Olive oil from the 79 A.D. Vesuvius eruption stored at the Naples National Archaeological Museum (Italy)

The study investigates whether a solidified organic residue preserved in a Roman glass bottle from the Vesuvius archaeological sites is authentic ancient olive oil and characterizes how its molecular composition evolved over nearly two millennia. Pompeii and Herculaneum, buried during the 79 A.D. eruption, preserve daily-life artifacts including food remains and containers. While many similar bottles are presumed to contain solidified olive oil, authentication is essential because some museum-held organic materials have proved to be of later origin. The purpose is to authenticate the Mann-S1 sample via radiocarbon dating and chemical biomarkers, and to map the extensive chemical transformations caused by the initial thermal shock and prolonged storage. This work is significant for food history, archaeometry, and lipid chemistry, potentially documenting the oldest bulk olive oil residue and providing insight into long-term lipid degradation pathways.

Previous attempts to characterize lipid residues in archaeological vessels used spectroscopic and chromatographic methods but often yielded inconclusive results, especially in distinguishing plant from animal fats due to profound compositional changes over time. NMR or GC/MS of fatty acids alone may not unambiguously determine origin because unsaturated fatty acids oxidize more rapidly, altering profiles. Some studies on prehistoric lipids and pottery residues highlighted these challenges. Additionally, prior assumptions of lipid time stability overlooked dramatic modifications induced by thermal shock and millennial storage. Reports exist of olive oil use and trade in Roman Campania and beyond, and early 19th-century sensory-based examinations of similar bottles, but robust chemical authentication and chronological placement were lacking.

Archaeological sample: A cylindrical glass bottle (Inv. 848747) from the MANN collection, likely from Herculaneum, half-filled with yellowish waxy matter shaped by solidification, was sampled (inner material, after scraping surface) using sterile tools. Aliquots were stored at −20 °C under N2. Portions were extracted in chloroform:methanol (2:1), filtered (0.22 µm nylon), and solvent-evaporated under N2. Analytical strategy: An integrated suite included UHPLC-ESI-MS/MS (Q Exactive Orbitrap; negative and positive modes; 200–1200 m/z; biphenyl column; water/methanol gradient), MALDI-TOF MS (UltraflexExtreme; DHB matrix; Na+ doping; reflector positive mode), 1H/13C NMR (600/150.9 MHz, CDCl3-d; DEPT and 2D experiments for assignments), high-resolution GC-FID of FAMEs (SP-2560 column; methanol/HCl transmethylation; temperature program 200–230 °C), free acidity by microtitration (expressed as palmitic acid equivalents) and peroxide value by modified iodometry, ATR-FTIR (PerkinElmer Spectrum 400; 4000–650 cm−1; comparison with EVOO), GC of sterols after saponification and extraction (RTX-5 column, FID), SPME-GC/MS of volatiles (DVB/CAR/PDMS fiber; 40 °C headspace; Supelcowax-10 column; 40–240 °C program; NIST spectral match), and radiocarbon dating (acid–alkali–acid pretreatment; AMS at INFN-LABEC; calibration with OxCal 4.3 using IntCal13). Fresh extra-virgin olive oil served as a reference for spectral and chromatographic comparisons.

- Authentication and dating: AMS radiocarbon dating placed the sample contemporaneous with the 79 A.D. Vesuvius eruption, supporting authenticity as an ancient artifact. The bottle contains approximately 0.7 L of residue, likely the oldest bulk olive oil remnant. - Lipid class transformation: No detectable glycerol-based lipids (mono-, di-, or triacylglycerols) were observed by UHPLC-MS/MS, MALDI-TOF MS, or NMR. 1H/13C NMR lacked glycerol backbone signals and TAG ester carbonyls; instead, a strong free fatty acid carbonyl at ~179.6 ppm was present. - Hydrolysis and oxidation: Free acidity was extremely high at 68.87% (as palmitic acid equivalents), indicating near-complete hydrolysis of TAGs; peroxides were not detected, consistent with a late-stage oxidative state. - Fatty acid profile (GC-FID): Dominant palmitic acid (C16:0) 91% and stearic acid (C18:0) 3.6%. Residual oleic acid (C18:1n-9c) 2.3% and its trans isomer elaidic acid (C18:1n-9t) 1.5%; linoleic (C18:2) and linolenic (C18:3) acids were not detected. Trace C20:0, C22:0, C24:0 mirrored olive oil ratios. The C16:0/C18:0 ratio was compatible with olive oil. Presence of elaidic acid suggests heat-induced cis/trans isomerization consistent with eruption thermal shock. - Estolide formation: UHPLC-HR-ESI-MS/MS and MALDI-TOF MS revealed rare estolides formed from hydroxy fatty acids, including dimers and trimers such as HS–HS (m/z 581.5168), P–HS (m/z 537.4901), O–HS (m/z 563.5067), P–HS–HS (m/z 819.7473), and O–HS–HS (m/z 845.7625). Diagnostic MS/MS fragment ions supported assignments. NMR corroborated estolide linkages (ester methine at ~4.83–4.86 ppm; ester carbonyl ~173–174 ppm; hydroxy-carbon resonances ~72–74 ppm). - Residual unsaturation: Very low-intensity olefinic signals in NMR and disappearance of olefinic C=C stretch in FTIR indicated near-complete saturation or modification of double bonds. FTIR showed shift of ester carbonyl (1745 cm−1) to free acid (1707 cm−1) and an estolide carbonyl shoulder (~1733 cm−1). A trans-double bond band (~967 cm−1) is consistent with some elaidic acid. - Volatile compounds: SPME-GC/MS indicated a complex volatile profile mainly derived from oleic acid breakdown (details not enumerated in the excerpt), consistent with advanced oxidation and possible glycerol-derived volatiles. - Macroscopic solidification: The residue’s solid state and inclined surface suggest in situ solidification; enrichment in saturated FAs and estolide formation likely contributed to solidification.

The combined radiocarbon and multi-omic analytical evidence confirms that the Mann-S1 residue is ancient olive oil dating to the 79 A.D. eruption. The findings address the research question by identifying definitive biomarkers of olive oil origin (FA distribution compatible with olive oil, sterol analysis method applied though results not detailed here) and by documenting the complete hydrolysis of triacylglycerols with extensive oxidation and secondary condensation reactions. The discovery of estolides formed from hydroxy fatty acids reveals a long-term transformation pathway of ancient oils and explains the residue’s physical solidity. The presence of elaidic acid at levels comparable to residual oleic acid indicates a significant thermal event consistent with the eruption temperatures inferred from archaeological contexts, supporting a scenario where initial heating and subsequent millennial storage shaped the current chemistry. These results are significant for archaeometry and food chemistry, offering a rare, well-preserved case that maps the long-term molecular evolution of olive oil and provides a validated framework for authenticating similar archaeological lipid residues.

This study authenticates a unique, bulk quantity residue of olive oil from the 79 A.D. Vesuvius eruption and elucidates its extensive molecular evolution over ~2000 years. Through integrated MS, NMR, FTIR, GC analyses, and radiocarbon dating, the authors show complete TAG hydrolysis, advanced oxidation, formation of hydroxy fatty acids and rare estolides, substantial loss of unsaturation, and evidence for thermal isomerization. The work establishes analytical benchmarks for identifying and aging archaeological oils and highlights estolide formation as a key long-term transformation. Future research should investigate mechanisms and kinetics of estolide formation in ancient matrices (thermal vs enzymatic/microbial), conduct systematic surveys across other MANN bottles and related finds, profile sterols and trace components as provenance markers, and expand volatile characterization to reconstruct ancient oil processing and storage conditions.

The study focuses on a single specimen, which may limit generalizability. The exact mechanism of estolide formation (thermal versus enzymatic/microbial or slow chemical processes) cannot be conclusively determined with the available data. Extensive degradation precluded detection of original glycerol-based lipids, constraining direct profiling of the initial triacylglycerol composition. Potential contamination over two millennia cannot be entirely excluded despite careful sampling and pretreatment. Some analytical results (e.g., detailed sterol and volatile profiles) are only briefly referenced in the provided text, limiting comprehensive interpretation here.