Understanding the corrosion mechanism of iron artifacts using mössbauer spectroscopy | Scientific Reports

Mar 25, 2025

Scientific Reports volume 15, Article number: 10207 (2025) Cite this article

Metrics details

Iron artifacts undergo complex corrosion processes, depending on the burial environment. Understanding the formation mechanism of corrosion products is crucial for preservation of artifacts and helps design strategies for future iron artifacts protection. Mössbauer spectroscopy was primarily utilized in this work to analyze the corrosion products formed on iron artifacts. The corrosion products were identified as consisting of goethite, lepidocrocite, magnetite, and maghemite. Low-temperature Mössbauer spectroscopy was performed for the accurate identification and quantitative analysis of superparamagnetic iron corrosion products. The results indicated that the surface corrosion products mainly consist of goethite and superparamagnetic goethite, with small amounts of lepidocrocite, magnetite, and/or maghemite. A cross-sectional analysis of the corrosion layers on an artifact was performed to better understand the corrosion products and their formation mechanisms. The products formed in different sections (metal, intermediate, and surface) of the corrosion layers on the iron artifact were identified, and a corrosion mechanism was proposed. The intermediate layer adjacent to the metal contains magnetite, maghemite, and lepidocrocite. The results presented in this study provide a deeper understanding of the iron corrosion process, laying a solid foundation for the development of an effective strategy for preserving iron artifacts.

Corrosion occurs when metallic iron undergoes oxidation in the presence of moisture and oxygen. Most iron artifacts buried for long periods of time undergo corrosion forming various iron compounds, depending on the environment. Observation and characterization of corrosion products are necessary for understanding the corrosion mechanisms of iron artifacts and assessing their preservation state. According to reports on nanophase goethite and the formation of protective rust on iron objects, goethite plays an important role in forming a protective and adhesive rust layer on exposed iron objects1. Reddish-brown iron oxide, called rust, typically formed when iron reacts with oxygen in the presence of water or moisture in the air. These adhesive properties are attributed to the small particles of nanophase goethite, which form a stable iron oxide layer on the surface of the iron object. This iron oxide layer protects the iron object from oxygen and moisture, slowing down the rate of corrosion. Therefore, the goethite is most likely the final stable oxide to form, and when present in large quantities, it provides corrosion protection2. Additionally, characterizing the corrosion products on iron artifacts helps us understand the corrosion mechanism and provides information about the environmental conditions to which the iron artifact has been exposed over time. Therefore, identifying the corrosion products formed on the artifact helps prevent further damage to the artifact and assists in predicting its long-term stability.

Common iron corrosion products formed under a variety of environmental conditions include goethite (α-FeOOH), magnetite (Fe3O4), and lepidocrocite (γ-FeOOH). When corrosion occurs under wet soil conditions, in the presence of dissolved oxygen, the iron surface corrodes forming rust on the surface. Rust often combines with clay and soil mineral particles to create a layer of corrosion products3. Generally, the outer corrosion layer consists of products, such as goethite and lepidocrocite, containing iron in a high-oxidation state4,5. Goethite is a typical corrosion product formed in an oxygen-rich, aerobic environment. Goethite is found in soils, sediments, etc., and promotes goethite formation in the presence of oxygen and moisture and neutral to mildly alkaline conditions (typically pH 7–8)6. In contrast, magnetite forms directly on the metal surface; however, magnetite may also be present in other layers of corrosion products4,7. Magnetite is found in partially oxygen-depleted environments and can form in anaerobic or slightly reducing environments, such as buried in soil, and under slightly alkaline conditions (pH 7–10)8. Lepidocrocite forms in moist, oxygen-rich environments with neutral or slightly acidic (pH 6–7) conditions and is often found in the early stages of corrosion9. Additionally, corrosion products that form in certain environments, depending on humidity, changes in oxygen levels, and the presence of certain ions, include iron carbonate, green rust, and iron sulfide. In CO2 environment, the iron carbonate (FeCO3) corrosion product is typically formed under anaerobic conditions lacking oxygen. This can occur in alkaline environments with neutral pH, such as groundwater or moist sediments10. Green rusts form in transitional environments between aerobic and anaerobic conditions. These mixed Fe(II)-Fe(III) hydroxides are commonly found in submerged environments such as coastal areas or local wetlands and are associated with neutral or slightly alkaline aqueous media and fluctuating oxygen levels11. Green rust may form if chlorides or sulfates are present. Iron sulfide (FeS2 and Fe1 − xS) is formed in strongly reducing, anaerobic environments in the presence of sulfide ions. These environments are found primarily in marine sediments, wetlands, and soils with decaying organic matter. The pH range for optimal bacterial oxidation of iron sulfide is 1.8 to 2.512,13. Sulfur compounds, which are generally formed under submerged or immersed conditions, and a reducing environment are essential for iron sulfide formation. In addition, corrosion products generated in aggressive environments can degrade the quality of iron artifacts, resulting in the loss of preserved information. Akaganeite (β-FeOOH) can form if the surrounding environment contains sufficient chloride ions14,15,16. The formation of akaganeite indicates active corrosion of iron under the corrosion layer, directly causing physical damage to the buried artifact as chloride ions dissolve in the moisture that fills the pores and gaps of the artifact. Hematite (α-Fe2O3) is not commonly formed as a corrosion product under normal ground conditions and is rarely observed in iron artifacts17. These iron corrosion products are key indicators of the environmental conditions under which iron corrosion occurs, with pH playing a critical role in determining which specific corrosion product will form.

To assess the corrosion state of iron artifacts, it is important to determine the relative proportions of various corrosion products—iron oxides and oxyhydroxides. X-ray diffraction (XRD), Raman spectroscopy, and Mössbauer spectroscopy are the commonly used methods for analyzing iron corrosion products18,19,20,21. Raman spectroscopy and hyperspectral imaging of corrosion scale have been used to study atmospheric iron artifacts exposed to air for long periods of time and corrosion products formed in marine and hypogea environments22. XRD is the most widely used method for identifying oxides in rust, and through careful calibration of the obtained data, the amount of different types of oxides constituting the corrosion product can be determined. However, XRD is not sensitive to low-crystalline or amorphous phases. Moreover, both magnetite and maghemite have cubic lattice structures at room temperature and almost identical lattice parameters; therefore, they cannot be easily distinguished using XRD23. Mössbauer spectroscopy is also used to determine the amount of different types of iron oxide in corrosion products. Mössbauer spectroscopy can be used to uniquely identify components with different magnetic properties. Separate identification of magnetite and maghemite has been demonstrated using Mössbauer spectroscopy24.

Several Mössbauer spectroscopic studies of iron objects corrosion have been reported18,19,20,21,22,23. Mössbauer spectroscopy was used to characterize rust on iron pillars from ancient structures in Delhi, India18. This study showed that rust contains crystalline iron hydrogen phosphate hydrate and amorphous superparamagnetic iron oxyhydroxide and magnetite. However, separate identification of superparamagnetic goethite (by magnetic relaxation due to small particle volumes) and other iron oxides using Mössbauer spectroscopy at room temperature is difficult since their superparamagnetic parameters are similar25. To solve this problem, we performed low-temperature Mössbauer measurements to analyze the temperature-dependent changes in superparamagnetic relaxation, which allowed us to identify the particles constituting the oxide phase and estimate their size.

This study mainly focuses on the quantitative analysis of iron corrosion products formed on iron artifacts. Mössbauer spectroscopy was performed to determine the composition of corrosion products and measure the magnetic properties of various oxides as a function of temperature. A cross-sectional analysis was performed to investigate the products formed in each layer of the corrosion structures. This study provides a method to quantify iron corrosion products and develops an understanding of iron corrosion mechanisms for effectively evaluating the multilayered corrosion structures formed on iron artifacts.

During the excavation from 2006 to 2007 in C section of Eunpyeong New Town 2nd District, the cemetery from the Joseon Dynasty (South Korea, mid-15th–early 20th centuries) was discovered. The Jinkwan-dong area in Eunpyeong-gu, Seoul, is an inland region located slightly away from the Han River, on the outskirts of the historical center of Seoul. Various historical artifacts from the Joseon Dynasty have been discovered in this area. Iron artifacts were discovered among the artifacts excavated from the Joseon Dynasty cemetery, and two iron forgings and four iron castings were selected as samples (In Fig. 1, the iron forgings; J-D and J-E, the iron castings; J-F, J-G, J-H, and J-I). Figure S1 shows the locations (Google Maps) of Jinkwan-dong in Seoul, South Korea.

The photographic images of the iron artifacts. The names assigned to the artifacts are also indicated. Yellow circles indicate the locations from which the corrosion layer was scraped off for analysis.

The photographic images of the iron artifacts in Fig. 1 show that the artifacts are covered with a thick layer of iron corrosion products. The corrosion product powders were scraped from the surface of the iron artifacts and ground in an agate mortar to a uniform and fine size. A total of ten powders were collected from different locations on surface of the iron artifacts as indicated in Fig. 1. The selected sampling points on the surface of the iron artifact were chosen to reflect the typical corrosion patterns of the entire artifact, with samples taken from various locations. This approach ensures a more accurate understanding of the overall corrosion patterns present on the surface of the artifact. Additionally, cross-sections covered with thick corrosion layers were cut to observe the internal corrosion patterns of the artifact. Samples were collected from different sections (indicated as 11, 12, and 13 in Fig. 1) of the corrosion layer on the artifact designated as J-I. The names assigned to the collected pieces of artifacts and the identification of the locations from which powders were scraped are also presented in Fig. 1. The internal structure and cross-sectional observations of the boundaries between the metal layer and the corrosion layer of some iron artifacts were reported through neutronic analysis in other studies26,27.

To analyze iron oxides and iron oxyhydroxides in the collected corrosion products, iron reagents were selected as reference samples. Magnetite and goethite were purchased from Sigma–Aldrich. Hematite and lepidocrocite were purchased from Alfa Aesar. XRD was used to confirm the composition and phase properties of the reagents and products.

A Rigaku Ultima IV X-ray diffractometer, equipped with CuKα radiation source (λ = 1.5406 Å), was used to measure XRD patterns. Samples were scanned in the 2θ range of 10–80° with a step size of 0.04°. The phase identification was performed using X′Pert HighScore software with ICDD PDF-2 Database. Raman spectra were recorded using Nanophoton Raman-11i, with an excitation wavelength of 532 nm. The Raman spectra were measured after 20 s of light exposure and the data was averaged six times over a range of 100–2000 cm− 1. Mössbauer spectroscopy was performed in constant acceleration mode and transmission geometry using a 57Co source (Rh matrix). A commercial proportional counter filled with a mixture of xenon and carbon gases was used. The measurements were performed over a velocity range of 12 mm/s. The spectra were recorded in the range of 4.2–295 K using a He closed-cycle cryostat to separately identify the iron oxides that were either superparamagnetic at 295 K or magnetically ordered at low temperatures. The isomer shift is quoted relative to α-Fe foil at 295 K and the spectral data are fitted with Lorentzian line shapes using least-square method.

The XRD patterns of the reference samples are shown in Fig. S2. Goethite, hematite, and magnetite exhibit perfect crystalline phases, whereas lepidocrocite exhibits a relatively low crystalline phase. The references exhibited a single phase and were confirmed to be devoid of impurity phases. The XRD patterns of the corrosion product powders, along with the XRD patterns of the references, are shown in Fig. 2. The relative proportions (%) of the crystalline phases in the samples are calculated based on the Rietveld refinement method using X’Pert Highscore software. The results of semi-quantitative XRD analyses of the samples are summarized in Table 1. The XRD analysis of the diffraction intensities indicates that the corrosion products from the J-E-4 sample contain 69.6% goethite and 30.4% magnetite (and/or maghemite). As per the XRD analysis, all the corrosion products are mainly composed of goethite and magnetite (and/or maghemite). XRD analysis of the products from J-F-6 and J-G-7 samples reveals the presence of quartz, which is probably incorporated from the soil on their surfaces. The corrosion products from the J-F-6 sample contain 93.3% quartz, 4.9% goethite, and 1.8% magnetite. The corrosion products from the J-G-7 sample do not contain any other iron oxides or oxyhydroxides other than goethite. Hematite is not detected in any of the collected products using XRD analysis. The XRD pattern of the corrosion products from the J-I-10 sample indicates the presence of goethite and magnetite, with 9.2% lepidocrocite. The nanosized corrosion products containing goethite and lepidocrocite are difficult to detect using XRD as they are poorly crystalline or amorphous.

XRD patterns of the corrosion product powders collected from the iron artifacts. Refer to Fig. 1 for identifying the samples and locations using the indicated names.

The Raman spectra enable phase analysis of the corrosion products scraped from the surfaces of iron artifacts. Raman spectra of reference samples for iron corrosion products were measured, and the spectra of the references with identified peak positions of each phase are shown in Fig. S3. The obtained results correspond well with those reported in the literature28,29,30,31,32,33,34,35,36. The Raman spectrum of goethite exhibits bands centered at 300, 398, 474, and 553 cm− 1. The Raman spectrum of lepidocrocite exhibits well-defined peaks at 255 (most intense), 311, 379, 528, and 648 cm− 1. The Raman spectrum of hematite exhibits intense bands centered at 226, 245, 292, 411, 497, and 612 cm− 1. The magnetite reference sample exhibits the Raman band centered at 672 cm− 1. The magnetite reference sample exhibits a spectrum with relatively weaker intensities and fewer vibrational bands. When the laser power is increased to a relatively high value, the magnetite sample starts to exhibit bands related to hematite, indicating phase change due to laser heating; therefore, the power was adjusted to prevent sample heating34.

Figure 3 shows the Raman spectra of the corrosion products acquired from the iron artifacts. According to the Raman data, the corrosion products contain goethite, lepidocrocite, and magnetite. The Raman spectrum of the samples is mainly composed of goethite. In addition, the spectrum consists of broad bands in the range of 650–750 cm− 1. The superposition of the most intense peak of magnetite at 672 cm− 1 with the broad maghemite band centered at 721 cm− 1 is probably responsible for the observed broad line shape with a shoulder in the range of 650–750 cm− 135,36. This suggests that the corrosion products on the surface may consist of a mixed phase of maghemite and magnetite, but the Raman spectrum results alone are not conclusive. Hematite is not observed in any of the samples.

Raman spectra of all the corrosion products obtained from iron artifacts. Refer to Fig. 1 for identifying the samples and locations using the indicated names.

The identification and proportion of each oxide in corrosion products vary depending on the environment. To accurately identify the corrosion products, Mössbauer spectra were obtained for representative reference samples of iron corrosion products, as shown in Fig. 4. Mössbauer spectrum of hematite at 295 K was fitted with a sextet, as shown in Fig. 4a. The Mössbauer parameters for isomer shift (IS), quadrupole splitting (QS), and magnetic hyperfine field (Hhf) are 0.25 mm/s, − 0.10 mm/s, and 515.4 kOe, respectively. At 295 K, Mössbauer spectrum of magnetite was fitted with two sextets according to tetrahedral (A) and octahedral (B) sites. The area ratio of A and B sites is 4:6. Magnetite has an inverse spinel structure, with Fe3+ and Fe2+/3+ ions at A and B sites, respectively; therefore, the spins at A and B sites are antiparallel. The IS and Hhf values are 0.15 mm/s and 490.4 kOe for the A site and 0.56 mm/s and 458.9 kOe for the B site, respectively. The Mössbauer spectrum of goethite consists of two sextets that are magnetically ordered at 295 K. As shown in Fig. 4b, the measured goethite particles are larger than 15 nm20 and have a well-defined magnetic sextet. In nature, goethite consists of particles with sizes ranging from bulk to small sizes. As the particle size decreases to a size below the bulk limit, the magnetic relaxation time changes, leading to changes in the magnetic field37. The Hhf values of goethite at 4.2 K are 506.1 and 501.4 kOe, respectively. Figure 4c shows the Mössbauer spectra of lepidocrocite fitted with paramagnetic doublet at 77 and 295 K. Lepidocrocite is antiferromagnetic, with a Néel temperature (TN) below 77 K38,39,40. The doublet at 295 K has IS and QS values of 0.25 mm/s and 0.56 mm/s, respectively. The spectrum at 4.2 K was fitted using a broad sextet having an Hhf value of 450.8 kOe. The results clearly indicate that lepidocrocite is paramagnetic above TN. In contrast, the Mössbauer spectrum below TN shows an unusually broad distribution. Table S1 lists the Mössbauer parameters corresponding to the spectra shown in Fig. 4. Akaganeite is reported to be paramagnetic at 295 K41,42. Separate identification of lepidocrocite and akaganeite in corrosion products is impossible because their QS values are similar at 295 K41. However, the TN of akaganeite is close to 295 K, and its spectrum at 77 K corresponds to a magnetically ordered sextet42. Therefore, Mössbauer spectroscopy at 77 K facilitates separate identification of lepidocrocite and akaganeite.

Mössbauer spectra of the references: (a) hematite and magnetite at 295 K; (b) goethite and (c) lepidocrocite at 4.2, 77, and 295 K.

At 295 K, the recorded Mössbauer spectra (Figs. 5(a–f)) of the corrosion products from iron artifacts show a superposition of an intense quadrupole doublet and several sextets. The Mössbauer spectra of the corrosion products from J-E-3, J-F-5, and J-F-6 samples, presented in Figs. 5(a), 5(c), and 5(d), respectively, were analyzed for goethite and magnetite using sextets, and for lepidocrocite and superparamagnetic particles using doublet. The Mössbauer spectra of the corrosion products from J-E-4 and J-H-9 samples, presented in Figs. 5(b) and 5(f), respectively, confirm the presence of maghemite. In contrast, the Mössbauer spectrum of the corrosion products from the J-G-7 sample, presented in Fig. 5(e), consists only of goethite, lepidocrocite, and superparamagnetic particles. The Mössbauer spectra of J-D-1 and J-I-10 samples are shown in Fig. S4.

Mössbauer spectra, recorded at 295 K, of the corrosion products obtained from iron artifacts: (a) J-E-3; (b) J-E-4; (c) J-F-5; (d) J-F-6; (e) J-G-7; (f) J-H-9. Refer to Fig. 1 for identifying the samples and locations using the indicated names.

The Mössbauer spectrum of corrosion products from the J-H-9 sample exhibits characteristic peaks corresponding to maghemite, magnetite, goethite, lepidocrocite, and superparamagnetic particles. The maghemite was analyzed with a sextet having Hhf = 496.9 kOe, QS = -0.05 mm/s, and IS = 0.33 mm/s. The maghemite sextet shows quadrupole splitting that is either nearly zero or slightly asymmetric due to its cubic symmetry. In contrast, the hematite sextet exhibits quadrupole splitting in the range of -0.1 to -0.2 mm/s, with such nonzero quadrupole splitting resulting in an asymmetric sextet, where the first and sixth peaks shift differently from the other four peaks. The magnetite was analyzed with two sextets having Hhf and IS values of 487.4 kOe and 0.19 mm/s for the A site and 458.7 kOe and 0.51 mm/s for the B site, respectively. The A site of magnetite is filled with Fe3+ ion, and the B sites are occupied with Fe2+ and Fe3+ ions. When exposed to oxygen, the magnetite is readily oxidized to the maghemite phase, with the Fe2+ ions partially converted to Fe3+ ions. Magnetite and maghemite cannot be easily distinguished using XRD and Raman spectroscopy. However, Mössbauer spectroscopy can provide quantitative information on the oxidation state of Fe species and measure the contribution of Fe ions occurring at each site. The hyperfine parameters of magnetite and maghemite reported in the literature are as follows24: Magnetite was reported as two sextets with Hhf and IS values of 495 kOe and 0.31 mm/s for the A site and 456 kOe and 0.65 mm/s for the B site, respectively, and maghemite was reported as sextet with Hhf = 491 kOe, QS = -0.05 mm/s, and IS = 0.30 mm/s. The reported hyperfine parameters are similar to the results we analyzed. The Mössbauer spectrum of the J-H-9 sample shows that the hyperfine values of magnetite A site (Tet) and maghemite are similar, leading to an increase in the spectrum intensities. Therefore, it can be concluded that both magnetite and maghemite phases coexist in the J-H-9 sample. Owing to the wide distribution of goethite particle sizes in the sample, the goethite was analyzed with a broad sextet43. The lepidocrocite and superparamagnetic particles were analyzed using doublet with QS and IS values of 0.62 mm/s and 0.27 mm/s, respectively. When the particle size is very small (a few nanometers), the particles are characterized by a doublet at 295 K owing to the superparamagnetic behavior. Since the hyperfine parameters of lepidocrocite and superparamagnetic particles are almost identical at 295 K23, it is very difficult to distinguish between them. Table S2 shows the Mössbauer parameters corresponding to the spectra shown in Fig. 5.

The Mössbauer spectra, recorded at temperatures ranging from 295 to 4.2 K, of the corrosion products from the J-D-2 sample are shown in Fig. 6a. As the temperature decreases, the area under the doublet, which occupies a significant portion at 295 K, decreases, and a magnetically aligned broad spectral component appears in the spectrum. The temperature dependence of the Mössbauer spectrum strongly suggests that the doublet observed in the spectrum at 295 K originates from nanoparticles in the superparamagnetic state. When the goethite particle size is smaller than 15 nm, the Mössbauer spectrum is superparamagnetic doublet at 295 K20. At low temperatures, the spectrum exhibits a magnetic hyperfine field owing to temperature-dependent changes in the relaxation time. The magnetic relaxation slows down at low temperatures, and the peaks in each sextet become sharper, resembling a well-defined magnetic structure. Meanwhile, the spectrum at 77 K still shows the presence of a doublet. The weak doublet at 77 K can be attributed to lepidocrocite. Finally, in the spectrum recorded at 4.2 K, all phases, including lepidocrocite, were analyzed using sextets. Figure 6c shows the relative areas of the iron corrosion products from the J-D-2 sample as a function of temperature. The areas under the spectrum corresponding to lepidocrocite and superparamagnetic goethite, which account for 26.2% at 295 K, respectively, decreases rapidly with decreasing temperature. The superparamagnetic goethite, which is a doublet at 295 K, consists of sextets and doublets below 180 K. The area of magnetic component in the Mössbauer spectrum of goethite increased from 46.1% at 295 K to 67.3% at 77 K. This indicates that in the J-D-2 sample, 46.1% is goethite larger than 15 nm, and 21.2% is superparamagnetic goethite smaller than 15 nm. Since the doublet spectrum at 77 K corresponds only to lepidocrocite, it is confirmed that 5% lepidocrocite is present in the corrosion products from the J-D-2 sample.

Mössbauer spectra of the corrosion products obtained from (a) J-D-2 and (b) J-G-8 samples at various temperatures. (c-d) Histogram presenting the relative areas of various iron compounds present in the corrosion products obtained from J-D-2 and J-G-8 samples as a function of temperature.

The Mössbauer spectra of the iron corrosion products from the J-G-8 sample as a function of temperature are presented in Fig. 6b. The spectrum at 295 K consists of a weak sextet and an intense doublet. As the temperature decreases, the area under the doublet decreases while the area under the sextets increases. The temperature-dependent presence of maghemite and magnetite is not observed in the spectrum. Low-temperature experiments confirmed that the doublet at 295 K is composed of superparamagnetic goethite and lepidocrocite. Figure 6d shows the relative areas of the iron corrosion products in the J-G-8 sample as a function of temperature. The spectrum at 295 K was analyzed for goethite using sextet (48.6%) and superparamagnetic using doublet (51.4%). The spectrum at 77 K was analyzed for lepidocrocite (5.2%) and goethite (goethite + superparamagnetic goethite, 94.8%). The results show that the corrosion products from the J-G-8 sample contains 46.2% superparamagnetic goethite, with a particle size of less than 15 nm. Therefore, to quantitatively analyze the corrosion products formed on iron artifacts, low-temperature Mössbauer analysis was used to identify iron oxides and determine their respective proportions. From the obtained Mössbauer analysis results, it was confirmed that the corrosion products formed in iron artifacts are mainly nano goethite containing superparamagnetic phase. Nano goethite plays a crucial role in forming a protective corrosion layer on iron artifacts, as it chemically stale and represents the final corrosion product in the corrosion environment. Based on these results, it appears that nano goethite forms a stable iron oxide layer on the surface of iron artifacts in this study, which slows down the corrosion rate and helps protect the artifacts from further deterioration.

Figure 7a shows the Mössbauer spectra, measured at 295 K, of the surface, intermediate, and metal sections that are labeled as 11, 12, and 13, respectively, in Fig. 1. The surface layer consists of goethite, superparamagnetic goethite, and lepidocrocite. The low-temperature spectrum indicates that the relative areas under the spectrum corresponding to goethite, superparamagnetic goethite, and lepidocrocite are 68.5, 26.1, and 5.4%, respectively (Fig. S5). The corrosion products in the intermediate layer are identified as magnetite (79.2%), maghemite (15.9%), and lepidocrocite (4.9%). Magnetite is the dominant phase in the intermediate layer. Typically, magnetite forms under conditions of limited oxygen availability or within layers of corrosion products that do not contain sufficient oxygen to form Fe oxyhydroxides. In Fig. 7a, the Mössbauer spectrum obtained from the metal layer of the J–I sample (Sect. 13 in Fig. 1) was fitted with a sum of several sub-spectra corresponding to magnetite, α-Fe, cementite (Fe3C), Fe-C, FexC (para), and γ-Fe(C). In the spectrum of the metal layer, small amounts of magnetite (4.5%) was unintentionally introduced from the intermediate layer during the separation process between the intermediate layer and the metal layer. Magnetite does not originally exist in the metal layer. The sextet with Hhf and IS values of 330 kOe and − 0.10 mm/s, respectively, corresponds to metallic α-Fe with a body-centered cubic lattice. Cast iron contains carbon, forming metastable states, such as cementite, with a fixed iron-carbon ratio44. Cementite can exist as an interstitial solid solution in iron with a non-cubic structure42. The cementite was fitted with two sextets, I and II, with Hhf values of 207.9 and 200.8 kOe, respectively45,46,47. When liquid cast iron solidifies, it decomposes into austenite (γ-Fe), and most of the carbon precipitates in the form of graphite or cementite. On further cooling, the carbon concentration in the austenite decreases as more cementite or graphite precipitates from the solid solution48, and the remaining liquid decomposes into cementite and ledeburite. Below the eutectoid temperature at which one single solid phase transforms into two solid phases simultaneously upon cooling, the austenite transforms into pearlite (α-Fe + Fe3C), and ledeburite (γ-Fe + Fe3C) turns into transformed ledeburite (pearlite + ledeburite). In the obtained spectrum of the metal from the cast iron artifact (sample J–I), the relative areas of α-Fe and cementite are determined to be 11.6 and 68.7%, respectively. Paramagnetic FexC (6.3%) exhibits a poorly crystalline phase, which may be due to the intergrowth structure of cementite49,50. The Fe-C was fitted using a sextet with a Hhf value of 258.8 kOe. A very small amount of iron carbides (2.5%) is observed. Furthermore, the doublet with IS and QS values of 0.05 and 0.43 mm/s, respectively, is attributed to γ-Fe(C), which has faced centered cubic iron with carbon structural inclusions51,52. Since the J-I sample appears to be composed of pearlite, ledeburite, and transformed ledeburite, it is estimated that the cooling rate of the cast iron was slow and the weight% carbon ranged from 2.11 to 4.32%48. The hyperfine parameters corresponding to the spectra in Fig. 7a are listed in Table S3.

(a) Mössbauer spectra acquired from different sections of J-I sample at 295 K. The surface, intermediate, and metal denote the sections labeled as 11, 12, and 13, respectively, in Fig. 1. (b) Mössbauer spectra of the metal (Sect. 13 in Fig. 1) at various temperatures. (c) Schematic showing the various corrosion products observed via cross-sectional analysis of the J-I sample (cast iron), along with possible chemical reactions.

Figure 7b presents Mössbauer spectra of the metal from the J–I sample at various temperatures. Low-temperature measurements were performed to clearly identify the paramagnetic FexC phase and γ-Fe(C) in the spectrum obtained at 295 K. The features observed in the low-temperature spectrum are similar to those of the spectrum at 295 K. The γ-Fe(C) maintains a doublet even at 4.2 K. The area under the spectrum corresponding to cementite at 4.2 K increases by 75.1% compared with that at 295 K. The paramagnetic FexC phase is estimated to be cementite with poor crystallinity.

Based on the experimental results, the formation mechanism of corrosion products on iron artifacts is proposed. In this study, the iron artifacts were excavated from a cemetery in an inland area, and since the burial conditions did not involve specific ions such as Cl⁻ or S⁻ and environments, corrosion products like iron carbonate and iron sulfide were not considered in this corrosion mechanism. During the corrosion of iron, the release of Fe2+ ions varies significantly depending on the pH of the environment. In acidic conditions (pH < 6), Fe2+ ions are easily released and can precipitate as Fe(OH)2 or transform into other iron oxides. In neutral conditions (pH 6–8), Fe(OH)2 forms and may further oxidize into lepidocrocite or magnetite, depending on the availability of oxygen and moisture. In alkaline environments (pH > 8), Fe(OH)3 forms, which can then oxidize into magnetite or goethite. At first, the Fe2+ ions are generated through a cathodic reaction, and then Fe(OH)2 is generated through an anodic reaction53. Gradually, Fe(OH)2 oxidizes and converts to Fe(OH)3 and then to magnetite and lepidocrocite. For the formation of magnetite, moisture and oxygen must be removed from Fe(OH)3. Magnetite forms when oxidation proceeds slowly, and the oxidation time is sufficiently long to remove oxygen from the crystal lattice. However, when oxidation proceeds rapidly, lepidocrocite with distorted cubic oxygen packing forms quickly during the early stages of corrosion. Over time, especially in conditions with low oxygen and a reducing environment, lepidocrocite transforms into magnetite. This process occurs when lepidocrocite is present near a metal surface and a reduction reaction by electron transfer occurs due to its reaction with the metal. Lepidocrocite is formed in a neutral or slightly acidic environment, whereas magnetite is formed in a reducing environment, so conversion to magnetite can be promoted when the pH is between 6 and 9. Therefore, lepidocrocite is easily reduced to magnetite at the metal interface and partially oxidized to maghemite. Additionally, magnetite formed by reduction of lepidocrocite is reoxidized to maghemite rather than lepidocrocite54,55,56,57,58,59. However, it is still unclear whether magnetite and lepidocrocite form simultaneously during the initial corrosion product formation process. The Mössbauer spectroscopy results confirm that the final corrosion product is goethite, a thermodynamically stable compound.

Figure 7c presents a schematic of the cross-section of the J-I sample, along with the identified corrosion products labeled on different sections (indicated as 11, 12, and 13 in Fig. 7a). The metal in the J-I sample, which is cast iron, consists of pearlite, cementite, and transformed ledeburite. When the metal surface begins to corrode, the magnetite is formed, some of which oxidizes and converts to maghemite. The magnetite formation process varies depending on the oxidation and reduction conditions. When the reduction of Fe(OH)3 proceeds slowly, magnetite is formed. However, when oxidation proceeds rapidly, lepidocrocite is formed from Fe(OH)3. Subsequently, lepidocrocite readily reduces and converts to magnetite. Therefore, magnetite and maghemite, which are oxidized iron compounds, exist in the intermediate layer. The surface layer is mainly composed of goethite, which is formed when magnetite oxidizes. The low-temperature Mössbauer spectroscopy results showed that the produced goethite consists of particles having different sizes. Other corrosion products that are present on the surface include magnetite, maghemite, and lepidocrocite. The analysis of the Mössbauer spectra of the surface corrosion products indicate the presence of ~ 5% lepidocrocite. Lepidocrocite is mainly located in the crevices of relics or at the interface with soil. Lepidocrocite forms during the early stages of corrosion; however, as the oxidation time increases, the amount of lepidocrocite decreases due to its reduction into magnetite, which ultimately changes to goethite. The schematic corrosion pattern in Fig. 7c is limited to the J-I sample, so the corrosion mechanism may vary depending on the environment.

This study focused on the quantitative characterization of corrosion products formed on iron artifacts, emphasizing new insights gained through the application of Mössbauer spectroscopy. The analysis revealed that the surfaces of the iron artifacts predominantly consist of goethite, with minor amounts of lepidocrocite, magnetite, and/or maghemite. Notably, hematite and akaganeite were absent in the analyzed samples. A key advancement of this study was the precise quantification of superparamagnetic goethite and lepidocrocite within the surface corrosion layers using low-temperature Mössbauer spectroscopy. This approach enabled the detection of otherwise challenging-to-identify phases, significantly enhancing the accuracy of phase composition analysis. A cross-sectional analysis, including the metal, intermediate, and surface layers, was performed on the corrosion layers of iron artifact, and the corrosion products in each layer were identified. The temperature-dependent Mössbauer spectroscopy analysis revealed the phase composition of the metallic core, offering critical data for assessing the metallurgical properties of cast iron. Based on these findings, a comprehensive formation mechanism of corrosion products on iron artifacts was proposed, grounded in the oxidation process. The novel application of Mössbauer spectroscopy not only improved the understanding of iron corrosion processes but also established a scientific foundation for advancing archaeological research and developing effective preservation strategies for iron artifacts.

All data generated or analyzed during this study are included in this article and its supplementary file.

Cook, D. C., Oh, S. J., Balasubramanian, R. & Yamashita, M. The role of goethite in the formation of the protective corrosion layer on steels. Hyperfine Interact. 122, 59–70 (1999).

ADS CAS MATH Google Scholar

Cook, D., Oh, C. & Townsend, S. J. H. E., the Protective Layer Formed on Steels after Long-Term Atmospheric ExposureCorrosion98343 (NACE-International, 1998).

Neff, D., Dillmann, P., Bellot-Gurlet, L. & Beranger, G. Corrosion of iron archaeological artefacts in soil: characterization of the corrosion system. Corros. Sci. 47, 515–535 (2005).

CAS Google Scholar

Balasubramaniam, R. On the corrosion resistance of the Delhi iron pillar. Corros. Sci. 42, 2103–2129 (2000).

CAS MATH Google Scholar

Jegdic, B., Radovanavic, S. P., Ristic, S. & Alil, A. Corrosion processes nature and composition of corrosion products on iron artefacts of weaponry. Sci. Tech. Rev. 61, 50–56 (2011).

CAS Google Scholar

Chen, S. A., Heaney, P. J., Post, J. E., Eng, P. J. & Stubbs, J. E. Hematite-goethite ratios at pH 2–13 and 25–170°C: A time-resolved synchrotron X-ray diffraction study. Chem. Geol. 606, 120995 (2022).

CAS Google Scholar

Misawa, T., Kyuno, T., Suëtaka, W. & Shimodaira, S. The mechanism of atmospheric rusting and the effect of Cu and P on the rust formation of low alloy steels. Corros. Sci. 11, 35–48 (1971).

CAS MATH Google Scholar

Ishikawa, T., Kondo, Y., Yasukawa, A. & Kandori, K. Formation of magnetite in the presence of ferric oxyhydroxides. Corros. Sci. 40, 1239–1251 (1998).

CAS MATH Google Scholar

Turner, C. & Chi, C. L. Formation of Corrosion Products of Carbon Steel under Condenser Operating Conditions. International Conference on Nuclear Plant Chemistry NPC 2021. (2012).

Matthiesen, H., Hilbert, L. R. & Gregory, D. J. Siderite as a corrosion product on archaeological iron from a waterlogged. Environ. Stud. Conserv. 48, 183–194 (2003).

CAS MATH Google Scholar

Refait, P., Abdelmoula, M., Génin, J. M. R. & Sabot, R. Green rusts in electrochemical and microbially influenced corrosion of steel. C R - Geosci. 338, 476–487 (2006).

CAS Google Scholar

Bosecker, K. Bioleaching: metal solubilizaiton by microorganisms. FEMS Microbiol. Rev. 20, 591–604 (1997).

CAS Google Scholar

Mahmoud, A., Cézac, P., Hoadley, A. F. A., Contamin, F. & D’Hugues, P. A review of sulfide minerals microbially assisted leaching in stirred tank reactors. Int. Biodeterior. Biodegrad. 119, 118–146 (2017).

CAS Google Scholar

Schmutzler, B. & Ebinger-Rist, N. The conservation of iron objects in archaeological preservation-Application and further development of alkaline sulphite method for conservation of large quantities of iron finds. Mater. Corros. 59, 248–253 (2008).

CAS Google Scholar

Réguer, S., Dillmann, P. & Mirambet, F. Buried iron archaeological artefacts: corrosion mechanisms related to the presence of Cl-containing phases. Corros. Sci. 49, 2726–2744 (2007).

Google Scholar

Ma, Y., Li, Y. & Wang, F. The effect of β-FeOOH on the corrosion behavior of low carbon steel exposed in tropic marine environment. Mater. Chem. Phys. 112, 844–852 (2008).

CAS MATH Google Scholar

Ståhl, K. et al. Lanschot. J. On the Akaganéite crystal structure, phase transformations and possible role in post-excavational corrosion of iron artifacts. Corros. Sci. 45, 2563–2575 (2003).

MATH Google Scholar

Balasubramaniam, R. & Kumar, R. Characterization of Delhi iron pillar rust by X-ray diffraction, fourier transform infrared spectroscopy and Mössbauer spectroscopy. Corros. Sci. 42, 2085–2101 (2000).

CAS MATH Google Scholar

Kamimura, T. & Nasu, S. Mössbauer spectroscopic study of rust formed on a weathering steel exposed for 15 years in an industrial environment. Mater. Trans. 41, 1208–1215 (2000).

CAS MATH Google Scholar

Cook, D. C. Application of Mössbauer spectroscopy to the study of corrosion. Hyperfine Interact. 153, 61–82 (2004).

ADS CAS MATH Google Scholar

Wagner, U. et al. The stabilizataion of archaeological iron objects: Mössbauer and XRD studies. Hyperfine Interact. 208, 111–1116 (2012).

ADS CAS MATH Google Scholar

Liu, W., Cheng, X., Wu, N. & Wang, K. Comparison of semiquantitative methodologies using Raman mapping for corrosion products on iron artifacts. J. Cult. Herit. 64, 167–175 (2023).

MATH Google Scholar

Kamimura, T., Nasu, S., Kuzushita, K., Tazaki, T. & Morimoto, S. Mössbauer spectroscopic study of rust formed on a weathering steel and a mild steel exposed for a long term in an industrial environment. Mater. Trans. 43, 694–703 (2002).

CAS Google Scholar

Winsett, J. et al. Quantitative determination of magnetite and maghemite in iron oxide nanoparticles using Mössbauer spectroscopy. SN Appl. Sci. 1, 1636 (2019).

CAS MATH Google Scholar

Morcillo, M. et al. Marine atmospheric corrosion of carbon steels. Rev. Metal. 51, e045 (2015).

MATH Google Scholar

Kim, L. Y., Kim, T., Kim, J. U. & Cho, N. C. Study of corrsion system on iron artefacts through neutronic analysis. J. Conserv. Sci. 40, 113–122 (2024).

MATH Google Scholar

Kim, L. Y. Study on the corrosion system of excavated Iron artifacts utilizing neutronic analysis. (Master’s thesis, Kongju National University) (2024).

Pingitore, G. et al. Structural characterization of corrosion product layers on archaeological iron artifacts from vigna Nuova, Crotone (Italy). J. Cult. Herit. 16, 372–376 (2015).

Google Scholar

Thibeau, R. J., Brown, C. W. & Heidersbach, R. H. Raman spectra of possible corrosion products of iron. Appl. Spectrosc. 32, 532–535 (1978).

ADS CAS MATH Google Scholar

Neff, D., Reguer, S., Bellot-Gurlet, L., Dillmann, P. & Bertholon, R. Structural characterization of corrosion products on archaeological iron: an integrated analytical approach to Establish corrosion forms. J. Raman Spectrosc. 35, 739–745 (2004).

ADS CAS Google Scholar

Bernabale, M. et al. A comprehensive strategy for exploring corrosion in iron–based artefacts through advanced multiscale X–ray microscopy. Sci. Rep. 12, 6125 (2022).

ADS CAS PubMed PubMed Central MATH Google Scholar

Froment, F., Tournié, A. & Colomban, P. Raman identification of natural red to yellow pigments: ochre and iron-containing ores. J. Raman Spectrosc. 39, 560–568 (2008).

ADS CAS Google Scholar

Salem, Y., Oudbashi, O. & Eid, D. Characterization of the microstructural features and the rust layers of an archaeological iron sword in the Egyptian museum in Cairo (380–500 A.D). Herit. Sci. 7, 1–12 (2019).

Google Scholar

Hanesch, M. Raman spectroscopy of iron oxides and (oxy)hydroxides at low laser power and possible applications in environmental magnetic studies. Geophys. J. Int. 177, 941–948 (2009).

ADS CAS Google Scholar

Molchan, I. S. et al. S. Corrosion behaviour of mild steel in 1-alkyl-3-methylimidazolium tricyanomethanide ionic liquids for CO2 capture applications. RSC Adv. 4, 5300 (2014).

ADS CAS MATH Google Scholar

Bellot-Gurlet, L. et al. Raman studies of corrosion layers formed on archaeological irons in various media. J. Nano Res. 8, 147–156 (2009).

CAS Google Scholar

Fock, J., Hansen, M. F., Frandsen, C. & Mørup, S. On the interpretation of Mössbauer spectra of magnetic nanoparticles. J. Magn. Magn. Mater. 445, 11–21 (2018).

ADS CAS MATH Google Scholar

Johnson, C. E. Antiferromagnetism of γ FeOOH: a Mössbauer effect study. J. Phys. C: Solid State Phys. 2, 1996–2002 (1969).

ADS CAS MATH Google Scholar

Bauer, P., Genin, J. M. & Rezel, D. Mössbauer effect evidence of Chlorine environments in ferric oxyhydroxides from iron corrosion in chlorinated aqueous solution. Hyperfine Interact. 28, 757–760 (1986).

ADS CAS Google Scholar

Hirt, A. M., Lanci, L., Doboson, J., Weidler, P. & Gehring, A. U. Low-temperature magnetic properties of lepidocrocite. J. Raman Spectrosc. 107, EPM 5-1-EPM 5–9 (2002).

Dézsi, I., Keszthelyi, L., Kulgawczuk, D., McInár, B. & Eissa, N. A. Mössbauer study of β- and δ-FeOOH and their disintegration products. Phys. Status Solidi B Basic. Res. 22, 617–629 (1967).

ADS Google Scholar

Génin, J. M. R., Abdelmoula, M., Ruby, C. & Upadhyay, C. Speciation of iron; characterisation and structure of green rusts and FeII–III oxyhydroxycarbonate fougerite. C R Geoscience. 338, 402–419 (2006).

ADS Google Scholar

Madsen, D. E. et al. Magnetic fluctuations in nanosized goethite (α-FeOOH) grains. J. Phys. : Condens. Matter. 21, 016007 (2009).

ADS CAS PubMed Google Scholar

Bhadeshia, H. K. D. H. & Cementite Int. Mater. Rev. 65, 1–27 (2020).

CAS Google Scholar

Ron, M., Shechter, H. & Hirsch, A. A. On the Mössbauer study of cementite. Phys. Lett. 20, 481–483 (1966).

ADS CAS MATH Google Scholar

Gyulasaryan, H. et al. Iron-cementite nanoparticles in carbon matrix: synthesis, structure and magnetic properties. J. Magn. Magn. Mater. 559, 169503 (2022).

CAS MATH Google Scholar

Liu, X-W. et al. Mössbauer spectroscopy of iron carbides: from prediction to experimental confirmation. Sci. Rep. 6, 26184 (2016).

ADS CAS PubMed PubMed Central Google Scholar

Reed-Hill, R. E. & Abbaschian, R. Physical Metallurgy Principles. Chapter 18: the Iron-Carbon Alloy System588–591 (PWS Publishing Company, 1991).

Paalanen, P. P., Vreeswijk, S. H., Dugulan, A. I. & Weckhuysen, B. M. Identification of iron carbides in Fe(-Na-S)/α-Al2O3 Fischer-Tropsch synthesis catalysts with X-ray powder diffractometry and Mössbauer absorption spectroscopy. Chem. Cat Chem. 12, 5121–5139 (2020).

CAS Google Scholar

Kumar, R. & Sahoo, B. One-step pyrolytic synthesis and growth mechanism of core–shell type Fe/Fe3C-graphite nanoparticles-embedded carbon globules. Nano-Struct Nano-Objects. 16, 77–85 (2018).

CAS MATH Google Scholar

Melo, I. N. R. et al. Influence of Niobium adding on the microstructure and abrasive wear resistance of a Heat-Treated High-Chromium Near-Eutectic cast iron alloy. Mater. Res. 25, e20210562 (2022).

MATH Google Scholar

Aparicio, C., Machala, L. & Marusak, Z. Thermal decomposition of Prussian blue under inert atmosphere. J. Therm. Anal. Calorim. 110, 661–669 (2012).

CAS Google Scholar

Bai, Z. et al. Effect of iron diffusion on the corrosion behavior of carbon steels in soil environment. RSC Adv. 8, 40544 (2018).

ADS CAS PubMed PubMed Central Google Scholar

Ponomar, V. P., Bagmut, M. M., Kalinichenko, E. A. & Brik, A. B. Experimental study on oxidation of synthetic and natural magnetites monitored by magnetic measurements. J. Alloys Compd. 848, 156374 (2020).

CAS Google Scholar

Cudennec, Y. & Lecerf, A. Topotactic transformations of goethite and lepidocrocite into hematite and maghemite. Solid State Sci. 7, 520–529 (2005).

ADS CAS MATH Google Scholar

Hellige, K., Pollok, K., Larese-Casanova, P., Behrends, T. & Peiffeer, S. Pathways of ferrous iron mineral formation upon sulfidation of lepidocrocite surfaces. Geochim. Cosmochim. Acta. 81, 69–81 (2012).

ADS CAS Google Scholar

Zegeye, A., Abdelmoula, M., Usman, M., Hanna, K. & Ruby, C. In situ monitoring of lepidocrocite bioreduction and magnetite formation by reflection Mössbauer spectroscopy. Am. Mineral. 96, 1410–1413 (2011).

ADS CAS Google Scholar

Ponomar, V. P., Antonenko, T. S., Vyshnevskyi, O. A. & Brik, A. B. Thermally induced changes in the magnetic properties of iron oxide nanoparticles under reducing and oxidizing conditions. Adv. Powder Technol. 31, 2587–2596 (2020).

CAS Google Scholar

Hanesch, M., Stanjek, H. & Petersen, N. Thermomagnetic measurements of soil iron minerals: the role of organic carbon. Geophys. J. Int. 165, 53–61 (2006).

ADS CAS Google Scholar

Download references

The authors would like to thank the Kongju National University for providing the samples. This work was supported by the National Research Institute of Cultural Heritage grant funded by the Korea Heritage Service (No. RS-2021-NC100301).

HANARO Utilization Division, Korea Atomic Energy Research Institute, Daejeon, 34057, Republic of Korea

Hyunkyung Choi, Gwang Min Sun & Young Rang Uhm

Department of Heritage Science and Technology Studies, Graduate School of Cultural Heritage, Korea National University of Heritage, Buyeo, 33115, Republic of Korea

Min Su Han

Department of Cultural Heritage Conservation Science, Kongju National University, Gongju, 32588, Republic of Korea

Nam-Chul Cho & Heewon Hwang

You can also search for this author inPubMed Google Scholar

You can also search for this author inPubMed Google Scholar

You can also search for this author inPubMed Google Scholar

You can also search for this author inPubMed Google Scholar

You can also search for this author inPubMed Google Scholar

You can also search for this author inPubMed Google Scholar

H. C. performed all experiments, and analyzed all datasets and figures and interpretation of the analytical data as well as preparing and editing the manuscript. M. S. H., N.-C. C., and H. H. prepared the archaeological data of the site and samples. G. M. S. contributed to sampling and designed the mechanism. Y. R. U. reviewed the manuscript, acquired funds, and supervised the overall project. All authors reviewed the manuscript.

Correspondence to Young Rang Uhm.

The authors declare no competing interests.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Below is the link to the electronic supplementary material.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

Choi, H., Han, M.S., Cho, NC. et al. Understanding the corrosion mechanism of iron artifacts using mössbauer spectroscopy. Sci Rep 15, 10207 (2025). https://doi.org/10.1038/s41598-025-95196-3

Download citation

Received: 21 August 2024

Accepted: 19 March 2025

Published: 25 March 2025

DOI: https://doi.org/10.1038/s41598-025-95196-3

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative