so that, if all goes well, 1mm of rock stays on the glass. It is by no means certain that all does go well and on occasions, I still have parts of the sample and even parts of the glass slide being cut away.

A jig for holding the mounted sample in the correct position for cutting

Rough and ready jig for holding the mounted sample in the correct position for cutting.

Video: Trimming the sample to 1mm thick

When the sample block has been cut, the narrow cutting wheel leaves enough left over to make a good hand specimen or, if the slide was damaged, to use as a backup for a second attempt. Of course, the left over piece has the label on, so I keep both it and the mounted sample together until such time as I label the slide.

Trimmed samples and remaining labelled pieces

Trimmed samples and remaining pieces with labels.

Thinning the section to 30 microns

My first step here is to use the lapping wheel to thin the section down to the point where the quartz (if there is any) is showing first order blue and red interference colours. From that point, I continue the thinning by hand.
If the rock is weak and the crystals are beginning to become cracked or holed, I stop machining immediately.
The video shows how I rotate the sample on the lapping wheel to even out the tendency for the trailing edge to thin more quickly than the leading edge. I sometimes start with a 150 or 350-grade diamond disc but most of the work is done using a 500 or 600 grade disc.
The video also shows me using a series of progressively finer wet and dry papers, but currently I am using a 1200-grade diamond disc mounted on plywood followed by a 3000 grade disc that is mounted next to it. I get better results when I use a thin oil when hand thinning rather than water.
When all or most of the quartz in the sample has lost its first order yellow interference colour (if there is no quartz, I look for the feldspar at the edge of the sample beginning to change from white to grey) the thinning is complete and I lightly polish it using 5000 grade wet and dry paper and clean it with Isoclene.I don’t apply a cover slip.
It is time consuming work but I find it satisfying when I manage to make a good thin section and interesting to see the crystal boundaries, colours and textures becoming clearer in the process of thinning.

Video: Machine and hand sanding to 30 microns

Showing the lack of definition and high interference colours when the section is too thick

The section being thinned is basalt.
The lack of definition and very high interference colours of the pyroxenes indicateit is still too thick.

Showing the higher definition and correct interference colours at 30 microns

The same basalt sample now thinned to about 0.30mm.
Without the presence of quartz it is difficult to be sure, but the yellow-orange orthopyroxenes and the second order blue inthe clinopyroxenes suggest it is about right.

Labelling slides

My labelling system identifies each thin section and relates it to the location where I collected the sample and the photomicrographs that I’ve made. I use PowerPoint to design and print the paper labels and PVA to stick them on the slides.

The labelled slide

The labelled slide without a cover slip.

Microscopy and Photomicrography

I use two microscopes, both of them by Brunel Microscopes; one is their monocular SP75, the other is their trinocular XP1500. Of the two, the SP75 produces slightly sharper photomicrographs while the XP1500’s two oculars make the direct viewing of thin sections a much more enjoyable experience. The XP1500 has a much better build quality and it also allows for reflected light viewing of samples without cover-slips.
The camera I use is a Nikon D5100 that I couple to the microscopes using adapters bought from Brunel Microscopes along with a x2.5 photo ocular.

The SP75 monocular microscope

SP75 monocular microscope with camera attached

The xp1500 trinocular microscope

XP1500 microscope

Whole slide photography in polarised light

In order to take whole-sample PPL and XP images that help in the semi-quantitative analysis of the samples, I attach a 49mm macro lens to my camera fitted with a linear polarising filter. I’ve made a rough and ready microscope slide-holder that fits in the cover of my iPad and positions the slide at a distance from the screen that, like most tablet screens, emits polarised light.
On the iPad, I open a pre-prepared, saved photo that displays a slide-sized white rectangle on a black background. This illuminates the whole sample in a relatively even polarised light. By rotating the linear polarising filter on the camera lens I can take pictures of the sample in both plane polarised light and with crossed polars.
I set a shallow depth of field on the lens and position the slide at a distance from the iPad screen so that screen pixels don’t appear in the image.

Taking photos of the whole sample in polarised light

Set up for photographing whole samples in polarised light

Whole section photographs in PPL and XP

Whole section photographs in PPL and XP

Determining Grain Sizes

Adobe Photoshop has some handy tools that help determine the grain size and mineral content proportions of a sample.
The ‘Analysis’ menu in Photoshop allows the custom calibration of a ruler tool by using the tool to stretch a measuring line between two points in the image ( e.g. from side to side) and then inputing the actual distance between the same two points on the physical thin section (e.g. 37mm). The software converts the number of pixels between the two points into a ‘logical’ measurement in millimetres. After calibration, the ruler can then be used to accurately measure any part of an image in any direction regardless of the zoom level of the image.
To calculate the average grain-size, I’ve made a Photoshop layer of 100 dots that I can overlay on any whole-section photograph. I measure each grain under every dot and use a Microsoft Excel table to input measurements and calculate averages.
For some crystals, particularly twinned ones, l have to refer to the actual slide viewed under the microscope to be sure of the crystal boundaries.

Using the ruler tool in Adobe Photoshop to accurately measure grain size

Using the ruler tool in Adobe Photoshop to accurately measure grain size.
In this example the section was 37mm from side to side, Photoshop displays the length of the stretched measuring line at the top of the screen, in this case the feldspar phenocryst is 5.29mm in length.

Determining the relative proportions of minerals in a sample

The approximate percentage by area of mafic minerals present in a thin section can be fairly quickly calculated in Photoshop by converting the whole-section PPL image first into greyscale and then, using the ‘Posterise’ tool, into a two, three or four tone image. The exact number of tones is a matter of judgement as to the best representation of the boundaries of the mafic minerals identified in the section under the microscope. In a three or four tone image, the opaque minerals appear completely black and the biotite and pyroxenes, dark grey.
In Photoshop’s ‘Photography’ mode, I clear the cache for the pixel count by clicking on the warning triangle displayed in the top corner of the histogram chart, select the area occupied by the whole section, refer to the information and make a note of the number of pixels it contains.
Then, using the Select -‘Colour Range’ option, I select the tonal range occupied by the mafic minerals and Photoshop displays the combined number of pixels in all the shadow/mafic areas, and so calculate of the percentage by area.

A thin section in PPL and the same section posterised to two tones to show the mafic content

Below: A thin section in PPL
Above: Posterised show the mafic content.


Photoshop’s colour range selection and pixel count.
The colour range is set forshadows and the pixel count is showing the total for the selected whole sample.


The pixel count shows the number of pixels occupied by the selected ‘shadow’ range.

Approximate proportions of the various mafic and felsic minerals can be determined in a slightly less straightforward way.
I carefully select all the areas in the image that are occupied by crystals of a particular mineral and then fill these areas in with flat colour. Then the Select – Colour tool enables me to see the number of pixels occupied by that colour so that I can calculate the percentage of the total area of the section that is occupied by the mineral.
Distinguishing between altered plagioclase and alkali feldspars can be very difficult and is an obvious source of error.

The same sample with plagioclase highlighted

The same sample with the plagioclase selected and highlighted in blue for pixel counting.

The same sample with the quartz highlighted

Here, the quartz is selected and highlighted in red.

The whole section PPL image overlaid with the quartz and plagioclase selections

Here the plagioclase select and quartz select layers overlay the whole-section view.

No vestige of a beginning, – no prospect of an end