Sampling and Pretreatment of Tooth Enamel Carbonate for Stable Carbon and Oxygen Isotope Analysis

The following protocol follows the guidelines of the Light Isotope Mass Spectrometry Laboratory at the Max Planck Institute for the Science of Human History. Appropriate ethics permissions from national and international committees should be sought for analyses involving endangered modern or historical faunal specimens, and for the use of archaeological and faunal material of interest to contemporary stakeholders. In this paper, the samples used were archaeological and fossil specimens. No living humans were used in this study and full ethical, institutional, and governmental permissions have been gained for any destructive analysis.

1. Bulk Sampling

Note: For humans and animals, the basic method of bulk tooth enamel sampling is the application of a drill to the buccal edge of the tooth.

1. Clean the outer surface of the tooth (~0.1 mm) using an aluminum shot blaster or by gentle abrasion using the drill set-up. Use a hand-held drill with a clean diamond-tipped drill bit (see below), of even shape, attached via a chuck to the drill set-up. Ensure that gentle and even abrasion is performed as a groove parallel to the entire growth axis (Figure 1 and Figure 2).
2. Attempt, where possible, to apply the method to the same tooth portion for each human or animal sampled.
   Note: In human paleodietary research, third molars are generally favored as representative of late juvenile/adult diet while first molars are avoided due to weaning effects²³. The preferred tooth for each taxon, and the period of life represented, will vary with the available samples and question under study.
3. Drill in a well-ventilated room while wearing a mask to avoid the inhalation of tooth enamel powder. Wear goggles to protect eyes. Be aware that the amount of sample powder drilled will vary depending on the equipment used for analysis, pretreatment protocol, and size of the tooth. Drill at low speed in order to prevent heating the sample.
   Note: Approximately, 4-8 mg of sample powder is an appropriate goal for multiple analyses using an online gas preparation and introduction system for isotope ratio mass spectrometry and diagenetic testing. Be aware that the chosen drill speed will depend on the fragility of the sample and will require some level of trial and error. In general, on a hand-held drill, two out of five bars of strength are sufficient. Drill the tooth powder over a piece of clean aluminum foil or weighing paper.
4. Collect the resulting enamel powder and transfer it into a 1.5 mL micro-centrifuge tube that has been tared on an appropriately accurate balance prior to drilling. Label each tube with the sample designation (Figure 3). Record the sample numbers and enamel weight in a laboratory notebook as well as an electronic database.
5. Before drilling each new tooth, clean the utilized drill bits first using 0.5 M HCl on a robust tissue. Then wash the drill bit with an organic solvent (such as acetone, methanol, or ethanol) using a robust tissue.
6. Following a drilling sitting, clean the workspace thoroughly using a dedicated vacuum cleaner, or a dustpan and brush. Wipe the sampling space with methanol. Vacuum or lightly spray the drill with compressed air to remove dust and sample powder between samples.

2. Sequential Sampling

Note: Sequential sampling can be approached in a variety of forms and will depend on the taxon being sampled, the size of the tooth, as well as the desired temporal resolution.

1. Clean the surface of the tooth being sampled by removing a very thin (~0.1 mm) layer of outer tooth enamel.
2. Drill the samples along the buccal surface perpendicular to the growth axis of the tooth, running from the enamel root junction to the top of the crown. Drill horizontal bands perpendicular to the growth axis, with each sample resulting in a 1-2 mm wide groove across the enamel layer (Figure 1 and Figure 4). Take care to avoid drilling through the enamel into the dentine as it would contaminate the sample and the resulting measurement.
   Note: The number of samples taken will vary depending on the resolution desired but c. 10 represents an appropriate goal to study seasonal changes in δ¹³C and δ¹⁸O values for a hypsodont the size of domesticated cattle.
3. Choose the appropriate drill bit for each sample, as the width of the sampling line is determined by the diameter of the drill bit.
4. Use a rig set-up for more delicate teeth requiring a greater number of incremental samples (Figure 2) (e.g., human teeth), as well as for more stable drilling. Set up the rig stand that holds the drill securely, with the drill bit pointing downwards.
   Note: While four samples can easily be obtained from a permanent human molar using a hand-held approach²⁴, more refined sampling requires a rig or CO₂ laser ablation approach that will not be discussed here¹⁴. Another common approach is to use a semi-automated microsampling²⁵.
5. Hold the tooth or attach to a clamp. Press the tooth enamel against the drill bit and apply pressure. Samples are drilled incrementally along the buccal surface from the apex (occlusal surface) to the base of the crown at the location of the enamel-root junction (ERJ). Replicate the same sample drilling strategy multiple times drilling each consecutive line parallel to the previous sample line. Use a small cylindrical diamond drill bit for delicate teeth, such as sheep molars, with a diameter of 1 mm (Figure 1C).
6. Measure the distance of each sampling line from the ERJ using calipers and record this distance for comparison.
7. Follow steps 1.3-1.6 above.
   Note: If small, delicate teeth are being analyzed and pretreatment and diagenetic check methods are not applied, samples as small as 1-4 mg will produce reliable results on an on-line gas preparation and introduction set-up for carbonate analysis. Modified automated-peripheral systems can facilitate the use of even lower sample weights, but are inevitably limited by the proportion of carbonate in the studied enamel (4 to 5 wt%)^26,27. Researchers should consult with the laboratory where samples are going to be measured to determine the amount of sample required.

3. Fourier Transform Infrared Spectrometry/Attenuated Total Reflectance method

1. For the Attenuated Total Reflectance method, establish a sample chamber background, either under vacuum or in normal atmospheric conditions.
2. Place approximately 1 mg of tooth enamel over the diamond crystal in the sample chamber using a spatula. Lower and secure the sample post until there is a firm connection between the post, the sample, and the diamond plate. Close the sample chamber, with or without a vacuum being established, depending on the set-up or availability.
3. Measure the FTIR spectra of the sample 64 times for the wavenumber ranging between 400 and 4,000 cm^-1. The desired number of replicates will vary based on the goals of the study, though the analysis of three replicates, ideally with different aliquots being taken and then returned to the microcentrifuge tube housing the sample, will ensure robust results.
4. Clean the spatula with methanol between each aliquot and between each sample.
5. Perform a baseline correction using the available software. The software automatically subtracts the sample chamber background from the resulting FTIR profile. To ensure better reproducibility, only enamel spectra with a minimum absorbance of 0.06 for the highest phosphate band at ~1035 cm^-1 should be taken into account.
6. Monitor the presence of the common secondary contaminant carbonate calcite in all samples by checking for a peak at 711 cm^-1.
7. Following the analysis, carefully return the samples to their microcentrifuge tubes using a spatula or suction device and take onto the next stage of pretreatment or analysis.
8. Export the raw wavelength and intensity data as a .csv file (or similar) and use to calculate the indices of interest (commonly used indices include A-site Phosphate Index, B-site Phosphate Index, Phosphate Crystallinity Index, Water-Amide on Phosphate Index, and CO₃/PO₄ index^17,18,19) (Table 1).
9. Compare the FTIR results from the fossil or archaeological samples to growing datasets of modern faunal tooth enamel now available in the literature^19,20 to determine the nature and extent of diagenetic alteration to the enamel sampled.

4. Simple Acetic Acid (0.1 M) Pretreatment

1. Weigh the enamel powder for each sample in a microcentrifuge tube using a balance as appropriate. Label the tube accordingly. Enamel powder must be ground to the same degree, with particles of a similar size.
   Note: Many researchers utilize an oxidizing agent, such as dilute bleach (NaClO) or 30% hydrogen peroxide (H₂O₂) to remove organics from a sample and this should be added at this point. However, not only may these reagents alter the stable carbon and oxygen isotope values of a sample, but spectra of bone collagen samples in an on-line gas preparation and introduction system demonstrate that the 100% phosphoric acid utilized in carbonate measurement does not react with organic samples (Figure 5), suggesting that this step is unnecessary.
2. Add 0.1 mL of 0.1 M acetic acid (per 1 mg of enamel) to each sample using a pipette. Avoid touching the samples with the pipette. If the sample and pipette come into contact, use a new pipette for the next sample to avoid cross contamination.
3. Agitate, either through shaking or an electronic agitator (Figure 3) and then leave the samples to sit for 10 min. Enamel powder should not be left in acetic acid for an extended period of time (>4 h).
4. Place the samples in a microcentrifuge for 2 min at 13,700 x g.
   Note: A low-cost alternative is the use of a mini-centrifuge for 4 min at 3,500 x g, which holds fewer samples but is less time consuming when dealing with large sample sets.
5. When the microcentrifuge program stops, replace the acetic acid with 2 mL of ultrapure water using a clean pipette, then microcentrifuge the samples for 2 min at 13,700 x g.
6. Change the ultrapure water two more times, as before, and test for neutrality. Remove the remaining fluid using a pipette (without disturbing the supernatant).
   Note: A total of three washes with ultrapure water is the standard procedure to get to neutrality. The sample should be washed until neutrality is reached.
7. Cut sheets of Parafilm (1 cm × 1 cm) and place over each microcentrifuge tube. Make a small hole in the center using a sharp object so that the sample dries appropriately.
8. Place the tubes in a freeze drier to remove any remaining fluid.
   Note: If a freeze drier is not available, gentle oven drying (40 °C) is also a possibility and should not have any adverse effects on the stable carbon and oxygen isotope data.
9. Once dry, remove the Parafilm and close the microcentrifuge lids. Make sure that the tube labeling is correct and re-write if necessary.
   Note: This is the final stage prior to storage for weighing out, which in labs with limited funds may be the final stage prior to the distribution of the samples for analysis elsewhere.

5. Weighing and Measuring Samples and Standards

1. Using a spatula, weigh out approximately 2 mg of enamel powder onto a tin disc on a balance sensitive to 0.001 mg (Figure 6). Then carefully transfer the sample into a phosphoric acid-resistant borosilicate glass vial. This amount of enamel is required to yield reliable results due to the relatively small proportion of carbonate in tooth enamel.
2. Clean the spatula between samples using methanol and use a fresh, clean disc or weighing vessel for each sample.
3. Weigh out 0.2 mg of an international standard such as IAEA NBS18, IAEA 603, IAEA CO8, or Merck CaCO₃. An in-house standard made from the homogenized enamel of a large tooth can be used as well²⁰. For a full run of 76 samples and standards, use 61 samples and 15 standards, which should be evenly distributed over the whole run. Several studies have reported appropriate numbers of standards and details in archaeological research^20,28.
   Note: The repeated pretreatment and analysis of an enamel in house standard, across a number of runs of archaeological samples, is also likely to provide an appropriate measure of accuracy for the samples being studied²⁰.
4. Tighten the specially-designed lids, accessible to the needle until the septum is tight but not sucked into the vial. Do not over tighten the lid as the pressure on the septum will get too strong and can ultimately result in a broken needle during analysis. Place the vials in a listed order within a heating block at 70 °C. The order and weights of samples must be reliably recorded for entry into the computer software.
5. Measure the samples by following three steps: 1) flush fill the samples with pure Helium, 2) add phosphoric acid, 3) measure the sample.
   1. Flush the samples with pure helium (grade 5.0) with a flow of 100 mL/min for 5 min to flush out ambient air.
   2. Add 4-5 drops of phosphoric acid (99%, density 1.85 g/mL) to the carbonate-containing sample. Reaction between the sample and H₃PO₄ begins. During the reaction, CO₂ is released which carries the isotopic value of the carbonate ion CO₃^2- of the sample.
   3. Wait 1 h per sample until the equilibration of CO₂ is reached. After 1 h, the samples should be measured to obtain reliable results. During the acquisition, a mixture of sample gas and He passes to the device. A drying stage removes water from the sample gas mixture.
   4. Measure the sample by starting with three peaks of CO₂ reference gas with a known isotopic composition. Peak intensity should be 4000 mV to match the intensity of the sample peaks (between 4000 to 8000 mV). The three reference gas peaks should be 20 s long and 30 s apart. Follow with defining the measurement interval of the sample. Measure the sample 10 times for 5 s each and 55 s apart. Ensure that peak height decreases over time indicating proper transport of the sample/helium mixture (Figure 5).
      Note: During measurement, the software calculates the isotopic composition of the sample by comparing the known isotopic value, the peak height, and area of the peak of the reference gas with the peak height and area of the peaks of the samples. The raw ¹³C/¹²C and ¹⁸O/¹⁶O composition of the sample is calculated.
6. Normalize the samples after measurement.
   Note: This is important because the reference gas does not undergo the same chemical and physical path as the sample gas when introduced into the mass spectrometer. Therefore, it is essential that the samples are additionally calibrated to international standards (step 5.3) that undergo the same physical and chemical treatment within a run as the sample itself²⁰. These international carbonate standards enable the two-point calibration of samples onto the international delta measurement scale and the assessment of precision and repeatability throughout a given run.