Zen and the Craft of Thin-Sectioning


On a brief visit to the northern hemisphere, and after a pleasant few days in Tübingen discussing the Lower Palaeolithic of Arabia (more later), the morning of the Day of Archaeology 2016 found me once more in the Mary Cudworth Lab at the Department of Archaeology, University of York, sanding away at a resin block attached to a small glass slide in the name of ‘Science’.

The slides I was preparing are used for thin-section micromorphology, that is, the examination of sediments or soils at the microscopic scale. By looking at the arrangement of the sediment particles, inclusions and any coatings or areas of dissolution of material, micromorphological analysis can be used to tell how a sediment deposited, and how it has been altered after it was deposited, information that can be used to reconstruct past environments and the means by which archaeological deposits were formed and preserved.

My PhD used this technique to examine at the changing sedimentation in the Haua Fteah, a Libyan cave site, during the Middle to Later Stone Age, and what these changing processes meant for environments during this transition. implications of these. The slides I was working on today are from an exposure of shallow marine deposits in SW Saudi Arabia, at Dhahaban Quarry, as part of the SURFACE project. These marine deposits were covered by windblown sediments. Micromorphological analysis of samples from the sediments will allow confirmation of these field interpretations, and, in establishing the contact between the marine and aeolian deposits, mark a past sea level, information that is incredibly important to understanding past coastal change, and human activity within it.

But before analysis, the thin sections have to be made, and it is this time-consuming, and at times frustrating, thin section-making process which is about as far away from Indian Jones as you can get…

Manufacture of the thin sections is technically straightforward:

  • Remove an intact block of the deposit to be studied, transport it to the lab and dry it out.
  •  Soak it in crystallitic polymer resin, and place it for a day or so in a vacuum, to make sure the resin fully impregnates the sample before it hardens.
  •  Wait a few months for the resin to harden fully.

Ready for the fun to start! A resin-impregnated sample ready for thin-sectioning. Photo: R. Inglis.



Cut resin blocks – time to choose the face to be mounted. Photo: R. Inglis.

  • Cut a slice off the resin block and trim it down to the size of a slide.
  • Grind one face of this slice so it is flat and even.

Grinding the blocks on the Logitech. Photo: R. Inglis.

  • Clean the face and glue it to a 2mm-thick glass slide using epoxy resin.

On the press – not too hard, not too light…Photo: R. Inglis.

  • Remove all but 1-2mm of the slice by cutting off the excess sediment/resin.

Cut the sample as thin as you can stomach to avoid long grinding times. Photo: R. Inglis.

  • Grind off all but 30µm (0.003mm – about the thickness of clingfilm) of the sample using a grinding machine, such as a Logitech as we have at York.

Diagnosis: human error. I set the jigs up wrong on the left-hand samples, causing loss of large areas of the samples. Time to start over! Photo: R. Inglis.

  • Check for thickness using a microscope (quartz grains under cross-polarized light should look grey or white).
  • Polish the surface of the slide, and there you have one lovely thin section, ready for analysis and interpretation!

Not perfect, but a finished slide ready for analysis. Photo: R. Inglis.

The whole process should, after the resin has hardened, take only a few days to complete. It varies a little based on the equipment used – at Cambridge, the Brots we used produced ‘mammoth’ thin sections (13x7cm, or 7x5cm – see Julie Boreham’s excellent Facebook site ‘Hidden Worlds: Off the Bench’ –  whereas the Logitechs produce 11x7cm slides and grind the samples in a slightly different way, using water rather than oil as lubricant. It’s therefore been a learning curve adapting my skills to the Logitech equipment after spending my PhD on the Brots.

Thin section manufacture is more a craft than a rigid science. At every step, these slides can be tricky: the chemical composition of the sediment or what it’s packed in may react with the resin, preventing it from hardening; too thick resin may not be drawn to the centre of the sample, meaning you have to impregnate it again. And that is just the resin. Cutting the blocks to the right size, and mounting them on the presses under a pressure firm enough to press the sample to the glass, but not so firm it cracks the glass, is the next hurdle. Further challenges include: cutting enough of the excess off so the grinding doesn’t take a whole day to remove the rest, but avoiding tearing the entire sample off the slide; making sure you’ve set the machine up properly (I am very much still learning about the Logitech’s finer points!); and whilst grinding, making sure the sample remains fully attached to the slide to avoid losing your sample at the last stages. And of course, the machines have their own foibles…

It’s this latter challenge which has me this morning gently sanding away microns of sample to get to that magic 30µm thickness – after it seized up on my last visit, our resident Logitech whisperer, Dr Carol Lang has been hard at work fixing it alongside her own research (on East African agriculture and Scottish hillforts), and we are on the road to perfect function once more. Yet whilst we are on that road, the monitoring continues and my slides come off a little thick, just in case, hence the sanding or ‘hand-finishing’. But slowly and surely, the beautiful structures and the sought-after grey/white quartz grains begin to appear, another thin section will be in the box, and I’ll be one step closer to unravelling the mysteries of the Dhahaban Quarry sediments.

And to the title of this blog. Sometimes thin sectioning goes perfectly, other times everything that can go wrong will, leaving you weeping over an amusing (to everyone else) set of six slides that have each broken at different stages. But it is also satisfying to work through a problem sample, coaxing it along and working out ways to adapt the process to the awkward reality of each sediment whilst finding out multiple new uses around the lab for the ultimate multi-tool, the large screwdriver. And then, sometimes, you just have to accept that a sample is lost and start again – and that this is OK.