Introduction

In the process of archaeological excavation, soils and sediments play a crucial role, not only constituting the physical environment of archaeological sites but also containing rich information1,2,3. Organic residues preserved in archaeological soils or sediments provide valuable clues for interpreting past human activities, including DNA4,5, proteins6, lipids7, and other biomarkers derived from human activities, animal and plant remains, or environmental deposits. By analyzing these residues, researchers can identify human activities such as hearths8,9, rituals10, and manuring7, thereby providing important evidence for exploring ancient dietary structures, food processing methods, and paleoenvironments.

However, research on organic residues in archaeological soils or sediments has long been neglected. These organic residues continuously degrade over long historical periods and mix with soils or sediments, making them generally indistinguishable to the naked eye, which often leads to insufficient attention during archaeological excavations. Organic residues are easily degraded and contaminated11,12,13,14, and the complex composition of soils or sediments and open environments further facilitate the degradation and mixing of organic residues, increasing the difficulty of analysis. Moreover, extracted archaeological soils or sediments are also susceptible to contamination, and in the absence of proper preservation, the scientific value of organic residues is often further limited due to their easily degradable nature. Although methods such as freezing can effectively preserve organic residues under laboratory conditions, archaeological sites often lack the necessary equipment and conditions to achieve timely and effective preservation. Therefore, how to effectively preserve these precious organic residues at archaeological excavation sites has become a critical issue that urgently needs to be addressed.

Currently, research on preservation methods for organic residues in archaeological soils or sediments remains limited. In recent studies, international scholars have begun to attempt applying resins to the detection and analysis of archaeological soils or sediments. Diyendo Massilani et al. evaluated the preservation of ancient mammalian DNA in sediment samples from 13 archaeological sites across Europe, Asia, Africa, and North America after resin impregnation, finding that resin-impregnated loose materials yielded more DNA than non-impregnated materials15. Caterina Rodríguez de Vera et al. demonstrated the feasibility of analyzing archaeological compounds in polyester resin-impregnated micromorphological thin sections using lipid biomarkers through simulation experiments and archaeological sediment detection analysis16. Some scholars have proposed using freeze-drying and low-temperature conditions to preserve archaeological sediments17, but considering the feasibility of extraction at archaeological sites, the method of resin impregnation is a promising research direction in terms of convenience for application in archaeological field conditions.

From this, it can be seen that preservation methods for DNA in sediments are relatively mature, research on lipids is ongoing but still incomplete, while research on preservation methods for protein residues remains lacking. Given the limitations of existing methods, this study aims to further explore preservation methods for organic residues in archaeological soils—specifically, the resin impregnation method—to develop a simple and applicable preservation method suitable for various archaeological field conditions, overcoming the limitations of freezing preservation methods in field applications. The main objectives of the research include evaluating the effectiveness of the resin impregnation method in preserving organic residues (especially proteins and lipids) in soil and discussing the feasibility of applying this method in archaeological field conditions, providing a basis for the preservation practice of organic residues in archaeological work. The overall experimental design of this study is illustrated in Fig. 1.

Fig. 1: Sample processing flow chart.
figure 1

PRI polyester resin impregnation, ERI epoxy resin impregnation.

Full size image

Methods

Sample preparation and experimental design

Fresh beef (30 g) was selected as the experimental substrate based on two key considerations: (1) its archaeological prevalence as a significant protein source in ancient human activities, making it relevant for paleod- iet research; and (2) its biochemical composition encompassing diverse analyzable biomolecules including proteins, fatty acids, and amino acids, which serve as comprehensive markers for understanding historic dietary patterns and environmental conditions18,19. The beef sample was buried in controlled soil conditions (temperature 16.7 °C, moisture 24.5%, measured with a digital soil thermometer and moisture meter) and maintained in dark, moist conditions. After three months of decomposition, the beef had fully degraded and became visually indistinguishable from the surrounding soil matrix. The soil samples, containing the degraded beef, were kept buried for a total of six months before being extracted for analysis. Approximately 0.5 g aliquots were collected and subjected to seven parallel preservation treatments: (1) room temperature storage, (2) freezing preservation (−40 °C), (3) dehydration followed by polyester resin impregnation, (4) dehydration followed by epoxy resin impregnation, (5) dehydration treatment alone, (6) direct polyester resin impregnation, and (7) direct epoxy resin impregnation. An unmodified soil sample was maintained as a negative control. With the exception of the freezing preservation group, all samples underwent room temperature aging for six months.

Two synthetic resins were selected based on their established utility in archaeological applications: unsaturated polyester resin and epoxy resin. This selection was informed by previous research in soil micromorphology20, ancient DNA preservation in sediments15, and lipid biomarker conservation16. The physicochemical properties of the selected resins are detailed in Table 1.

Table 1 Impregnation resin information
Full size table

Bone fragment samples recovered from an archaeological hearth context (F4) at the Lushanmao site in Yan’an City, Shaanxi Province, a significant large-scale settlement from the Miaodigou Phase II Culture (late Neolithic period, ca. 2300–2200 BCE), were utilized to validate the preservation methodologies (Fig. 2). The sampling location was precisely documented through detailed site mapping and stratigraphic positioning (Fig. 2). After photographic documentation and metric recording, the bone fragments were pulverized using a mortar and thoroughly homogenized to ensure sample uniformity. The homogenate was divided into seven equivalent portions: one portion was immediately analyzed as an initial reference, while the remaining six were allocated to two experimental groups (three replicates per group). One group underwent epoxy resin impregnation, and the other was stored in centrifuge tubes at ambient temperature. Both epoxy resin-impregnated and non-impregnated archaeological samples were maintained under identical environmental conditions (at an ambient temperature of (23 ± 2) °C and relative humidity of approximately 40–60%) for six months. Environmental parameters were regularly monitored throughout the preservation period. Upon completion of the storage duration, samples from both preservation conditions were extracted for comparative analysis.

Fig. 2: Archaeological bone fragment recovered from hearth context prior to processing.
figure 2

Scale bar: 1 cm.

Full size image

The epoxy resin impregnation protocol was implemented as follows: Centrifuge tube interiors were pre-treated with a silicone-based mold release agent (Polydimethylsiloxane, CAS: 63148-62-9, Buehler, Lake Bluff, IL, USA) and allowed to dry completely. Subsequently, a Bisphenol A-type epoxy resin (Truer MC002 Quikfirm, CAS: 25068-38-6) was prepared to a viscous consistency and introduced to the bottom portion of each tube. Pulverized sample material was evenly distributed across the resin surface, positioned toward the periphery to facilitate future sampling. Additional epoxy resin was then gradually dispensed to fully immerse the sample. Under ambient laboratory conditions ((23 ± 2) °C), the resin achieved preliminary polymerization within approximately 90 minutes, permitting safe handling and transport. Following a 24-hour complete curing period at room temperature, the intact resin blocks containing the impregnated samples were easily extracted from the centrifuge tubes due to the pre-applied release agent.

For analytical access to preserved specimens, sterile dental drill bits were utilized to extract approximately 50 mg of powdered sample material directly from the resin blocks. Prior to drilling, the resin block surfaces were disinfected with 75% ethanol to prevent contamination. For non-impregnated samples, equivalent amounts of powder were directly collected from the centrifuge tubes. The epoxy resin impregnation process is illustrated in Fig. 3, showing both the preliminary solidification phase with samples still in centrifuge tube molds (Fig. 3A) and the final resin blocks after removal (Fig. 3B). As shown, the resin-impregnated samples maintain their spatial integrity within transparent blocks, allowing for visual inspection of sample distribution while providing protection from environmental degradation factors. The cylindrical shape of the resulting blocks facilitates both storage and subsequent sampling procedures, as specific areas of interest can be targeted using dental drill bits as described in the sampling protocol.

Fig. 3: Epoxy resin impregnation of archaeological soil samples.
figure 3

A Archaeological soil samples impregnated in epoxy resin prior to removal from centrifuge tube molds (S4, S5, S6 represent different experimental samples). B The same samples after removal from the molds, showing the intact resin blocks with impregnated soil samples. Scale unit in cm.

Full size image

Amino acid analysis

Amino acid analysis was performed using an amino acid analyzer (Biochrom 30+, Biochrom, UK). The ninhydrin reaction method was employed21, with the AAS18 standard diluted twice with sodium citrate buffer for machine analysis, resulting in a concentration of 1.25 μmol/mL after dilution. For amino acid chromatography of protein hydrolysates, three buffers with different pH values (pH 3.20, pH 4.25, pH 6.45, all 0.2 M sodium citrate based) and one regeneration solution (0.04 M sodium hydroxide, CAS: 1310-73-2) were used for stepwise elution. Sample preparation: 50 mg of powdered sample was weighed into a hydrolysis tube, 10 mL of 1:1 analytical grade hydrochloric acid (CAS: 7647-01-0) was added, nitrogen was introduced into the hydrolysis tube for 30 seconds and sealed, then hydrolyzed in a 110 °C oil bath for about 24 h. After hydrolysis, the sample was cooled to room temperature and filtered through a 0.45 μm membrane into a 50 mL volumetric flask for volume adjustment. 2 mL of sample was taken for deacidification treatment until a small amount of solid residue remained, then dissolved in 2 mL of sample buffer (0.2M Sodium Citrate, pH 2.2), filtered through a 0.45 μm filter, and analyzed by the amino acid analyzer. All water used was molecular biology grade.

Lipid analysis

Fatty acid analysis was conducted using an Agilent 7820A-5977B GC-MS system with a CP-Sil 88 column (100 m × 0.25 mm × 0.25 μm). Sample preparation was as follows: An appropriate amount of powdered sample was weighed into an EP tube, 4 mL of chloroform solution (CAS: 67-66-3) was added, and the mixture was shaken for 30 seconds to ensure uniform mixing. Then, 0.9% NaCl (w/v) solution (CAS: 7647-14-5) was added, and the operation was repeated to obtain a mixed sample. The sample was centrifuged at 3500 rpm at room temperature for 15 minutes, allowed to settle, and the lower liquid was extracted. The remaining sample was subjected to a second extraction with 2 mL of dichloromethane (CAS: 75-09-2), repeating the above operation. The lower liquids from both extractions were combined and dried under nitrogen gas (specific conditions). After drying, 2 mL of methanol (containing 5% sulfuric acid, CAS: 77-75-8, CAS: 7664-93-9) was added, shaken for 30 seconds, and incubated in an 80 °C water bath for 2 h. After cooling, 2 mL of n-hexane (CAS: 110-54-3) and 1 mL of water were added, mixed, shaken for 30 seconds, and centrifuged at 2000 rpm for 5 min. The supernatant was further treated with 1 mL of water, repeating the shaking and centrifugation steps, and the resulting supernatant was dried under nitrogen gas. An appropriate amount of isooctane (CAS: 540-84-1) was added based on sample concentration, shaken for 30 seconds, allowed to settle, and the solution was transferred to a sample vial. An appropriate amount of the extracted sample was centrifuged at 12,000 rpm for 10 minutes, and the supernatant was used for analysis.

Protein quantification

Protein extraction was performed using a standardized SDS-TFA protocol. Briefly, samples were homogenized in 1 mL lysis buffer (4% SDS (w/v), 0.1% TFA (v/v) in water) using a high-throughput tissue grinder with three successive grinding cycles. The homogenates were then subjected to thermal denaturation at 95 °C with concurrent agitation (1000 rpm) for 30 minutes. Following centrifugation (16,000 g, 4 °C, 20 minutes), the supernatants were collected and proteins were precipitated by adding pre-chilled acetone at a 1:5 ratio (v/v) and incubating overnight at −20 °C. The precipitated proteins were collected by centrifugation (13,000 g, 4°C, 20 minutes), and the resulting pellets were washed three times with 90% pre-chilled acetone. The purified protein pellets were then resolubilized in lysis buffer, sonicated on ice for 3 minutes, and clarified by centrifugation (13,000 g, 4 °C, 20 minutes) to obtain the final protein extract for subsequent analyses.

Protein concentrations were determined using the bicinchoninic acid (BCA) assay with bovine serum albumin (BSA) as the reference standard. A standard curve was generated using seven BSA concentrations ranging from 0.05 μg μL−1 to 0.6 μg μL−1. BCA working reagent was prepared by combining reagents A and B at a 50:1 ratio immediately before use. For each measurement, 20 μL of protein extract was mixed with 200 μL of BCA working reagent and incubated at 37 °C for 20 min. Absorbance values were measured using a DeNovix micro-spectrophotometer. All samples were analyzed in triplicate, and the mean values were used for subsequent calculations.

ZooMS

For ZooMS (Zooarchaeology by Mass Spectrometry) analysis, 30 μg of protein from each extract was transferred to a new microcentrifuge tube and supplemented with 20 μL of digestion buffer (0.1 M Urea (CAS: 57-13-6), 5 mM TCEP (CAS: 51805-45-9), 10 mM CAA (CAS: 543-34-0)). The samples were heated at 95 °C for 5 min and then cooled to room temperature before adding trypsin/lysine-C protease mixture. Enzymatic digestion proceeded at 37 °C with agitation (800 rpm) for 2 h and was terminated by adding 10% TFA solution (CAS: 76-05-1). The resulting peptides were desalted using C18 solid-phase extraction columns with sequential elutions (2 × 30 μL elution buffer). The purified peptides were vacuum-concentrated and subjected to matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALdi-TOF-MS) analysis on a Bruker autoflex MALDI-TOF mass spectrometer (Bruker Daltonics, Bremen, Germany). Species identification was performed by comparing the acquired mass spectra with a reference collagen peptide database compiled from the supplementary data of Xia et al. (2024)22.

Results

Amino Acid Analysis

Amino acid analysis was performed on a Biochrom 30+ amino acid analyzer (Biochrom, Cambridge, UK). The analysis is based on ninhydrin post-column derivatization21. An AAS18 amino acid standard solution (Sigma-Aldrich, St. Louis, MO, USA) was diluted 1:1 with sodium citrate loading buffer to a final concentration of 1.25 μmol mL−1 for calibration. Protein hydrolysates were separated using a stepwise elution program with three sodium citrate buffers (0.2 M at pH 3.20, 4.25, and 6.45) and a regeneration solution (0.04 M NaOH, CAS: 1310-73-2). For sample preparation, 50 mg of powdered sample was hydrolyzed in 10 mL of 6 M HCl (CAS: 7647-01-0) under a nitrogen atmosphere at 110 °C for 24 h. After cooling, the hydrolysate was filtered (0.45 μm), brought to a final volume of 50 mL, and an aliquot was dried to remove acid. The residue was redissolved in sample loading buffer (pH 2.2), filtered again (0.45 μm), and injected for analysis. All aqueous solutions were prepared with molecular biology grade water.

Figure 4 The isoleucine content in all samples is similar to or slightly higher than that in fresh beef samples, a phenomenon consistent with the characteristic of high isoleucine content in blank soil samples. This may be due to certain soil microorganisms or plant residues being rich in isoleucine, and the resin impregnation process better retained the amino acid components of the soil itself, leading to higher detection results for isoleucine18. This finding suggests that when interpreting the amino acid composition of archaeological samples, the potential influence of soil microbial communities needs to be considered. Additionally, the content of valine in buried samples is markedly lower compared to fresh beef (one-sample t-test, p < 0.001), which may be related to microbial activity, specific soil chemical environments, or selective loss, requiring further research to elucidate its mechanism (Fig. 4).

Fig. 4: Amino acid concentration under different methods.
figure 4

Heat map showing amino acid concentration (mg/g) across different preservation methods. Color intensity represents concentration levels, with brighter colors indicating higher concentrations.

Full size image

Figure 5 The amino acid analysis results of this study indicate that resin impregnation, especially epoxy resin impregnation, has certain advantages in preserving the amino acid composition of organic residues in archaeological soils. This method not only slows down the degradation of amino acids but may also help retain the characteristic components of the soil itself. Although the amino acid detection results include amino acids produced during the hydrolysis process and cannot completely distinguish between free amino acids originally present in the sample and protein hydrolysis products, this does not prevent the assessment of sample preservation conditions and the effects of different treatment methods through the overall content and composition of amino acids. The results show that resin-impregnated samples, especially epoxy resin-impregnated samples, have amino acid compositions closer to that of fresh beef, indicating their advantage in preserving organic residues (Fig. 5).

Fig. 5: Amino acids distribution under different methods.
figure 5

Radar chart displaying the relative compositional profile of amino acids, normalized as a proportion of the total amino acid content to highlight similarities with the reference sample. Each axis represents a different amino acid, and values are shown as percentages of total amino acid content.

Full size image

Protein quantification analysis

Figure 6 shows the distribution of protein concentrations in parallel sample groups using the bicinchoninic acid protein assay (BCA) under different preservation conditions. It is evident that different preservation methods exert statistically significant differences in their effects on protein content (one-way ANOVA, F(7, 8) = 343.34, p < 0.001). Direct impregnation with epoxy resin shows the best protein preservation effect, with an average concentration of 0.2906μg μL−1, far exceeding those of the other treatment methods. This result is consistent with previous amino acid analysis results, further confirming the advantage of epoxy resin in preserving organic residues. In contrast, direct impregnation with polyester resin and impregnation after dehydration show poor effects, with concentrations of only 0.062μg μL−1.

Fig. 6: Protein concentration distribution under different preservation conditions.
figure 6

Bars represent mean protein concentration (μg μL−1), while individual data points show replicate measurements. Error bars indicate standard deviation. Numbers above bars indicate fold-change relative to blank soil control. The highest value (ERI) is highlighted in yellow. The differences among the groups were statistically significant (one-way ANOVA, p < 0.001).

Full size image

Samples impregnated with resin after dehydration show considerable differences. The protein concentration in epoxy resin-impregnated samples (0.1655μg μL−1) is significantly higher than that in polyester resin-impregnated samples (Sample 3, 0.0522μg μL−1). This result is consistent with the trend in amino acid analysis, reflecting the superiority of epoxy resin in protecting biological macromolecules such as proteins. Room temperature storage (0.1524μg μL−1) and frozen storage (0.1238μg μL−1) are less effective than direct epoxy resin impregnation, indicating that traditional low-temperature freezing preservation methods have limitations in maximally protecting protein residues at archaeological sites. Simple dehydration treatment (Sample 5, 0.1036μg μL−1) shows poor preservation effect on proteins, which is attributed to protein denaturation and degradation during the dehydration process, leading to a loss of structural integrity23,24. This result is consistent with the previous amino acid analysis results, further confirming the limitations of simple dehydration treatment in preserving organic residues.

It should be noted that the BCA method detects the content of intact proteins in the sample, not the total amount of amino acids after complete hydrolysis. This differs from the amino acid analysis mentioned earlier. Amino acid analysis reflects the amino acid composition of all proteins and peptides in the sample after complete hydrolysis, while BCA determination mainly reflects the content of well-preserved proteins25. The results of these two methods complement each other, providing a more comprehensive assessment of sample preservation effects.

Samples directly impregnated with epoxy resin show the highest protein content in BCA detection, indicating that this method not only preserves amino acids but also maintains protein integrity well. In contrast, polyester resin-impregnated and dehydrated samples show higher total amino acid content (such as glutamic acid and aspartic acid) in amino acid analysis but lower protein content in BCA detection, suggesting that proteins in these samples may have partially degraded into short peptides or individual amino acids. These results collectively indicate that the direct epoxy resin impregnation method has significant advantages in preserving protein residues in archaeological soils.

Fatty Acid Analysis

To assess the impact of different preservation methods on organic residues in archaeological soils, this study conducted quantitative analysis of fatty acids in the samples. Gas chromatography-mass spectrometry (GC-MS) was used to detect and quantify medium and long-chain fatty acids in the samples, a method widely applied in the analysis of archaeological organic residues26. As seen in Fig. 7, which shows the distribution of specific fatty acid content under different preservation methods, epoxy impregnation and epoxy dehydration impregnation methods show the highest content of characteristic fatty acids, indicating that these two methods are overall most effective in preserving organic residues.

Fig. 7: Key fatty acid content under different preservation methods.
figure 7

Concentration (ng/g) of key fatty acids (C18:1n9c, C18:2n6c, C20:3n3) across preservation methods, highlighting differences in unsaturated fatty acid retention.

Full size image

The fatty acid profile of all samples was dominated by saturated fatty acids including C16:0 (palmitic acid) and C18:0 (stearic acid), consistent with the common fatty acid composition in soils27. C22:0 is present in high amounts in blank soil samples, suggesting that C22:0 in various samples may be inherent to the soil. Since these three saturated fatty acids are high in content, inherent in soil, and difficult to distinguish from organic acids resulting from sample degradation, we specifically focused on the distribution of fatty acids that serve as more reliable biomarkers (see Fig. 8). Oleic acid (C18:1n9c) and linoleic acid (C18:2n6c) were selected as primary indicators for evaluating preservation effectiveness because these unsaturated fatty acids: (1) originate predominantly from animal tissues rather than soil microorganisms28, (2) are particularly susceptible to degradation through oxidation processes, making them sensitive indicators of preservation quality. Additionally, eicosatrienoic acid (C20:3n3) was selected as a third indicator due to its status as a diagnostic polyunsaturated fatty acid present in animal tissues that rapidly degrades in suboptimal preservation conditions.

Fig. 8: Characteristic fatty acid distribution under different methods.
figure 8

Heat map showing the distribution of characteristic fatty acids, with color intensity proportional to concentration. Major saturated fatty acids (C16:0, C18:0, C22:1n9) were excluded to enhance visualization of characteristic fatty acids. Color gradient ranges from teal (low) to yellow (medium) to orange (high).

Full size image

It can be seen that in samples impregnated with epoxy resin (both direct impregnation and impregnation after dehydration), polyester resin impregnation, and epoxy resin impregnation after dehydration, the content of oleic and linoleic acids is significantly higher than in blank soil samples. This indicates that these preservation methods can effectively preserve organic residues from buried meat, with the direct epoxy resin impregnation method showing the best effect in preserving these two unsaturated fatty acids. In contrast, samples preserved at room temperature, frozen, and simply dehydrated show oleic and linoleic acid contents similar to blank soil samples, indicating that these traditional methods have limited effectiveness in protecting organic residues and may lead to severe degradation of organic matter or mixing with soil matrix. Notably, eicosatrienoic acid (C20:3n3) shows abnormally high content in polyester resin-impregnated and polyester dehydration-impregnated samples. C20:3n3 is a polyunsaturated fatty acid widely present in nature, found in animals, plant sources (such as flaxseed and hemp seed), and marine organisms (such as algae and fish)29,30. In this experiment, the high content of this fatty acid likely originates from the buried meat samples rather than the soil itself, suggesting that the polyester resin impregnation method may have unique advantages in preserving certain types of animal-derived fatty acids such as C20:3n3. Fig. 8 also shows that C14:0 (myristic acid) has higher content in dehydrated and epoxy-impregnated samples, which may reflect the advantages of these methods in preserving certain saturated fatty acids. C20:1 (11-eicosenoic acid) also shows relatively high content in epoxy-impregnated samples, further confirming the effectiveness of epoxy resin in preserving long-chain unsaturated fatty acids. These observations are consistent with Regert’s (2011) research results on the preservation effects of different methods on specific fatty acids.

These results support the superiority of epoxy resin impregnation, especially direct impregnation, in preserving organic residues in archaeological soils. This method not only effectively preserves unsaturated fatty acids (such as oleic and linoleic acids) but also shows good protective effects on certain long-chain fatty acids. Although polyester resin impregnation performs well in preserving certain specific fatty acids (such as C20:3n3), its overall effect is not as good as epoxy resin. Traditional preservation methods (such as room temperature storage and freezing) show obvious limitations in long-term protection of organic residues.

Protein preservation in archaeological bone samples

To validate the efficacy of epoxy resin impregnation for preserving archaeological materials, we analyzed protein concentrations in archaeological bone fragment samples stored under different conditions for six months. The initial protein concentration in freshly prepared bone samples averaged 0.2981μg μL−1 (Fig. 9). Following the six-month storage period, statistically significant differences in protein preservation emerged between the different preservation methods (one-way ANOVA, p < 0.01). Samples impregnated with epoxy resin maintained remarkably high protein integrity, with a mean concentration of 0.2753μg μL−1, representing 92.4% retention of the initial protein content. In contrast, samples stored at room temperature exhibited substantial protein degradation, with concentrations decreasing to 0.1756μg μL−1, corresponding to only 58.9% retention of the initial protein content (Fig. 9).

Fig. 9: Protein preservation in archaeological bone samples after 6-month storage.
figure 9

The box plot shows the distribution of protein concentrations for each condition (Initial, Epoxy Resin, and Room Temperature). Boxes represent the interquartile range with the median line, and whiskers extend to the data range. Individual data points (n = 3) are overlaid on each box. Mean values and protein retention rates are indicated for each group. The differences among the groups are statistically significant (one-way ANOVA, p < 0.01).

Full size image

Notably, the epoxy resin-impregnated samples demonstrated not only superior protein preservation but also greater consistency across replicates (Fig. 9), with protein concentrations ranging from 0.2597μg μL−1 to 0.2853μg μL−1. This narrow range suggests that epoxy resin impregnation provides stable and reliable preservation conditions. Room temperature storage, however, showed greater variability between replicates (0.1571μg μL−1 to 0.1955μg μL−1), indicating less consistent preservation under conventional storage conditions. These findings provide compelling evidence that epoxy resin impregnation significantly outperforms conventional room temperature storage for preserving protein content in archaeological bone material over extended periods.

ZooMS analysis successfully identified both the initial bone samples and those stored for 6 months as pig (Sus scrofa). Further peptide spectrum analysis revealed significant differences in preservation quality between storage methods (Fig. 10). Prior to visualization, all mass spectra were processed using mMass software, which included baseline correction and normalization to the most intense peak to facilitate comparison. The stacked spectral comparison clearly demonstrates variations in peptide marker retention across the three preservation methods. The direct analysis samples (top panel, purple in Fig. 10) exhibited strong signals for all four diagnostic peptide markers (e.g., m/z 1453, 2131, 2883, and 3034, corresponding to markers A, D, E, and F respectively, following the nomenclature of Brown et al. (2020))31, confirming the initial protein integrity. In contrast, samples preserved with epoxy resin (middle panel, yellow) maintained excellent signal strength at m/z 1453, with clearly detectable peaks at m/z 2131, 2883, and 3034, though at slightly reduced intensities compared to direct analysis. This indicates that epoxy resin impregnation effectively preserved the key diagnostic peptides required for species identification. Most notably, the room temperature-stored samples (bottom panel, teal) showed significant protein degradation. While the marker at m/z 1453 remained detectable with reduced intensity, and m/z 2131 was barely visible, the critical markers at m/z 2883 and 3034 were effectively lost in the background noise, making species identification substantially more challenging. The intense peaks observed below m/z 1000 are likely due to matrix ions or contaminants and were not considered for species identification. These findings demonstrate that epoxy resin impregnation not only preserves total protein quantity but also maintains protein structural integrity, retaining critical biological information. This is particularly significant for archaeological samples, where preservation of diagnostic peptide markers is essential for accurate species identification and functional protein studies.

Fig. 10: MALDI-TOF mass spectra comparing preservation methods for archaeological bone samples.
figure 10

The displayed spectra were baseline-corrected and normalized to the most intense peak. direct analysis control (top, purple), epoxy resin-preserved samples (middle, yellow), and room temperature storage without resin (bottom, teal). ZooMS analysis identified all samples as pig (Sus scrofa), with four diagnostic collagen peptide markers (m/z 1453, 2131, 2883, and 3034) highlighted by vertical shaded bars. Epoxy resin preservation maintains all diagnostic markers with signal intensities comparable to control samples, while conventional room temperature storage exhibits substantial signal attenuation for higher molecular weight markers (m/z 2883 and 3034). These results demonstrate that epoxy resin impregnation provides superior preservation of protein biomarkers essential for species identification in archaeological remains.

Full size image

Discussion

The experimental results of this study indicate that among the preservation methods for organic residues in archaeological soils, epoxy resin impregnation shows significant advantages. This method outperforms others in preserving protein integrity, maintaining amino acid composition, and protecting unsaturated fatty acids across multiple key indicators. Particularly in terms of protein preservation, the protein concentration in epoxy resin-impregnated samples (0.2906μg μL−1) is far higher than other methods, indicating its good effect in maintaining the integrity of macromolecular structures. Although freezing preservation and polyester resin impregnation also show certain advantages in some specific indicators, their overall performance is not as good as epoxy resin impregnation.

The superior preservation performance of epoxy resin impregnation likely results from multiple synergistic mechanisms. Resin impregnation forms a closed microenvironment that effectively isolates samples from external factors such as air, moisture, and microorganisms, thereby significantly reducing oxidation, hydrolysis, and biodegradation processes30,32,33. The molecular structure of epoxy resin contains epoxy groups (-C-O-C-) that form highly cross-linked three-dimensional networks during curing34,35,36, allowing it to react with various functional groups, including hydroxyl, amino, and carboxyl groups, enabling epoxy resin to form more chemical bonds with organic molecules (such as proteins) in the sample37. In contrast, polyester resin primarily consists of linear or branched molecules connected by ester bonds (-COO-)38 with a lower degree of cross-linking39, resulting in a looser network formed by polyester resin, with poorer barrier properties against water and other small molecules compared to epoxy resin. Additionally, epoxy resin’s lower shrinkage rate (2%–7% versus 5%–10% for polyester resin)40,41 better maintains the original shape and spatial relationships within samples, which is particularly important for preserving the microstructure of archaeological samples.

This study represents the first systematic validation of epoxy resin impregnation for preserving organic residues in actual archaeological bone samples. Using real archaeological bone samples, we confirmed the efficacy of epoxy resin impregnation for preserving organic residues. While protein content and ZooMS peptide marker intensities in resin-impregnated samples were slightly lower than in fresh samples—potentially due to dilution effects from resin powder during sampling—they significantly outperformed room temperature storage methods. Compared to conventional room temperature storage, epoxy resin impregnation significantly enhanced protein retention (92.4% versus 58.9% of initial content) while better maintaining structural integrity and diagnostic information. These findings align with our simulation experiments, confirming the reliability and effectiveness of epoxy resin impregnation as a preservation method for archaeological materials.

Despite its demonstrated advantages, a balanced discussion requires acknowledging several considerations for the application of the epoxy resin impregnation method. Firstly, the irreversible nature of the embedding process warrants consideration, as it may limit the applicability of certain future analytical techniques. A notable example is radiocarbon dating of bone collagen, where the introduction of exogenous carbon from the resin presents a significant challenge that future research could aim to address, for instance, through novel decontamination protocols. Secondly, while modern epoxy resins are generally inert, the potential for trace-level leaching of unreacted components should be considered for ultra-sensitive analyses, meriting further investigation. Thirdly, in terms of field logistics, the method is more resource-intensive in cost and labour compared to simple bagging, representing a trade-off between preservation quality and operational simplicity. Finally, the choice of preservation method should be guided by the primary research questions. For instance, while epoxy resin provides a superior physical barrier, less invasive methods like silica gel might be preferred if DNA analysis is the sole priority28. Therefore, the epoxy resin method is best viewed as a powerful tool within a broader toolkit of preservation strategies, to be selected following a careful evaluation of a sample’s analytical potential.

The epoxy resin impregnation method offers several significant advantages for preserving organic residues in archaeological soils. It maintains spatial integrity by fixing samples in their original state, preserving the relative positions of components—critical for studying distribution patterns of organic residues in sediments. The method creates a sealed preservation environment that significantly reduces degradation processes, while the chemical properties of epoxy resin provide additional protection through potential interactions with biomolecules. The dimensional stability of epoxy resin (with only 2-7% shrinkage) helps preserve the original morphology of samples, reducing physical stresses on organic molecules. Furthermore, our successful extraction and analysis of proteins from resin-impregnated archaeological samples demonstrates that this method not only preserves samples but remains compatible with subsequent biomolecular analyses, including micromorphological analysis, elemental analysis, and organic molecule extraction. This provides a practical solution for long-term preservation of field-collected archaeological samples.

Based on our experimental findings, we developed a simplified and efficient epoxy resin impregnation protocol specifically suited for archaeological field conditions. This protocol employs disposable plastic centrifuge tubes with silicone-based release agent treatment, followed by a layered impregnation technique using E51-type epoxy resin mixed with hardener in a 10:1 ratio. The resin reaches preliminary solidification within approximately 90 minutes at room temperature ((23 ± 2) °C), allowing safe sample transport on the same day. After 24 hours of complete curing, the pre-applied release agent allows easy removal of intact resin blocks without additional tools or container damage. These blocks can be directly labeled or stored in sealed bags at room temperature without specialized storage equipment. This streamlined protocol significantly reduces implementation complexity, featuring short processing times and minimal equipment requirements ideally suited for archaeological field applications.

Resin impregnation methods demonstrate clear advantages in preserving organic residues in archaeological soils. Among these, epoxy resin impregnation, particularly the direct impregnation method, shows excellent performance in preserving amino acids, proteins, and fatty acid content in both simulated samples and actual archaeological bone fragments. In contrast, traditional preservation methods such as room temperature storage, freezing, and simple dehydration have limitations in long-term protection of organic residues. The epoxy resin impregnation method not only effectively preserves the quantity of organic residues but also maintains their compositional characteristics and molecular integrity well, which is of significant importance for subsequent archaeological analysis. Therefore, the epoxy resin impregnation method has the potential to become one of the effective methods for preserving soil organic residues at archaeological sites, providing a feasible preservation solution for excavation sites lacking good preservation conditions.

Future research is still needed, including assessing the impact of resin impregnation on the types of organic residue molecules and their biological traceability, exploring longer-term preservation effects, refining on-site operational protocols, and determining how to minimize potential contamination and interference with samples during the preservation process. Crucially, further work must address the incompatibility of this method with key analyses like radiocarbon dating to either develop decontamination procedures or to define a clear decision-making framework for its application.