CONCLUSION Solid-state NMR spectroscopy provides information on the elimination of collagen and diagenetic carbonate as expected throughout the two carbonate purification protocols. These protocols affect differently carbon and oxygen stable isotope values of individuals from the catacomb of Sts Peter and Marcellinus and the St Benedict cemetery (inter- and intra-sites). For both sites we observed a general pattern for δ 13 C and δ 18 O values between 0.1 M acetic acid and 1 M acetic acid treatments corresponding to a depletion of δ 13 C values and an enrichment of δ 18 O values (except for one individual). Further analyses have to be conducted in order to confirm these results for different contexts (burial conditions, soil environments, diagenesis processes, etc.) and different periods. Finally, according to this study, using 1 M acetic acid treatment in archaeological context provides as valuable results as 0.1 M acetic acid treatment. Special thanks to Axelle Grélard, Claude Manigand and Joël Ughetto. V ALIDATION OF BONE APATITE PURIFICATION PROTOCOLS FOR S TABLE ISOTOPE ANALYSIS IN BIOARCHAEOLOGY BY SOLID-S TATE NUCLEAR MAGNETIC RESONANCE SPECTROSCOPY 1 Université de Bordeaux, CNRS, UMR 5199 PACEA Anthropologie des Populations Passées et Présentes, Talence, France 2 Muséum National d’Histoire Naturelle, CNRS, UMR 7209 Archéozoologie, Archéobotanique : sociétés, pratiques et environnements, Paris, France 3 Université de Bordeaux, Institut Polytechnique de Bordeaux, CNRS, UMR 5248 CBMN Chimie & Biologie des Membranes & Nanoobjets, Pessac, France Kevin SALESSE 1 , Vanessa URZEL 1,3 Élise DUFOUR 2 , Dominique CASTEX 1 , Jaroslav BRŮŽEK 1 and Érick J. DUFOURC 3 Corresponding authors: k.salesse@pacea.u-bordeaux1.fr, v.urzel@pacea.u-bordeaux1.fr -15 -10 -5 0 1 2 3 4 5 -15 -10 -5 0 1 2 3 4 5 -7 -5 -3 -1 1 1 2 3 4 5 INTRODUCTION Stable isotope analysis on mineral phase of bone is commonly used to investigate diet patterns and residential mobility of human past populations. The reliability of stable isotope values of the mineral phase of bone discovered in archaeological context is however regularly questioned due to potential isotopic effects caused by the chemical treatments used to eliminate exogenous and adsorbed carbonate in bone apatite. To date, two bone apatite purification protocols are frequently applied in bioarchaeological studies. Nuclear Magnetic Resonance (NMR) spectroscopy appears to be a fundamental tool for characterization of the mineral phase of archaeological bone throughout the different steps of bone apatite purification treatments. NMR spectroscopy can provide information about the composition, conformation and structure of many compounds of bone using the physical properties of atomic nuclei. Although its use has been extended to the comprehension of biological issues, NMR spectroscopy has been underused in domains such as physical anthropology and archaeology. OBJECTIVES This study aims to test the validity of two different bone carbonate purification protocols for carbon (δ 13 C) and oxygen (δ 18 O) stable isotope analysis by hydrogen ( 1 H) and carbon ( 13 C) solid-state NMR spectroscopy. MATERIAL This study was carried out on 10 archaeological human bones from mass graves found in different cemeteries and presenting several states of preservation: - 5 bones from the catacomb of Sts Peter and Marcellinus (SSPM) in Rome (Italy, 1 st -3 rd century AD) - 5 bones from the St Benedict cemetery (SB) in Prague (Czech republic, 18 th century AD) RESULTS & DISCUSSION THE CATACOMB OF STS PETER AND MARCELLINUS -7 -5 -3 -1 1 1 2 3 4 5 RESULTS & DISCUSSION THE S AINT BENEDICT CEMETERY Fig. 3: Superimposition of 13 C NMR spectra obtained for T0, T1, T2 and T3. Each curve is a summation of the 5 individual spectra from SSPM Fig. 5: δ 13 C values for each individual from SSPM according to T0, T1, T2 and T3 Fig. 6: δ 18 O values of each individual from SSPM according to T0, T1, T2 and T3 METHODS Protocols: Bones were sampled with a hand drill. Bone samples were mechanically cleaned to remove surface contaminations. Samples were frozen in liquid nitrogen and powdered with a MM400 cryomill. Powder was soaked in 2-3 % NaOCl solution for 48 h at room temperature to oxidize the organic matter and was subsequently rinsed five times with distilled water. The remaining powder was treated with 0.1 M CH 3 COOH for 4 h or with 1 M CH 3 COOH for 1 h at room temperature to eliminate exogenous and adsorbed carbonate and was then rinsed five times with distilled water. Stable isotope analysis was performed on 580–630 µg of untreated and treated bone powders via a Thermo Scientific Delta V Advantage Isotope Ratio Mass Spectrometer interfaced with a Kiel IV Carbonate Device. NMR analysis was performed on 100 mg of untreated and treated bone powders via a Bruker Avance 500 MHz spectrometer under Magic Angle Spinning (MAS) at 10 kHz. For 1 H spectrum acquisition, we used a classical sequence with 156 scans, a pulse length of 2 μs, and a relaxation delay of 5 s, representing a total acquisition of approximately 15 min. For 13 C spectrum acquisition, we used a Cross-Polarization under Magic Angle Spinning (CPMAS) pulse sequence with 18 k scans, a contact time of 3 ms, and a relaxation delay of 2 s, representing a total acquisition of approximately 12 h. δ 13 C and δ 18 O values as well as 1 H and 13 C NMR spectra were obtained for: T0 = untreated bone powder T1 = treated bone powder with 2-3 % NaOCl during 48 h T2 = treated bone powder with 2-3 % NaOCl during 48 h and 0.1 M CH 3 COOH during 4 h T3 = treated bone powder with 2-3 % NaOCl during 48 h and 1 M CH 3 COOH during 1 h Lipids OH of hydroxyapatite HCO 3 of hydroxyapatite + H 2 O Hyp γ Pro α, Hyp α Hyp δ Ala α, Pro δ Hyp β Gly Pro β + CH 2 Pro γ Ala β Lateral chains of amino acid in organic matter Fig. 1: 13 C NMR spectrum gives information about the organic fraction of bone (namely type-I collagen) Carboxyl groups of organic matter COOH CO COMPARISON T0 VS T1: 13 C NMR spectra: NaOCl removes the major part of collagen. 1 H NMR spectra: NaOCl, as show the similarity of spectra, did not affect the mineral fraction. COMPARISON T1 VS T2/T3: 1 H NMR spectra: OH signal intensity is preserved. HCO 3 signal intensity decreased for both T2 and T3 attesting elimination of CO 3 , presumably adsorbed and exogenous CO 3 . COMPARISON T2 VS T3: 1 H NMR spectra: HCO 3 signal intensity is lower for T3 than T2, suggesting the elimination of a slightly larger amount of CO 3 with T3. δ 13 C values: There is a systematic 13 C offset between δ 13 C values of T2 and δ 13 C values of T3. The mean spacing between δ 13 C values of T2 and T3 is 0.6 ± 0.3 ‰ δ 18 O values: Except for the individual 5, all studied individuals show a 18 O enrichment of δ 18 O values between T2 and T3. The mean spacing between δ 18 O values of T2 and T3 is 0.2 ± 0.2 ‰. EFFECTIVENESS OF NaOCl TREATMENT For each individual, on three sub-samples of the same bone, we performed NaOCl treatment in order to remove organic matter. Despite the use of identical analytical conditions (immersion time, NaOCl concentration, NaOCl volume, temperature), we observed that organic matter is sometimes completely removed of all the sub-samples of a same individual, whereas sometimes organic matter remains in only 1 or 2 sub-samples. When the bone collagen is particularly well-preserved, a special care about vortex mixing of aliquot (bone powder in NaOCl solution) has to be devoted to this step in order to avoid this kind of issues. δ 13 C δ 13 C δ 18 O δ 18 O COMPARISON T0 VS T1: 13 C NMR spectra: The collagen is removed by NaOCl. 1 H NMR spectra: Despite the decrease of the peak intensity in the HCO 3 region (5 ppm), that could be ascribed to a loss of water, spectra obtained for T0 and T1 can be considered as similar. COMPARISON T1 VS T2/T3: 1 H NMR spectra: A small decrease in the intensity of OH peak is observed. It could be due to a loss of OH in HA that should not occur with T2 and T3. Further analysis need to be conducted to confirm this hypothesis. CO 3 elimination (probably diagenetic CO 3 ) by T2 and T3 is observed and highlighted by the decrease of the signal intensity. COMPARISON T2 VS T3: 1 H NMR spectra: OH and HCO 3 signal intensities are lower for T3 than T2, indicating a stronger effectiveness of T3 than T2 or a potential attack of the mineral matrix with T3. δ 13 C values and δ 18 O values: There are systematic offsets between T2 and T3 for both δ 13 C values and δ 18 O values. The mean difference values are 0.2 ± 0.1 ‰ for δ 13 C and 0.5 ± 0.2 ‰ for δ 18 O. Fig 2: 1 H NMR spectrum gives information mainly about the mineral fraction of bone, about the water presence, and in a lesser extent about the lipid content Fig. 11: Zoom on 13 C NMR spectra showing a differential elimination of collagen for the same individual Fig. 7: Superimposition of 13 C NMR spectra obtained for T0, T1, T2 and T3. Each curve is a summation of the 5 individual spectra from SB Fig. 4: Superimposition of 1 H NMR spectra obtained for T0, T1, T2 and T3. Each curve is a summation of the 5 individual spectra from SSPM Fig. 8: Superimposition of 1 H NMR spectra obtained for T0, T1, T2 and T3. Each curve is a summation of the 5 individual spectra from SB Fig. 9: δ 13 C values for each individual from SB according to T0, T1, T2 and T3 Fig. 10: δ 18 O values of each individual from SB according to T0, T1, T2 and T3 82 nd annual meeting of the American Association of Physical Anthropologists, 9-13 April 2013, Knoxville, Tennessee T0 T1 T2 T3