Introduction

In paleontology, silicification is a natural process in which organic materials are replaced by silica1,2,3. It was primarily understood as the result of the infiltration of dissolved silica into wood fossils, which over time, undergoes silica phase changes from amorphous silica, such as opal, to higher-phase crystals, such as quartz4,5,6,7,8,9,10. Wood silicification has also been recognized worldwide in various fossil records11,12,13,14,15. Previous research also indicates that organic material is seemingly replaced by silicate minerals1,13,16,17,18. To date, this process is believed to preserve the original structure of wood anatomy, allowing them to remain intact for millions of years1,3,14,19,20. Silicification is thought to play a crucial role in naturally preserving wood fossils and preventing decomposition2,13,21,22. Therefore, it can be considered a form of “natural sealing”.

Based on their anatomical structure, wood can be categorized according to its pore structure. These pore structures are the fluid transport components of living trees. Fluid transport in gymnosperms is carried out by tracheid cells, while in angiosperms, these flows are facilitated by vessel components23,24. These vessels are typically encountered in Dipterocarpaceae trees as angiosperms, which are easily recognizable in their microstructure25. The porous characteristics of woods conceivably facilitated the infiltration and accumulation of dissolved silica in the internal structure of the wood over time during fossilization. However, more detailed research is still needed to understand the geochemical proportions in wood fossils and their interaction with the host rock, the implications on silica phase proportions, the availability of organic residues in the host rock, and evidence of silicification at the tissue level. Despite its importance, such studies are relatively rare, yet they hold the key to comprehending the silicification process within organic tissues. Thus, further studies on wood fossils are necessary to provide a more detailed perspective, at least at the micron level.

Our research employs microanalytical approaches to acquire better paleontological perspectives on wood fossilization, notably focusing on the wood silicification. The objective of the study was to examine the roles of the silicification process of wood fossils through microanalysis. Specifically, we compare the proportions of oxide compositions and trace elements between wood fossils and the host rock. Additionally, we determine the crystallization and silica phase proportions in silicified wood, track the presence of organic residues, and perform observations and element mapping of silicified wood at the micron scale. This study utilizes microanalytical techniques, including X-ray Fluorescence (XRF), inductively coupled plasma mass spectrometry (ICP-MS), X-ray Diffraction (XRD), Fourier Transform Infrared (FT-IR) spectroscopy, and Field Emission-Electron Probe Microanalysis (FE-EPMA).

Results

Geological mapping

In the Jasinga region, West Java, the Genteng Formation is characterized by rock orientations to the northwest (NW) and northeast (NE) dip directions. This formation can be traced along the Ciherang River transect from the southern to the southwest region (see Fig. 1). The transect area for this study was along the western side of the dextral strike-slip fault (N214°E/83 NW). Observations and measurements of the lithostratigraphic profiles were primarily conducted on the west of the fault line to avoid errors in determining the rock sequences. Most fossil wood in this area is associated with host rocks.

Figure 1
figure 1

Field documentation (left), geological map (mid), and lithostratigraphic profile (right) of the study area in Jasinga, West Java. Red arrows in the field documentation (left) show the presence of wood fossils in the host rock. The number of wood fossil symbols represents the samples collected for this study. The coordinates of the object are also mentioned in the UTM (48S) format. The map on this figure created using Licensed Desktop ArcGIS 10.x (980940/Institut Teknologi Sumatera), URL: www.esri.com.

Full size image

The Genteng Formation in the study area encompasses three rock units (oldest to youngest): The Breccia-laharic rock unit, the Sandstone rock unit, and the Conglomeratic rock unit. All rock units tend to have a northeast (NE) dip direction. The fossil wood taxa in this transect mainly belong to the Dipterocarpaceae family.

The composition of the wood fossils and its host rock

The silicification of wood fossils has been confirmed through XRF and ICP-MS analysis (Table 1). SiO2 is the dominant component, with concentrations ranging from 91.86% to 92.79% w/w. Other components tend to be much lower and typically do not exceed 5%, such as Fe (1.55% to 4.47%), Al2O2 (1.60% to 1.77%), CaO (0.35% to 2.41%), P2O3 (0.72% to 0.74%), and K2O (0.05% to 0.19%). The MgO content in these wood fossils tends to be below 0.01%, while compared to their host rocks, the MgO content in the host rocks ranges from 0.25% to 0.79%. Additionally, sodium (Na) found in the fossilized wood falls within the range of 304 to 112 ppm, while trace elements such as lanthanides do not exceed 10.0 ppm. Besides silica, the consentration of these geochemical components tends to be lower when compared to the concentrations in the host rocks. This indicates that wood fossils tend to serve as media for Si-enrichment, considering the initial material of wood fossils comes from organic carbon material, not silica.

Table 1 Composition of wood fossils and its host-rock in different rock strata, Jasinga, Java Island.
Full size table

The characteristics of geochemical composition between wood fossils and host rocks in this study are evaluated through composition ratios between wood fossil (Fo) and host rock (Hr). These ratios were calculated using the weight values (wt.%) of each calculated composition. The Fo/Hr value represents the ratio of geochemical concentration w/w between the fossil specimen and its host rock, write in decimal value. Fo/Hr > 1.00 indicates a richer concentration of components in the wood fossil, while Fo/Hr < 1.00 indicates the opposite. A value of Fo/Hr = 1.00 indicates balance proportion between the two. The results show that the SiO2 in fossilized wood is significantly more abundant than in its host rock (Fo:Hr = 1.36 to 1.89). This value indicates that the silica composition in fossilized wood is 36% to 89% more abundant than in the host rock. Interestingly, the same condition occurs for sodium (Na) where Fo/Hr = 8.72 to 25.35, indicating that Na in the silicified wood is 11 to 25 times than in the host rock. This suggests that not only silica but also sodium undergoes enrichment along with the silicification process. Phosphate (P2O5) appears tends to maintain a balanced proportion between the fossil and host rocks. Meanwhile, other geochemical components tend to be lower in silicified wood compared to the host rock, indicating that components such as iron, aluminum, and calcium are not enriched in silicified wood.

Correlation analysis was conducted to examine the straightforward relationship between the chemical proportions of fossilized wood and its host rock. The correlation results are presented in a matrix table (Table 2); each value represents the correlation coefficient (r). This correlation matrix shows that the concentration of SiO2 in the host rock positively controls the abundance of SiO2 in fossilized wood (r =  + 0.82). Conversely, the total iron (as FeO) content in the host rock significantly negatively affected the abundance of SiO2 in the fossilized wood (r = − 0.72). These results indicate that SiO2 concentrations in the host rock are related to the potential Si-enrichment in the fossilized wood. In contrast, the increased total iron in the host rock decreased the SiO2 concentration in the fossilized wood. Based on this, it can be concluded that the degree of silicification in fossilized wood is strongly influenced by the high concentration of SiO2 and low concentration of total iron in the host rock. Furthermore, the total iron content in the fossilized wood appears to be influenced by the abundance of total iron in the host rock (r =  + 0.61), as well as the low enrichment of P2O5 (r =  − 0.82) and SiO2 (r =  − 0.74) in the fossilized wood.

Table 2 Matrix correlation of fossil composition vs. fossil and host rock vs. fossil.
Full size table

The correlation coefficient (r) among the samples (wood fossil vs. host rock) indicates a relationship between the abundance of SiO2 in the host rock and the abundance of SiO2 in the fossil wood, as demonstrated by the correlation between SiO2 in wood fossils and SiO2 in the host rock (r =  + 0.82). Biologically, silica in woody plants is rarely present, at around 0.5 to 2.0% wt26,27,28,29. However, the results demonstrate that the analyzed wood fossils contain 92.79 ± 2.032% SiO2, indicating that the analyzed wood fossils have undergone carbon-to-silica replacement, whereas the silicas have replaced the original organic component of these woods during fossilization1,30. The major source of silica in wood fossils likely originates from the silica present in the sediments during its early burial process, representing its host rock characteristics. Some lithologies of rocks containing wood fossils are tuffaceous sedimentary rocks associated with a paleovolcanic environment rich in silica (Fig. 1). Silica from the surrounding environment can dissolve and infiltrate the wood body, filling the vessel of wood. Over time, silica accumulation becomes mineralized into silicate minerals during the lithification and rock diagenesis processes, as the temperature and pressure gradually increase. The abundance of silica in fossil wood also seems to correlate with the decreasing composition of other components, such as organic materials, potassium, and iron, which are replaced by silica.

Silica more easily infiltrated the wood structure in the dissolved state. In this dissolved state, silica fills micron-sized spaces within the wood structure. The solubility of silica tends to be higher under alkaline conditions (high pH) and begins to precipitate at low pH31,32. The alkaline conditions of the depositional environment during the early burial stage provided a sufficient amount of dissolved silica. This surface process controls the characteristics of the host rocks of wood fossils. As the silica infiltrates the decaying wood, the dissolved silica tends to become immobile due to the acidic conditions of the decaying wood. This acidic condition may trigger the accumulation of silica within the pores wood. However, silica only mineralizes once adequate temperature and pressure are reached, which are required to form minerals such as quartz. When the intensity of the burial process increased, the temperature and pressure began to increase, followed by the sediment lithification and diagenesis processes. This stage causes the silica within the pores wood to form more mature silica phases and well-oriented silicate minerals. Under these conditions, other dissolved silica may re-infiltrate the wood structure from the surrounding rock but at higher pressures and temperatures, deep from the surface, due to several geological processes such as deep burial and tectonics events. The silica within the wood structure mineralizes and gradually replaces the previous composition of the wood organic material simultaneously with the organic decomposition process.

Proportion of silica phase in silicified wood

Our study demonstrates that X-ray Diffraction (XRD) analysis indicated that quartz was the predominant silica phase in the silicified wood of Jasinga (Fig. 2). This examination revealed variations of silica phases across wood fossil specimens from different geological strata. XRD data were collected from specimens obtained from the lower (B-3/G54), middle (C1/G28 and C-5/G55), and upper (D-2/G43) strata. Wood fossil samples were collected from different lithologies to understand the role of the host rock lithology in wood silicification. The XRD analysis showed 19 peaks for quartz ranging from 5° to 90° at 2θ, with the peak at 26.7° showing the highest intensity compared to the other peaks, indicating the presence of quartz. The peaks at 19.6° (629 cps) and 20.9° (4240 cps) suggest the presence of a small amount of amorphous silica, particularly opal-CT (opal-cristobalite-trydimite). Opal-CT was significantly higher in wood fossil B-3/G54 from the tuffaceous breccia in the lower strata. The results indicated that the crystallinity of wood fossils from conglomerate beds was notably lower than that of wood fossils from other beds. In this lithology, the quartz abundance in wood fossils tended to be lower than that in other lithologies. As shown in Table 3, wood fossils within the tufaceous layers tend to have high crystallinity (> 70%) and a much more abundant presence of quartz. Furthermore, when comparing the characteristics of the grain size of its host rocks, it is clear that the intensity of quartz is higher in finer grains of host rock such as sandstone than in larger grains of host rock such as breccia and conglomerate. These results demonstrate that lithology plays a significant role in wood silicification.

Figure 2
figure 2

The XRD spectra of the wood fossils from Jasinga, Java Island, along with their comparison to the spectra of other silica phases, also include the values of the 2-θ degree and relative peak height (cps) for each peak listed in the specimen.

Full size image
Table 3 Comparison of crystallinity and quartz intensity and presence of amorphous silica of silicified wood from different rock strata of Jasinga, Java Island.
Full size table

It can be observed that wood fossils found in tuffaceous rocks (C-5/G55, C-1/G28, and B-3/G54) exhibit higher crystallinity compared to those found in non-tuffaceous rocks. The characteristics of tuffaceous rocks significantly influence the crystallinity of silicate minerals in wood fossils, as these rocks contain an abundance of tuff grains rich in silica. Tuff, a type of rock that originates from volcanic ash ejected during volcanic eruptions, consists of silica-rich grains. These grains are highly likely to have been deposited simultaneously during wood burial, thus acting as the primary source of silica for wood silicification. Tuff and other volcanic materials are good sources of dissolved silica in depositional environments because they contain silica in the form of volcanic glass. However, in the upper strata composed of conglomerate rocks, although silica-containing grains are likely present in siliciclastics, they do not effectively supply silica sources to wood fossils. Siliciclastic conglomerates usually consist of more mature grains with higher material resistance, making the silica within them difficult to dissolve. However, the source of silica from this type of lithology could be related to the presence of silicate as intergranular cement in the host rock, which forms due to the precipitation of dissolved silica in intergranular pores during sediment lithification.

This phenomenon is associated with the comparison of quartz intensities among wood fossils. Fossils found in rocks with fine grains, such as sandstone, tend to exhibit a higher quartz intensity (ranging from 43,952 to 41,635 cps) compared to those found in rocks with larger grains, such as conglomerate and breccia (ranging from 13,247 to 22,962 cps). This result reaffirms that the lithological characteristics, particularly the texture and composition, affected wood silicification. These factors are related to the formation of porosity and permeability, where the infiltration significantly depends on both. Additionally, the presence of opal-CT in wood fossils in Jasinga remains to be determined. Opal-CT appears in older wood fossils, indicating that these fossils underwent later recrystallization in the pores due to the brittle conditions formed. Older wood fossils tend to be more brittle, especially with extended geological dynamics. Deeper burial positions allow for a more intense presence of hydrothermal veins, where dissolved silica can precipitate and settle within.

Non-crystaline material in silicified wood

In this study, in addition to the Si–O bonds, we successfully identified residues of organic material in the form of spectral carbon bonds using Fourier Transform Infrared (FT-IR) analysis (Fig. 3). FT-IR analysis produces specific spectra when the infrared wave number directly interacts with tracked atomic bonds. In this study, fossil wood specimens from different strata were compared with the FT-IR results from previous studies33,34,35,36.

Figure 3
figure 3

FT-IR spectra of wood fossils from different rock strata in Jasinga, Java Island.

Full size image

Spectral peaks appeared at wavenumbers of 458 cm–1, 693 cm–1, 778 cm–1, and 1070 cm–1. Spectral peaks at 458 cm–1 and 693 cm–1 indicate the abundance of Si–O bonds in quartz minerals. Meanwhile, the peaks at 778 cm–1 and 1070 cm–1 indicate Si–O bonds from tridymite minerals. The FT-IR results in this study confirmed the interpretation of the XRD results, where not only quartz but also silica can be found as other silica phases. Furthermore, several spectral peaks were observed at wavenumbers ranging from 1500 to 4000 cm–1, indicating the presence of carbon bonds characteristic of hydroxyl (–OH), methyl (–CH2), and carbon chains. These characteristics were observed in the sample at wave numbers 1584 to 1637 cm–1, 1876 cm–1, 2372 cm–1, 2972 cm–1, and 3408-3631 cm–1. These results are related to the FT-IR of the cellulose and lignin molecules. A comparison of fossil wood samples from different age strata showed a relative trend of increasing silica bond intensity from younger to older samples (Fig. 3). Meanwhile, the spectral peaks characteristic of organic bonds tended to decrease from the younger to older sample comparisons.

These results suggest that, while most silicified wood specimens are predominantly composed of silica, remnants of the original organic wood can still be detected. Wood silicification does not instantaneously convert the organic material into silica1,37. Instead, it is a gradual process where silica simultaneously replaces the organic mass of the wood. In the context of fossilization, this process is known as replacement. It involves dissolving the original organic structure and substituting gradually with others, including minerals like silica. As wood undergoes fossilization, the organic mass gradually diminishes throughout the silicification process, leading to the replacement of certain parts of the wood structure with silicate minerals.

Element mapping on fossilized wood structure

We selected specimens representing the interaction zone between wood fossils and their host rock. In this section, we observed interactions between the host rock and the wood fossil, particularly regarding Si-enrichment, and represents a series of compositional changes from predominantly carbon-based wood materials to silica-rich fossils. The body of the wood fossil and its host rock can be distinguished in Secondary Electron Image (SEI) and Backscattered Electron Image (BEI). Remnants of wood structure were visible, as indicated by the structure of wood fossil vessels from a cross-sectional view (Fig. 4). These vessels are characterized by elliptical and solitary shapes. The vessels appear parallel and in line with the rays of cells flanking them on either side. Meanwhile, sedimentary grains with angular shapes and poor sorting were visible in the host rock section. The body of the wood fossil and host rock can be differentiated from the BEI.

Figure 4
figure 4

Comparison of secondary electron image (left) and backscattered electron image (right) of wood fossil from Jasinga, Java Island. The images were captured at the boundary where the wood fossil and host rock meet and interact (yellow dotted lines).

Full size image

We selected sections of the sample to measure the intensity of silica and carbon directly through the transect lines. The EPMA results showed a sharp contrast between the body of the wood fossil and its host rock. Carbon can still be found in silicified wood (Fig. 5). EPMA imaging successfully located evidence of silica infiltration into the wood vessels. Silica seemingly infiltrates the internal structure of the wood through the wood vessel, which is directly exposed to the host rock. Further imaginary process also indicate that the silicon content is negatively correlated with the carbon intensity in a sigmoid pattern (Fig. 6). Otherwise, Elemental mapping of iron and Dy shows that these elements are also notable in the wood tissue through fractures (veins) of the host rock. Both metal elements appeared to fill the interconnected gaps between the parenchyma cells (Fig. 7). We have measured the total Light-REE (LREE) and Heavy-REE (HREE). An intensity analysis was also conducted to compare the distribution of LREE and HREE in wood fossils (Fig. 8). REE can infiltrate wood fossils through veins directly connecting the host rock and wood fossil. Geologically, veins are formed when a dilated fracture in a rock is filled with mineral constituents38,39. These minerals are transported by an aqueous solution within the rock and deposited through precipitation. The hydraulic flow facilitating this process is typically a result of hydrothermal activity. These activities are driven by increased temperature and regional geological pressure during rock diagenesis. These veins appear to carry rare earth elements (REEs) and fill the internal structure of the wood, particularly in the parenchyma tissue. These results suggest that the transformation of organic wood material into a fossil is heavily influenced by the host rock. By this case, wood acts as a sponge and absorbs certain geochemical elements from its host rock.

Figure 5
figure 5

Comparison of EPMA imaging probes for carbon (left) and silica (right). The images reveal silica infiltration from the host rock into wood vessels (red arrows). Some vessels remain in carbon (e.g., vessel A), while others have already been infiltrated by silica (e.g., vessels B and C). The white dotted lines represent the boundary between the wood fossil and its host rock.

Full size image
Figure 6
figure 6

Results of normalized difference index analysis between carbon (black) and silica (red) elements, index values presented in a color bar, along with regression outcomes on intensity values of the line transect (yellow line A-B) in the observed transition zone between the fossilized wood structure and its host rock.

Full size image
Figure 7
figure 7

Comparison of EPMA imaging probes for iron (left) and dysprosium (right). The images reveal these elements infiltrating the wood fossil bodies (yellow arrows) through veins that directly connect the host rock and wood fossil bodies (white arrows). The white dotted lines represent the boundary between the wood fossil and its host rock.

Full size image
Figure 8
figure 8

Regression results of LREE and HREE intensities (right), along with visualization of the total REE distribution (left) in the transition zone between the fossilized wood body and the fossilization substrate.

Full size image

Discussion

The Genteng Formation was deposited in the western part of Java. This formation lies on the Miocene Bojongmanik Formation as the angular unconformity40,41,42. This is associated with the tectonical event in back-arc basin of Java that occurred during the Pliocene43,44. The Genteng Formation can be distinguished by its lithological characteristics, including tuffaceous sandstone, tuffaceous siltstone, conglomerate, and breccia, which mostly contain fossilized woods. The wood fossils in this formation show intensive silicification. This study reveals that the silicification process occurring in wood fossils is closely linked to the composition of the host rock. The term “host rock” refers to a layer of rock containing fossilized wood that serves as a substrate for fossilization. The study confirms that wood fossilization within this formation significantly alters the organic composition of wood into inorganic matter. This process occurs concurrently with sedimentation and lithification of the host rock. The presence of veins in rocks associated with wood fossils suggests that silicification occurred during lithification of the rock. Veins cannot form during the initial stages of sedimentation and brittle conditions are required for their formation39,45. In this study, through elemental mapping analysis, we found evidence of silica-rich veins directly connected from the host rock to wood fossil, facilitating the permineralization process. Upon comparing the composition of the fossils with that of the host rock, it is clear that wood fossils tend to accumulate SiO2. Compared with other oxides, the SiO2 content in the wood fossil body is more abundant than in the host rock. The abundance of SiO2 in these wood fossils exceeds 90%, replacing carbon as the previous main component.

The wood specimens used in this study were fossils of the Dipterocarpaceae tree. Dipterocarpaceae is one of the hardwoods characterized by porous wood features46,47. Anatomically, porous wood is abundant with vessels48. Vessels are tubular structures that, in their life state, play a role in transporting water and nutrients throughout the tree23,25,49. Additionally, these woods have several thick-walled cells, such as fibers, parenchyma, and ray cells, which are rich in lignin, making them more resistant to decomposition50,51. The weathering process could also form cracks in the wood, facilitating infiltration more intensively. In the presence of pore structures, such as Dipterocarpaceae wood, silica accumulation can occur throughout the buried wood, even to the lumens of dead wood cells, replacing the organic components of the wood structure. This silica enrichment process must have occurred before the entire organic wood structure decomposed, especially to form silicified wood fossils entirely.

Silicified wood typically contains SiO2, which is predominantly composed of quartz. We also detected the presence of opal-CT, but its significance is less compared to quartz. The lithological characteristics of the host rock, such as its grain size and composition, significantly influence the crystallinity and intensity of quartz in wood fossils. For instance, silicified wood in tuffaceous rock exhibits higher crystallinity than in non-tuffaceous rock. The porosity and permeability of the host rock play a crucial role in the infiltration of dissolved silica, thereby affecting its precipitation in wood fossils. Rocks with finer grain sizes, such as sandstone, facilitate capillary pressure in rock porosity more effectively52,53 than conglomerates or breccias. Conglomerates and breccia, characterized by non-uniform grains and interlocking textures, tend to inhibit the infiltration of silica into the wood body due to their lithological characteristics, which have less porosity and poor permeability.

Silica can become enriched within wood tissue in a dissolved state, especially during early burial. This dissolved state allows silica to fill both the lumen of the cells and spaces within the wood structure2,3. Alkaline conditions are required to maintain higher solubility of silica at surface temperatures, particularly in depositional environments. For silica to precipitate, low pH conditions are required, causing silica to tend to be immobile6,32,54. Decaying wood provides a suitable medium for silica precipitation because many organic components decompose and release organic acids under such conditions. In the next stage, the precipitation of silica in the wood body gradually crystallizes into silica crystals as disoriented lattice crystals in the form of amorphous (SiO2.nH2O)55. However, this silica precipitation must be able to fill the pores of the wood before the organic structure of the wood decomposes completely. Next, while certain conditions that play a role in the diagenesis of the host rock occur, the same conditions also affect the form of silica inside wood pores. At this stage, the amorphous silica in the wood body gradually began to dehydration (removal of H2O), allowing silica and oxygen to form well-oriented Si–O crystals, which later became quartz minerals (SiO2), as already identified in our study. Silica crystals can recrystallize into various forms, as demonstrated in the well-known silica phase diagram56,57,58 where in this case, the local geological setting could also control the pressure and temperature conditions of these processes. Some researchers have successfully identified the presence of moganite, tridymite, and cristobalite in wood fossils with various local geological setting15,30,33,59,60. In wood fossils, moganite can be found in chalcedony3,61,62, while tridymite and cristobalite are frequently recognized alongside opal, known as opal-CT (opal-cristobalite-tridymite)63,64. The presence of opal-CT indicates the simultaneous presence of tridymite and cristobalite, particularly as poorly oriented crystals. Stratigraphically, higher temperatures and pressures could be achieved from the burial depth during the lithification of the rock strata65,66. Along with the duration of burial on the geological time scale, the depth of burial, sub-surface temperature, and regional pressure gradually affect the recrystallization of silica in the host rocks, including any wood fossils within them. An appropriate source of silica from the surrounding rock grains could contribute to the successful process of wood silicification, allowing it to be preserved as a fossil.

The deeper the burial depth, the higher are the temperatures and pressures reached in the rock layers67,68,69. Under these conditions, mineral recrystallization occurs as part of the rock diagenesis. It is highly possible that wood fossils have undergone similar processes. In this context, the silica deposited in the wood fossil body undergoes recrystallization into higher phases. Silica can enter the wood fossil body during sedimentation in its depositional environment21,22, where dissolved silica infiltrates the pores and precipitates into SiO2.nH2O. The SiO2.nH2O in the wood vessels served as the precursor for the recrystallization of silica phases during rock diagenesis. The more abundant of sodium in the silicified wood, compared to the host rock, indicates that the environment enriched with silicon is related to alkaline conditions16,22,70,71.

These silica phase variations were significantly influenced by burial temperature and pressure conditions56,57,58. In nature, silica exists in the soluble form as H3SiO4−. The solubility of H3SiO4- depends on the pH72; an alkaline condition would make this silica more soluble54,73; however, under acidic conditions, it tends to precipitate. These ions can be transported alongside the water flow and infiltrate sedimentary beds sourced from silicate minerals or hydrothermal fluid weathering22,74. Silicate solutions precipitate on porous substrates during sedimentation22,71,74. These silicate solutions precipitate under acidic conditions and form SiO2.nH2O. This sediment layer would be buried deeper during the diagenesis process to form a sedimentary rock. The increased temperature and pressure in the diagenesis process include amorphous silica or Opal16,63,75. Some of these may have been associated with cristobalite and tridymite at higher temperatures and pressures, leading to the formation of opal-CT (opal-cristobalite-tridymite). Hydrated silica releases .nH2O as increasing in pressure and temperature58. These conditions culminate in the recrystallization of SiO2 to a higher phase, such as quartz. However, they can be distinguished using XRD and FT-IR techniques34,76. Tridimite and cristobalite were the higher silica phases, which formed only under higher temperature and pressure conditions56,77,78. The transformation of SiO2.nH2O to SiO2 involves dehydration, where H2O and OH molecules are separated from the silicate crystals. Therefore, in this study, we suggest that the abundance of -OH and -n(H2O) groups in silicified wood can be used as indicators of the maturation level of the silicification process, with higher degrees of silicification resulting in lower proportions of these molecular groups.

Understanding the compositional changes that occur in fossils at the cellular level is essential. Through FE-EPMA analysis, it is noticeable that wood silicification begins with the infiltration of silica into the wood structure. Silica tends to enter the vessel portions of wood and gradually replace carbon. However, it is not just silica; under specific conditions, metallic elements such as Fe and Dy can also infiltrate the wood body through fractures that connect the wood fossils and the host rock. Silica, iron, and dysprosium are not dominant in living vegetation. However, a series of fossilization processes allows them to infiltrate wood fossils because of direct interactions between the host rock and wood fossils. Silica minerals within wood vessels provide natural preservation over long periods in the geological timescale. Numerous studies on fossilized wood have revealed that some fossil within silicified wood retains its original form, unlike what typically occurs in carbonaceous wood fossils18,37,79. This difference arises because silica recrystallization occurs within the lumen of wood cells1,3,80. As a result, the hardness characteristics of silica render wood fossils more resistant to physical, chemical, and biological weathering. Consequently, this wood fossil can last millions of years.

In conclusion, this study demonstrated that microanalytical approaches for silicified wood fossils can reveal the phenomenon of silicification in burial depth history. This approach revealed Si-enrichment in silicified wood under alkaline conditions. The degree of silicification in fossilized wood is controlled by the lithology of the host rock. The sedimentary texture is responsible for the mechanism of silica infiltration, which is due to the capillarity of the host rock to the wood. Microanalysis studies are necessary to investigate the geochemical interactions between the host rock and wood fossils. Silica infiltration, the remnants of carbon, and veins which involve in permineralization can be traced using this method. The microanalytical approach in paleontological studies has provided new perspectives on silicification. Through this approach, researchers have better understood how organic materials become mineralized and preserved as geological records. We hope that this study encourages researchers not only to examine fossils for identification but also to interpretate the fossilization processes.

Methods

Field works

The samples used in this study were obtained from the Genteng Formation, Java Island, Indonesia. Wood fossil samples (n = 27) were collected directly from rock layers with known lithostratigraphic positions in the Jasinga region of West Java, specifically along the Ciherang River transect. The samples were then sorted from the oldest to the youngest rock strata. The sizes of the wood fossils found in the host rock vary; however, in this study, we selected samples of wood fossils that still resembled, with a minimum size of 10 × 10 × 10 cm. Some of the samples are documented, as presented in Fig. 1. This size was considered sufficient to meet the requirements for further analysis. The positioning of the sample findings was taken along with the rocks adhering to the fossil, which were cleaned for further analysis. This stage aims to obtain a representative sample from each stratum and provide an appropriate number of samples for further analysis. We preferred to avoid over-collecting wood fossil samples in the research area, considering the limited availability of wood fossils in the region. The number of collected wood fossil samples in each rock layer is represented by the number of symbols in the lithostratigraphic profile.

X-ray fluorescence (XRF)

XRF are used to determine the proportion of elements and oxides in wood fossils and their host rocks. The samples were powdered to a particle size of 200 mesh.The obtained powder was subsampled using the coning and quartering sampling methods. The 200-mesh powdered samples were analyzed using a Malvern PANalytical Epsilon 3XLE X-Ray Spectrometer at the Mineral Laboratory, National Research and Innovation Agency (BRIN) in Lampung, Indonesia. This instrument detects elements and oxides at concentrations greater than 0.01%.

Inductively coupled plasma mass spectrometry (ICP-MS)

For geochemical composition with concentrations below 0.01%, samples were tested using an iCAP-Q Quadrupole ICP-MS instrument. This procedure followed the chemical testing standards and was conducted at the Center for Geological Survey of Indonesia (PSG) in Bandung, Indonesia. The samples were dried at room temperature, then grind to obtain a particle size of 200 mesh. Each sample (0.1 g) was dissolved in ultrapure 60% HNO3, 40% HF, and 60% HClO4. The mixture was then heated to 200 °C prior to evaporation. An additional 5 ml of ultrapure 60% HNO3 was added, and the mixture was heated to 150 °C. The solution was cooled to room temperature and diluted to 50 ml using ultrapure water. To this solution, 1 ml of a 0.1 mL 1 ppm Be solution and 0.1 mL 1 ppm Bi solution were added and labeled with 2% HNO3. The proportions of each element and oxide were analyzed to determine the ratio between the wood fossils and their host rocks and to calculate the correlation coefficient.

X-ray diffraction (XRD)

X-ray Diffraction (XRD) was performed to determine the crystallization, presence, and intensity of minerals in the wood fossils. This analysis was conducted at Greenlabs Research and Development Laboratory in Bandung, Indonesia using the Miniflex 6th generation Benchtop Powder X-ray Diffraction Instrument. The XRD settings utilized a Cu-Kα radiation source, covering a 2θ range of 2° to 90°. The powdered specimen (200 mesh) to be analyzed was subsampled using the coning and quartering sampling method for the analysis of the XRD instrument. XRD results are presented in. xy, .raw, and .asc formats. The spectral XRD database was sourced from the Crystallography Open Database (http://www.crystallography.net/) and RRUFF Project Database (https://rruff.info/). The XRD spectra were subsequently analyzed using Profex open-source software.

Fourier transform infrared (FT-IR)

To detect various chemical bonds in wood fossils, FTIR was employed. This approach enables the identification of organic molecule remains and inorganic compounds. FT-IR analysis was conducted at the Department of Advanced Research Infrastructure for Material and Nanotechnology, Hokkaido University, Sapporo, Japan, using the IRPrestige 21 Shimadzu instrument. The powdered specimen (200 mesh) to be analyzed was subsampled using the coning and quartering sampling methods. The samples were mixed with potassium bromide (KBr) powder in a test cup, and then exposed to infrared radiation ranging from wavenumber 400 to 4000 cm−1 using the IRsolution software. The resulting spectral data were used to determine the peak values, and were compared with relevant references representing the organic compounds found in wood and other petrified objects.

Field emission-electron probe micro-analyzer (FE-EPMA)

It was used to detect the distribution and proportions of elements in fossil wood at the micron tissue scale, particularly the fossil body part that directly interacts with the host rock. The instrument employed for this analysis was JXA-8530 F by JEOL, located in the Division of Sustainable Resources of Engineering, Hokkaido University, Sapporo, Japan. Secondary Electron Images (SEI) and Backscattered Electron Images (BEI) were captured in the focal area of the sample for field-emission observation. It uses solid samples that have been cut and polished to a smooth surface with dimensions ranging from less than 4 × 4 × 4 cm. Subsequently, spectral extraction was performed on the selected elements. The acquired images were digitized into pixel values to determine the spatial position and intensity of the spectral signals within the specimen. This was achieved by using an open-source image-processing calculator available online at dcode.fr. The digital value per pixel (pixel value) represents the intensity of the analyzed elements. Data processing and presentation were conducted using Microsoft Excel for normalized index calculations.