If you are interested in collaborative research projects/commercial analysis then please email b.stern@brad.ac.uk
On-site sampling of archaeological ceramics for the purposes of lipid analysis has developed through dialogue between field and laboratory-based archaeologists. However, the dialogue continues so we would be very interested to receive feedback on the applicability and utility of the guidelines suggested below.
Sampling for organic residue analysis really depends on the sample. If you have digital photos of the sherds then that would give us a better idea of what we are dealing with. There are three main types of samples we can take: visible residues, ceramic absorbed residues and the matrix (usually soil). Sampling can be carried out in the field/museum although it is generally easier to send the sherds here where we can sample under controlled conditions - as the levels of archaeological organic residues are generally low we need to take precautions to avoid contamination.
Handling: Lipid analysis is based on the survival and recognition of molecules common to many sources around us in the contemporary world. For example, fingerprints will deposit lipid molecules, such as fatty acids, cholesterol and wax esters onto the surfaces of samples. Therefore, if at all possible minimize or even exclude handling of samples intended for analysis. Use the end of a trowel or tweezers (neoprene gloves may be hard to find in the field).
Washing: Avoid washing sherds intended for organic analysis. Washing could dislodge surface residues, remove more labile constituents of absorbed residues and introduce contaminants. Avoid cleaning sherds with hydrochloric acid as this is likely to bring about the removal of organic components. During recording, please avoid pen and varnish on the inner surface of the sherd.
Soil samples: If possible, soil adhering to the sherd should be bagged with the archaeological sample. Alternatively, several grams of soil from around the sherd can be collected and bagged separately. Soils do have their own lipid fraction, often completely different to that absorbed in the sherd matrix, and it is therefore useful to look for potential migration of contamination into the sherds.
Destructive sampling: Lipid analysis is destructive and is based upon solvent extraction of powdered ceramic-absorbed residues or visible residues.However, most analysts will leave a portion of the sample for future analysis. For ceramic-absorbed residues some early investigations used up to 100 g of powdered potsherd! We have experimented with very small samples (0.1 g) of sherd powder from both the exterior and interior of the sherd - to achieve this we drill an area of approximately 1X1cm. We use a Dremmel drill fitted with a tungsten abrasive bit (solvent cleaned between samples) to remove a fine powder to a depth of 2mm. We collect the sherd powder on aluminium foil before transferring to a glass vial. If requested this can be done in the centre of the sherd if you require the edges for reconstruction. However, when starting out, we would advise a ceramic sample size of between 1-2 grams. For visible residues we generally require a very small amount of visible organic residue for analysis. If it is pure organic material then a pinhead sized piece is about the minimum we can analyse. In both cases if you can spare more then this gives us greater options for repeat analysis or isotopic analysis etc and we will return any unused material. When possible it would be extremely helpful to have more than one sample coming from the same vessel. If handles are available they can be utilised as control samples, especially in the cases that soil samples are not available.
Archaeological considerations: In order to maximise the potential for archaeological interpretation, it is extremely helpful to note the fabric type and to know what part of the vessel the sherds derives (e.g. rim, body or base). Equally important is for sherds to be sampled from stratigraphically well-established contexts. Some use regimes (e.g. boiling) are likely to lead to differential deposition of lipid within the wall of the pot. Ideally, whole vessel profiles will allow multiple sampling of a single vessel, although such samples are not easy to come by so consideration must be given to whether up to three 1-2 gram samples can be removed and powdered. Some sherds display visible surface deposits on their interior surface. This is often burnt organic matter and lipids do tend to be preserved in these carbonaceous matrices. Separate extraction of the visible surface material does allow for comparison with the lipid absorbed in the pot matrix. We have also experimented with so-called ‘concentration gradients’ – longitudinal sections through the potsherd from the inside to the outside. This is helpful where soil samples are not available. High concentrations of lipid in the interior sections compared to the outer sections support the view that the vessel came into contact with lipid-rich substances. Finally, sooting on the outer surfaces of potsherds/vessels is useful in suggesting activities associated with heating the containers and may assist in interpreting the lipid distribution.
Storage: Please avoid storage in plastic bags/wrapping as plasticizers can adhere to samples and tend to swamp the generally low levels of lipid. To send organic residues preferably sample into solvent washed glass vials, to send sherds wrap the sample in aluminium foil, and then in a labelled zip-loc bag.
Other issues: Some countries require permits to export samples for analysis. The requirements vary but catalogues referring to as much detail as possible regarding the sherds to undergo analysis with photographs are normally needed.