Introduction

The excavated relics from the Sanxingdui Ruins in Guanghan, Sichuan, China, have garnered worldwide attentions for the first time due to the discovery of a large number of intact elephant tusks and ivory artifacts. This remarkable find not only helps in reconstructing the original appearance of ancient human life but also carries immense significance in examining the origin, progression, and decline of civilization. Furthermore, the excavated ivory provides crucial scientific information on ancient Chinese climate, geography, and the environment, adding a deeper layer of understanding to our historical knowledge [1].

Ivory is a typical natural organic–inorganic composite material, with collagen fibers dominating its organic component and hydroxyapatite dominating its inorganic component [2]. The complete structure of ivory consists of enamel, cementum, dentin and pulp cavity from outside to inside (Fig. 1). Only the tip of the ivory bears a conical layer of enamel that usually wears away until it disappears over the course of an elephant’s life cycle [3]. The enamel is highly mineralized, dense and smooth, and has the function of protecting the teeth. Ivory consists almost entirely of dentin, which is covered with a thin layer of cementum (a soft enamel derivative). The structural properties of ivory are mainly determined by the characteristics of the dentin, as it is the main component of the incisor [3, 4]. Dentin is composed of organic collagen, water and 70% minerals (mainly hydroxyapatite), which has mineralized collagen fibers (MCF) similar to lamellar bone. The large number of tiny dentinal tublues in the dentin structure endow the ivory’s flexibility and strength. The ivory has a unique pattern of intersecting lines across its cross section that form small diamond-shaped areas visible to the naked eye, the so-called Schreger pattern. According to existing reports, the Schreger pattern may be attributed to the three-dimensional arrangement of dentinal tubules [4, 5], or to the specific periodic orientation arrangement of the MCF [6, 7]. This unique Schreger band is closely related to the mechanical properties of ivory. With the evolution of underground buried environment for a prolonged period, the original collagen of ivory is depleted, resulting in a porous and loose structure. Once the ivory is excavated and exposed to the atmosphere, the subsequent water loss renders ivory vulnerable to irreversible damage such as shedding, pulverization, cracking, and disintegration. To ensure the long-term preservation of the excavated ivory, it is urgent to conduct a comprehensive research on its pore structure, water content, moisture state, and the deterioration caused by water loss on the basis of its chemical composition and crystal structure.

Fig. 1
figure 1

a The structure of ivory, b the real situation of the underground cold storage room (4 ℃, 95% RH) of excavated ivory from the Sanxingdui Ruins, c the excavated ivory half-wrapped in polymeric bandage, the measured residual samples (d) K7XY-181 and (e) K7XY-61

Full size image

To ensure the long-term preservation of the excavated ivory, it is urgent to conduct comprehensive research on its pore structure, water content, moisture state, and the deterioration caused by water loss based on its chemical composition and crystal structure.

To date, numerous studies on excavated ivory have primarily focused on its composition and crystal structure. For instance, the composition and crystal structure of the excavated ivory from the Jinsha and Sanxingdui ruins have been analyzed using a variety of testing techniques such as amino acid analysis, infrared spectroscopy (IR), and X-ray diffraction (XRD) [8, 9]. In addition, Raman spectroscopy and other testing methods have been used to identify and characterize mammoth ivory [10,11,12]. However, there are few studies on the pore structure, water content and moisture states of the excavated ivory, and most of them have largely unexplored.

Regarding pore and moisture analysis, there have been some relatively mature studies on waterlogged wooden cultural relics. For instance, the pore structure characteristics of waterlogged archaeological wooden can be detailedly investigated through N2 adsorption, mercury intrusion porosimetry, and thermal porosimetry [13]. Additionally, the moisture analysis of waterlogged archaeological wood can be explored in detail by using low-field nuclear magnetic resonance and dynamic water vapor sorption [14]. Inspired by the analysis of pore and moisture characteristics of waterlogged archaeological wood, this paper innovatively conducts a preliminary correlative study on the pore structure and moisture states of the excavated ivory. Firstly, the pore structures of the excavated ivory from the No.7 Sacrificial Pit (K7) at the Sanxingdui Ruins were comprehensively analyzed using dry and wet porosity methods, respectively. Subsequently, a detailed evaluation and analysis of the water content and moisture states of the excavated ivory from K7 at the Sanxingdui Ruins was conducted. The correlation mechanism between pore structure and moisture states was further explored through the water loss deterioration properties of excavated ivory from K7 at the Sanxingdui Ruins. This study provides theoretical support for the long-term preservation of the excavated ivory in the future and offers new perspectives for the conservation of fragile bone and horn cultural relics.

Materials and methods

Materials

The experimental research object of this article was selected from the excavated ivory at the K7 site in the Sanxingdui Ruins. The preservation environment of the ivory after excavation and some test samples of the excavated ivory (K7XY-181 and K7XY-61) are shown in Fig. 1. Special instructions for naming test samples. K7XY-181in and K7XY-61in refer to the interior of K7XY-181 and K7XY-61 that is isolated from the outside world, respectively. Additionally, K7XY-181out and K7XY-61out refer to the outer surface of K7XY-181 and K7XY-61 in contact with the external environment, respectively.

Characterization

The dry method testing of excavated ivory underwent pre-treated by freeze-drying to preserve the original microscopic morphology as comprehensively as possible. Wet method testing preserves the original state of the excavated ivory and is directly used for characterization analysis.

Chemical composition and crystal structure analysis. Fourier Transform Infrared Spectrometer (FTIR, VERTEX 70, Bruker, Germany) was tested by transmission method with a scanning wave number range of 4000–400 cm−1 and a resolution of 4 cm−1. The excavated ivory were mixed with dried KBr powder in a ratio of 1:49 and then compressed into pieces. The elemental composition and chemical state were analyzed using X-ray Photoelectron Spectrometer (XPS, Escalab Xi + , Thermo Fisher Scientific, USA) with an Al Kα X-ray source, a tube voltage of 40 kV, and a tube current of 40 mA. All spectra were calibrated using the C1s binding energy of 284.8 eV. Surface elements were analyzed using Energy Dispersive Spectrometer (EDS, X-MAXN50MM2, Carl Zeiss AG, Germany) operated at a working voltage of 20 kV. Crystal structure analysis was performed using X-ray diffractometer (XRD, MiniFlex600, Rigaku, Japan) with Cu Kα radiation (λ = 1.5406 Å), tube current of 20 mA, tube voltage of 40 kV, a step size of 0.02°, a scanning speed of 10°/min, and a scanning range of 10–80°. The crystal structure and micro-morphology of the excavated ivory was further measured by transmission electron microscope (TEM, HT7800, Hitachi, Japan) with a test voltage of 120 kV, and selected area electron diffraction (SAED) analysis was conducted on local regions.

Pore structure analysis. The apparent morphology and pore structure of the freeze-dried excavated ivory were observed using a scanning electron microscope (SEM, Zeiss EV018, Carl Zeiss AG, Germany) at a working voltage of 15 kV. The specific surface area (SSA) and pore size distribution of the freeze-dried excavated ivory were evaluated by Brunauer–Emmett–Teller (BET, ASAP2460, Micromeritics, USA) method. The pore distribution, average pore diameter, and porosity of the freeze-dried excavated ivory were measured using mercury injection porosimeter (MIP, AutoPore 9500, Micromeritics, USA). The pore radius distribution and porosity of the original excavated ivory were analyzed using the non-destructive technique of low-field nuclear magnetic resonance (LF-NMR, MesoMR23-060H-I, Niumag Analytical Instrument Corporation, China). The resonance frequency for the test was 23 MHz, the magnet strength was 0.5 T, the magnet temperature was 32 ℃, the number of echoes was 5000, and the echo time was 0.2 ms.

Water content, moisture states, and water loss deterioration analysis. The dynamic water loss process and water content of the original excavated ivory at 25, 60, and 100 ℃ were preliminary conducted with the gravimetric method, respectively. The water content and moisture states of the original excavated ivory were further evaluated by thermal gravimetric analyzer (TGA, Q500, TA, USA) under a nitrogen gas flow rate of 50 mL/min, a temperature range of 35–800 ℃, and a heating rate of 10 ℃/min. The Differential Scanning Calorimeter (DSC, PerkinElmer 8500, TA, USA) was employed to estimate the moisture states and their binding strength with other components in the excavated ivory over a temperature range of 10–100 ℃ in a nitrogen atmosphere with a heating rate of 2 ℃/min. The moisture states and proportion of the original excavated ivory were non-destructive calculated by LF-NMR with a resonance frequency of 23 MHz and a magnet strength of 0.5 T. The water loss kinetics of the original excavated ivory was demonstrated by dynamic water vapor adsorption instrument (DVS, SPS11-10 μ, ProUmid, Germany) at a constant temperature of 25 ℃, within a humidity range of 25–95% RH, and with a humidity gradient of 10% RH. Combing the pore structure and water properties of the excavated ivory, the ultra-depth-of-field 3D microscopy (KEYENCE, OSM, VHX-2000C, Japan) was conducted to observe the surface morphology of the excavated ivory during the water loss process to verify the process of water loss deterioration.

Results and discussion

Chemical composition and crystal structure analysis

The chemical composition and elemental content of the interior and outer surfaces of K7XY-181 and K7XY-61 (For example, K7XY-181in refers to the interior of K7XY-181, while K7XY-181out refers to the outer surface of K7XY-181) were studied using FTIR and XPS. As shown in Fig. 2a, The FTIR results show that the main chemical structure is the same on both the interior and outer surfaces of the excavated ivory. The broad absorption peak is observed at 3440 cm−1 corresponding to the stretching vibration of OH− [15]. The presence of characteristic peaks at 1413 and 1463 cm−1 is attributed to the substitution of CO32− for a minor portion of PO43− in the hydroxyapatite structure [16]. Furthermore, the peak at 874 cm−1 indicates a partial substitution of CO32− for OH− within the hydroxyapatite lattice [17]. The peaks at 1041 and 1105 cm−1 are assigned to the ν3 symmetric stretching vibration of PO43−, while the peaks at 563 and 603 cm−1 correspond to the ν4 bending vibration of PO43−. Lastly, the peak at 468 cm−1 is associated with the ν1 asymmetric stretching vibration of PO43− in the hydroxyapatite [18, 19]. Based on the detection of CO32−, PO43−, and OH−, it is speculated that the main component of the excavated ivory from the K7 site may be carbonated hydroxyapatite. Both K7XY-181 and K7XY-61 are mainly composed of C, O, P, and Ca elements (Fig. 2b, h). Notably, the content of O element exhibits the highest (59 ~ 65%). Furthermore, the content of C element is higher on the outer surface of the excavated ivory compared to its interior. This may be attributed to the outer surface being more exposed to the environment, thus undergoing a higher degree of mineralization [20]. The N content of K7XY-181 and K7XY-61 is significantly less than 1% (Fig. 2h). This indicates that the organic collagen fibrin component within the excavated ivory have undergone basically degradation after thousands of years of underground burial. As shown in Fig. 2c–g, the high-resolution XPS spectra of C, N, O, P, and Ca for K7XY-61in are presented. In the C 1 s spectrum of K7XY-61in, the binding energy of C–C (284.8 eV) is fixed and used as a reference [21]. The peak at a binding energy of 289.4 eV is attributed to CO32− [22]. For the N 1 s spectrum of K7XY-61in, due to the low nitrogen content of the sample, the signal-to-noise ratio is too low to decompose it. In the O 1 s spectrum of K7XY-61in, the inorganic salt components of CO32− and PO43− appear at 530.6 eV [23], while the Ca-OH bond in hydroxyapatite appears at 531.4 eV [24]. Tthe P 2p spectrum of K7XY-61in show an intense P 2p 3/2 peak at 133.0 eV and a P 2p 1/2 peak at 133.9 eV [25]. Similarly, in the Ca 2p spectrum of K7XY-61in, 346.6 and 350.0 eV correspond to the Ca 2p 3/2 and Ca 2p 1/2 peak [25], respectively. These results collectively indicate that the original collagen fibers of the excavated ivory from the K7 site have essentially degraded after a long period of underground burial, resulting in hydroxyapatite and carbonated hydroxyapatite as the primary constituents.

Fig. 2
figure 2

a FTIR, b typical XPS survey spectra, and (h) elemental atomic content of K7XY61 and K7XY181. c–g High-resolution XPS spectra of C, N, O, P and Ca of K7XY-61in

Full size image

Taking K7XY-61 as an example, the difference of element distribution of the excavated ivory was further verified by SEM–EDS [10, 26]. The elements of C, O, P, and Ca are widely distributed in both K7XY-61in and K7XY-61out (Fig. 3), which are the basic components of carbonated hydroxyapatite. However, the distribution of N element is relatively lower in both samples, indicating its lower content. The overall element distribution of K7XY-61in appears relatively uniform. By comparison, the distribution of Ca and P elements in K7XY-61out is more uneven. This uneven distribution may be ascribed not only to the surface irregularities of the samples but also to the potential effects of prolonged exposure of the external surface of the excavated ivory from the K7 site to environmental conditions.

Fig. 3
figure 3

SEM images and corresponding EDS spectra of (a) K7XY-61in and (b) K7XY-61out

Full size image

As shown in Fig. 4a, the main diffraction peaks of K7XY-181 and K7XY-61 are corresponding to hydroxyapatite [12]. Their main diffraction peaks are 25.9°, 28.9°, 31.8°, 32.9°, 39.7°, 46.7°, 49.5°, 53.2°, 64.0°, which correspond to the (002), (210), (211), (300), (310), (222), (213), (004), (304) crystal planes of hydroxyapatite standard card (JCDPS: PDF# 09–0432). The crystal lattice parameters of K7XY-181 and K7XY-61 are obtained through XRD Rietveld (Fig. 4b). Four sets of crystal lattice parameters are similar to the standard crystal lattice parameters of hydroxyapatite, conforming to the hexagonal lattice structure of P63/m hydroxyapatite [10]. The slight differences are mainly due to the partial substitution of CO32− for PO43− at the B site (B-type substitution) and OH− at the A site (A-type substitution) in the hydroxyapatite structure, respectively [27, 28]. Therefore, it is indicated that the components of the excavated ivory from the K7 site contain some carbonated hydroxyapatite. As shown in Fig. 4c, the crystal sizes of the four samples were calculated using the Scherrer Eq. (1) [29,30,31,32] through selecting the peak of the (002) crystal plane at 2θ = 25.9° that does not overlap with other peaks. The result shows that the crystal sizes of the four samples are 21.06, 22.39, 21.77, and 22.23 nm, respectively.

$${\mathbf{L}}\, = \frac{{\user2{K\lambda }}}{{{\varvec{\beta}}\cos {\varvec{\theta}}}}$$
(1)

where L represents the nanocrystal size. K is the shape factor, which is typically set to 0.89 for ceramic materials. λ is the radiation wavelength λCuKα = 0.15406 nm. β is the full width at half maximum (FWHM). And θ is the diffraction angle of the peak. As shown in Fig. 4d–g, K7XY-61 exhibits a mixed crystal structure with mainly needle-like and secondarily lamellar structure. Two types of lattice fringes with fringe spacings of 0.272 nm and 0.344 nm can be observed from the HRTEM image of K7XY-61 (Fig. 4h), which corresponding to the (300) and (002) crystal planes of hydroxyapatite, respectively. In Fig. 4i, SAED image of K7XY-61 is revealed diffraction rings corresponding to the (002), (300), and (004) crystal planes of hydroxyapatite [32, 33]. These results collectively indicate that the main components of the excavated ivory from the K7 site are hydroxyapatite and carbonated hydroxyapatite.

Fig. 4
figure 4

a XRD patterns, b cell parameters and (c) grain size of K7XY-181 and K7XY-61. d–g TEM images at different magnifications, h high-resolution TEM (HRTEM) image, and (i) selected area electron diffraction (SAED) patterns of K7XY-61

Full size image

Pore structure analysis

Ivory, being a typical natural organic–inorganic composite material, primarily comprises hydroxyapatite as its inorganic constituent and collagen fibrils as its organic component [2]. The organic collagen fibrils of the excavated ivory have been mostly degraded after thousands of years of diagenesis, while the hydroxyapatite has been retained. Consequently, the excavated ivory exhibits special characteristics such as fragility, dampness, porosity, and looseness [34,35,36]. Therefore, the micro-pore structure of the excavated ivory was comprehensively analyzed through dry methods (including SEM, BET, and MIP) and wet methods such as LF-NMR.

Combined with Fig. 5a, b, d, e, the dentinal tubules pore structure of the excavated ivory is slender and clearly visible. The dentinal tubules pores in the internal longitudinal profile of the excavated ivory are uniformly distributed and relatively well-preserved (Fig. 5a). In contrast, the dentinal tubules pores in the external longitudinal profile are more scattered and in a less-than-ideal state of preservation (Fig. 5d). There are numerous pores and a relatively concentrated pore size distribution on the internal cross-section of the excavated ivory, with an average pore size of 0.788 ± 0.278 µm (Fig. 5b, c). Conversely, the outer cross-section of the excavated ivory exhibits fewer pores, some collapsed pores and partial pores filled with dirt and other impurities (Fig. 5e). The average pore size on the outer cross-section is 1.038 ± 0.257 µm (Fig. 5f). It can be concluded that the pore size distribution of the excavated ivory is wide, ranging from hundreds of nanometers to several micrometers, and the average pore diameter of the external cross-section is larger than those of internal cross-section. The small number and more scattered distribution of pores on the outer cross-section may be mostly attributed to collapse and filling during burial of excavation [36,37,38,39]. The pore size distribution of micro- and mesopores in materials can be effectively characterized by BET method [40]. The pore structure and specific surface area results of two parallel samples of K7XY-61 (K7XY-61–1 and K7XY-61–2) as shown in Fig. 5g–h. Both K7XY-61–1 and K7XY-61–2 exhibit nitrogen adsorption–desorption curves that fall within the H3 subtype of type IV isothermal adsorption–desorption curves [41,42,43,44]. The adsorption curve and the desorption curve do not completely overlap, indicating capillary condensation, and an adsorption hysteresis loop appears in the middle section of the curve without a distinct saturation adsorption platform. This indicates that K7XY-61 belongs to the porous structural material dominated by mesopores, but the pore structure is irregular. The pore size distribution of K7XY-61 is primarily concentrated in the range of 1–120 nm, according to the density functional theory (DFT) method (Fig. 5h). Among them, the majority are mesopores (2–50 nm) and macropores (> 50 nm), with a very small proportion of micropores (< 2 nm). The specific surface areas of K7XY-61–1 and K7XY-61–2 are 122.29 m2/g and 113.74 m2/g, respectively, while their average pore sizes are 16.49 and 18.40 nm, respectively. Both the specific surface areas and average pore sizes are similar, which is consistent with the results obtained from the adsorption–desorption curves. The cylindrical model based on the MIP method exhibits higher precision in analyzing the macropores size [45, 46]. As shown in Fig. 5i, the pore structures of K7XY-61–1 and K7XY-61–2 are similar, mainly consisting of mesopores and macropores with a wide range of pore sizes, mainly distributed between 5 and 100 nm. There are pores with sizes greater than 200 nm, with the maximum pore size reaching 1000 nm. The porosity of K7XY-61–1 and K7XY-61–2 is 61.59 and 62.61%, respectively, with average pore sizes of 22.5 and 23.3 nm, respectively. In a nutshell, the results of the dry method porosity analyses reveal that the pore structure of the excavated ivory from the K7 site is predominantly mesopores and macropores, with subordinate micropores. The main pore size distribution range and porosity of K7XY-61 are 5–100 nm and approximately 62%, respectively, with an average pore size of approximately 20 nm.

Fig. 5
figure 5

SEM images and pore size normal distribution of K7XY-61: a internal longitudinal section, b internal cross-section, c internal cross-section pore size normal distribution, d external longitudinal profile, e external cross-section, f external cross-section pore size normal distribution. g Nitrogen adsorption–desorption curves, h DFT pore size distribution, BET specific surface area and average pore size, i mercury injection pore size distribution, porosity and average pore size of K7XY-61

Full size image

Dry method porosity analyses require the drying pre-treatment of the samples, which may impact the internal pores of the excavated ivory. Under the framework of “principle of minimal intervention” of cultural relic conservation, how to accurately present the true internal pore structure of the excavated water-logged ivory is worthy of deep consideration. LF-NMR possesses the advantages of high speed, accuracy, and non-destructive testing, and its application has gradually expanded from traditional fields such as bioscience and geological structure to the water-containing porous materials [47,48,49,50]. As shown in Fig. 6a, b, the pore structure of K7XY-61 is divided into two parts. The first part is the pore structure with a pore radius of 10–630 nm, accounting for a total of 96.72%, among which the pore radius range of 25–100 nm with the highest proportion of 60.53%. The second part is the cracks with a radius of 4–16 μm, accounting for only 3.28% of the total. As shown in Fig. 6c, the porosity of K7XY-61 is 59.88%. The LF-NMR pore analysis shows that the pore structure of the excavated ivory from the K7 site is mainly composed of macropores, followed by mesopores, and somes cracks.

Fig. 6
figure 6

LF-NMR analysis of K7XY-61: a distribution of pore radii, b statistical graph of pore radii, and (c) distribution of porosity

Full size image

The results of both dry and wet methods reveal a consistent trend towards a predominance of mesopores and macropores for the excavated ivory from the K7 site. However, notable differences in pore size distribution emerge of the excavated ivory under different testing methods, indicating a diverse pore structure and considerable variation in size. This observation underscores the complexity and heterogeneity of the pore architecture within the excavated ivory from the K7 site.

Water content and moisture states

The water content and moisture state of excavated ivory is one of the key factors affecting its preservation state. Consequently, it is imperative to conduct rigorous research on the moisture states of excavated ivory, as well as its water content across various environments, for the preservation and conservation. The dynamic water loss curves of K7XY-61 were monitored by gravimetric method at 25, 60, and 100 ℃ to gain insights into its water loss rate and water content. As shown in Fig. 7a, b, due to the loose structure of the excavated ivory, the higher the heating temperature, the faster the water loss rate of K7XY-61. Specifically, K7XY-61 takes 2 h, 4 h and potentially 24 h to reach a constant weight through water loss at 100, 60 and 25 ℃, respectively. The average water content of K7XY-61 calculated by gravimetric method is 36.78 ± 1.42%. This indicates that it is extremely easy to lose water in the air environment without any protective measures. The water content and moisture states within K7XY-61 were in-depth investigated by thermal analysis (TG and DSC). As shown in Fig. 7c, the TG curve of K7XY-61 exhibits one decomposition stage from room temperature to 125 ℃, which is the loss of internal free and adsorbed water. The weight loss rate of K7XY-61 at 125 ℃ is approximately 37.36%, corresponding to its water content. Additionally, the peak of the DTG curve occurs at 82.3 ℃, indicating that the content of free water within K7XY-61 is higher than that of adsorbed water [51]. As shown in Fig. 7d, the DSC curve exhibits a single endothermic peak at 70.3 ℃, which is typically associated with the evaporation of free water and adsorbed water [52]. This also indicates that the content of free water within K7XY-61 is higher than that of adsorbed water. LF-NMR is an advanced analytical technique allows for rapid and non-intrusive detection of water states [53, 54]. This method relies on the analysis of the transverse relaxation time (T2) of hydrogen ions in various states within a magnetic field, which directly correlates with the mobility of water molecules within a sample. Specifically, a higher T2 value signifies greater fluidity of water molecules, whereas a lower T2 value indicates restricted mobility. Furthermore, each distinct T2 relaxation peak is indicative of a unique moisture state, offering a detailed representation of the moisture states within the material being analyzed. [55, 56]. As shown in Fig. 7e, K7XY-61 exhibits two relaxation peaks (T21 and T22), corresponding to adsorbed water (1–100 ms) and free water (> 100 ms), respectively. Based on the peak area reflected by the amplitude of the T2 spectrum signal, and subsequently reflecting the water content in various states [57]. It is determined that the proportion of free and adsorbed water in K7XY-61 is 70.25 and 29.75%, respectively (Fig. 7f). In summary, the water content of the excavated ivory from the K7 site is approximately 37%. The majority of this water content consists of free water, while a minor portion is adsorbed water. To gain a comprehensive understanding of the specific role played by the internal water within the excavated ivory, further rigorous academic research is imperative.

Fig. 7
figure 7

a Water loss curves and (b) corresponding water content of K7XY-61 at 25, 60, and 100 ℃. c TG-DTG curves, d DSC curves, e LF-NMR T2 relaxation spectroscopy, and (f) different moisture states content of K7XY-61

Full size image

water loss degradation analysis

The deterioration of excavated ivory is closely related the loss of water. The DVS technique has been utilized in the development of water dynamic models for wood and cellulose materials. This method typically involves measuring the mass response of a sample over time following a step change in relative humidity (RH) [57,58,59,60]. To gain a deeper understanding of the negative impacts of water loss on the structural integrity of excavated ivory, the kinetics of water loss in three parallel samples of K7XY-61 (K7XY-61–1, K7XY-61–2, and K7XY-61–3) were conducted by DVS under constant temperature of 25 °C and varying humidity conditions. As shown in Fig. 8, the first stage is a gradient dehumidification process of ambient humidity, reducing the RH from 95 to 25%. Continuous water loss of K7XY-61 due to decreased humidity during this process, especially at the fastest rate within 24 h. The maximum weight loss rate of K7XY-61 during the entire decreased humidity process is between 35 and 40%, which is consistent with the water content determined through the gravimetric method. Subsequently, the second stage is a gradient humidification process of ambient humidity, increasing the RH from 25 to 95%. The sample eventually reached equilibrium and stabilized at 95% RH. Although K7XY-61 has slightly increased its weight by absorbing water from the environment during this period, it has significantly decreased its weight compared to the initial state. The results indicate that the dynamic process of dehumidification and humidification of the excavated ivory cannot be restored to the original condition. The loss of internal water (free and adsorbed water) for the excavated ivory under relatively dry conditions has led to the collapse of its internal structure, resulting in irreversible damage. It shows that the internal water of the excavated ivory has a structural support function.

Fig. 8
figure 8

a–c DVS curves and d–f weight change curves of three parallel samples of K7XY61 (K7XY-61–1, K7XY-61–2, and K7XY-61–3)

Full size image

Based on the fact that the fastest water loss rate of excavated ivory mainly occurs within 24 h, LF-NMR was performed on K7XY-61 with dense structure at 0, 12, and 24 h of water loss at room temperature and humidity (25 ℃, 65% RH) to further explore the relationship between deterioration, porosity and water of excavated ivory. There are two distinct T2 relaxation peaks of K7XY-61 at 0 h of water loss corresponding to free and adsorbed water (Fig. 9a), which accounted for 41.55 and 58.45% of total water (Fig. 9d), respectively. And K7XY-61 mainly contained pores smaller than 100 nm and dentinal tubules pore structures larger than 1 μm (Fig. 9b, c). With the rapid migration and volatilization of free water, when the water loss time is extended to 12 or 24 h, the T22 relaxation peak of free water disappears, while the relaxation time and intensity of T21 relaxation peak of adsorbed water gradually increase. This change may be due to changes in water distribution and the interaction of water molecules with the pore surface structure. In the process of water loss, with the evaporation of free water, the remaining water molecules may be transferred from the strong adsorption site to the relatively weak adsorption site, resulting in a slight increase in the degree of freedom of molecular motion, thereby prolonging the T2 relaxation time [61]. Correspondingly, the dentin tubular pore structures could not be detected and only the pores smaller than 100 nm were detected after 12 and 24 h of water loss, indicating that the free water had completely evaporated or been lost. This is consistent with the T21 relaxation peak in the LF-NMR spectrum where is only adsorbed water. The results showed that the free water mainly existed in the dentinal tubules or larger pores, while the adsorbed water mainly distributed in the pores smaller than 100 nm.

Fig. 9
figure 9

LF-NMR analysis of K7XY-61 at different periods (0, 12 and 24 h) of water loss under room temperature and humidity (25 ℃, 65% RH): a T2 relaxation spectroscopy, b distribution of pore radii, c distribution of porosity, and d different moisture states content

Full size image

The surface micro-morphologies of three parallel specimens of K7XY-61 were analyzed under simulated continuous water loss for different periods in environment conditions (25 ℃ and 64% RH) by Ultra-depth-of-field 3D microscope and gravimetry to directly reflect the influence of water loss on excavated ivory [62,63,64]. As shown in Fig. 10a, in the process of water loss, the rapid migration and volatilization of free water caused the original cracks to widen and new cracks to appear, the surface of excavated ivory to become dry and crumbly, and parts to fall off. This corresponds to the weightlessness curve shown in Fig. 10b. The results combined with DVS and LF-NMR showed that the rapid migration and volatilization of free water in dentinal tubules, intercalations and cracks under room temperature and humidity could cause serious deterioration of the excavated ivory. And the degree of water loss was positively correlated with the degree of deterioration. And the degree of water loss was positively correlated with the degree of deterioration. Moreover, if the adsorbed water existing in the pores ( < 100 nm) is not properly preserved, it may be further lost over time, thus aggravating the deterioration of the excavated ivory.

Fig. 10
figure 10

a Ultra-depth-of-field 3D microscope images of three parallel samples of K7XY-61 (K7XY-61–1, K7XY-61–2, and K7XY-61–3) for different periods of continuous water loss at 25 ℃ and 64% RH. b The weight-loss rate curve corresponding to the ultra-depth-of-field 3D microscope test

Full size image

The comprehensive study of the pore structure and internal water of the excavated ivory indicates that due to long burial in a damp underground environment, the ivory has absorbed a significant amount of water, resulting in a high water content state. Schreger bands formed by MCF periodic orientation or three-dimensional dentinal tubules arrangement in ivory structure easily lead to mechanical weakness in tangential section. Moreover, the degradation of the original collagen fibers during long-term underground burial further aggravated the deterioration of the ivory structure, resulting in numerous pores, interlayer gaps, and cracks in the excavated ivory except for dentinal tubules structure. As shown in Fig. 11, free water is located in the cracks, interlayer gaps, and dentin tubular pore structures of the excavated ivory, while absorbed water occupies the pores with a diameter less than 100 nm in the excavated ivory. These water components play a crucial role in excavated ivory. Among them, the formation of cracks is closely related to the rapid migratory and volatile of free water. This is because free water provides volume support, adhesion between layers, maintaining the moist state of the excavated ivory, and balancing the external environment. Absorbed water is adsorbed onto the inorganic mineral components (hydroxyapatite and carbonated hydroxyapatite) of the excavated ivory through intermolecular forces and hydrogen bonds, providing support within its pores.

Fig. 11
figure 11

Schematic diagram of the relationship between pores, cracks and water in excavated ivory

Full size image

Conclusions

This paper focuses on the microstructure and water function of excavated ivory from the K7 site of the Sanxingdui Ruins, delving into the correlation between the pore structure and moisture states. The main phase of the excavated ivory are hydroxyapatite and carbonated hydroxyapatite, which exhibits a mixed crystal structure with primarily needle-like crystals and secondary lamellae. Porosity analyses using both dry and wet methods show that the pore structure of the excavated ivory is primarily composed of macropores and mesopores, with secondary micropores and a small portion of cracks. The water content of the excavated ivory varies between 35 and 38%, consisting primarily of free water and adsorbed water. Free water is situated within the cracks, interlayer gaps and dentin tubular pore structures of excavated ivory, whereas adsorbed water resides in its pores (d < 100 nm). Free water provides support, adhesion, and environmental balance for excavated ivory. Its rapid migration and evaporation can influence the expansion of existing cracks and the formation of new ones. Free water performs crucial functions such as providing support, bonding, and balancing the environment of the excavated ivory. Conversely, adsorbed water primarily serves as a structural support for the pores. Any loss of internal water in the excavated ivory can result in the loss of these vital functions, ultimately leading to irreversible damage. This finding offers valuable insights and a scientific basis for the effective long-term reinforcement and preservation of the excavated ivory, providing an idea for the scientific in-situ conservation of cultural artifacts.