Characterization of Delhi iron pillar rust by X-ray di raction, Fourier transform infrared spectroscopy and MoÈ ssbauer spectroscopy - PDF

Description
Corrosion Science 42 (2000) 2085±2101 Characterization of Delhi iron pillar rust by X-ray di raction, Fourier transform infrared spectroscopy and MoÈ ssbauer spectroscopy

Please download to get full document.

View again

of 17
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Graphics & Design

Publish on:

Views: 17 | Pages: 17

Extension: PDF | Download: 0

Share
Transcript
Corrosion Science 42 (2000) 2085±2101 Characterization of Delhi iron pillar rust by X-ray di raction, Fourier transform infrared spectroscopy and MoÈ ssbauer spectroscopy R. Balasubramaniam a, *, A.V. Ramesh Kumar b a Department of Materials and Metallurgical Engineering, Indian Institute of Technology, Kanpur , India b Electrochemistry and Corrosion Division, Defense Materials, Stores Research and Development Establishment, Kanpur , India Received 19 February 1999; accepted 21 March 2000 Abstract Rust samples obtained from the region just below the decorative bell capital of the Delhi iron pillar (DIP) have been analyzed by X-ray di raction (XRD), Fourier transform infrared spectroscopy (FTIR) and MoÈ ssbauer spectroscopy. The identi cation of iron hydrogen phosphate hydrate in the crystalline form by XRD was unambiguous. Very weak di raction from the oxyhydroxides/oxides of iron was observed indicating that these phases are most likely to be present in the amorphous form in the rust. The present XRD analysis of rust obtained from an inaccessible area of the DIP has also been compared with earlier analyses of DIP rust obtained from regions accessible to the public. FTIR indicated that the constituents of the scale were g-, a-, d-feooh, Fe 3 x O 4 and phosphate, and that the scale was hydrated. The unambiguous identi cation of the iron oxides/oxyhydroxides in the FTIR spectrum implied that they are present in the amorphous state, as XRD did not reveal these phases. The FTIR results have also been compared with earlier FTIR spectroscopic results of atmospheric rust formation. MoÈ ssbauer spectroscopy indicated that the rusts contained g-feooh, superparamagnetic a-feooh, d-feooh and magnetite, all in the amorphous form. The MoÈ ssbauer spectrum also con rmed that iron in the crystalline iron hydrogen phosphate hydrate, whose presence was con rmed by XRD, was in the ferric state indicating that it was a stable end corrosion product Elsevier Science Ltd. All rights reserved. * Corresponding author X/00/$ - see front matter Elsevier Science Ltd. All rights reserved. PII: S X(00) 2086 R. Balasubramaniam, A.V. Ramesh Kumar / Corrosion Science 42 (2000) 2085±2101 Keywords: Delhi iron pillar; Rust characterization; X-ray di raction; Fourier transform infrared spectroscopy; MoÈ ssbauer spectroscopy; Phosphate; Amorphous iron oxyhydroxides 1. Introduction The Delhi iron pillar (DIP) (Fig. 1) is a testimony to the high level of skill achieved by the ancient Indian iron smiths in the extraction and processing of iron. It has attracted the attention of archaeologists and corrosion technologists, as it has withstood corrosion for nearly 1600 years. Several theories, which have been proposed to explain its superior corrosion resistance, can be broadly classi ed into two categories [1±3]: the environmental and the material theories. The proponents of the environment theory state that the mild climate of Delhi is responsible for the corrosion resistance of the (DIP), while on the other hand, several investigators have stressed the importance of the material of construction as the primary cause for its corrosion resistance. These theories have been critically reviewed elsewhere [1±3]. The role of slag particles in enhancing the passivity in the material has also been mentioned earlier [2,3]. The precise reason for the corrosion resistance of the famous 1600-year-old DIP is not well understood. In order to glean insights into the nature of the passive lm that forms on the DIP (which is responsible for its excellent resistance to atmospheric corrosion [3]), a detailed characterization study of the undisturbed rust from the region just below the decorative bell capital of the DIP was undertaken. The characterization of rust samples by X-ray di raction (XRD), Fourier transform infrared spectroscopy (FTIR) and MoÈ ssbauer spectroscopy is the subject of the present communication. The reason for utilizing several di erent characterization techniques must be emphasized. XRD analysis would provide information about phases that possess long range order and not about amorphous phases. Therefore, XRD analysis alone would not provide the complete picture of the rust's nature, especially if it contains amorphous matter. As several iron oxides/oxyhydroxides that form on iron during atmospheric corrosion can be amorphous in nature, these phases would not provide any di raction peaks, and therefore, XRD is not the characterization technique that will reveal the presence of these amorphous phases in the rust. The characterization technique, which can reveal the di erent allotropic modi cations of iron oxides and oxyhydroxides even if they are amorphous in nature, is FTIR. The results of FTIR coupled with that of Raman spectroscopy (which provides complementary information) should provide information about the various forms of iron oxides and oxyhydroxides present, but they will not provide any information whether the phases are amorphous or crystalline in nature, as both these techniques utilize radiation (infrared in FTIR spectroscopy and visible in Raman spectroscopy) for determining the molecular R. Balasubramaniam, A.V. Ramesh Kumar / Corrosion Science 42 (2000) 2085± vibration (due to bending and stretching of the bonds) between iron and oxygen [4]. As the molecular vibration in all known iron oxides and hydroxides are well characterized and available in the literature, comparison of the FTIR spectra from the DIP rust with standards would provide the exact nature of the phases of iron oxides and oxyhydroxides present. IR spectroscopy has been successfully utilized Fig. 1. The Delhi iron pillar before the construction of the iron grill cage around the stone platform. 2088 R. Balasubramaniam, A.V. Ramesh Kumar / Corrosion Science 42 (2000) 2085±2101 to identify iron oxide/oxyhydroxide phases under atmospheric corrosion [5,6] and aqueous corrosion [7] conditions. MoÈ ssbauer spectroscopy can be used to obtain further insights on the structural (i.e. amorphous or crystalline) and chemical aspects of the rusts. MoÈ ssbauer spectroscopy is a powerful tool to understand the amorphous versus crystallinity controversy in the case of oxides/oxyhydroxides [8]. In case the phase that constitutes the rust is crystalline, it would lead to alignment of magnetic elds in the individual grains of the phase and provide a characteristic hyper ne splitting with characteristic magnetic elds. For example, in the case of crystalline magnetite, two sextets (10-line pattern) are to be expected in the MoÈ ssbauer spectrum at ambient temperatures. However, in case this phase is amorphous, the magnetic elds will not be aligned in the individual grains and this would result in collapse of magnetic elds or as a result of which a doublet would be obtained instead of a sextet [8±10]. A collapsed sextet would indicate the ne nanocrystalline nature of magnetite. Therefore, MoÈ ssbauer spectroscopy can be advantageously utilized to probe the crystal nature (i.e. amorphous or crystalline) of the phase. The use of MoÈ ssbauer spectroscopy for elucidation of the chemical nature of corrosion products is also well established [8,11±13]. Another advantage of utilizing MoÈ ssbauer spectroscopy is that the asymmetry in any peak of central doublet indicates the contribution due to some other species (phase). For example, Fe 2+ ions show high positive value for isomer shift and quadrupole splitting resulting in asymmetry in the central doublet in which a higher intense rst peak (from left) is obtained. The second peak of ferrous doublet will be in the more positive region of the spectrum [14]. 2. Experimental Rust samples were collected using a plastic scraper from several di erent locations in the region just below the decorative bell capital (Fig. 2(a)). Observation of the same region, nearly a year after the samples were collected (Fig. 2(b)), did not reveal any evidence for rust removal, thereby indicating that the passive lm is self-healing. This is the region where the rust layer on the exposed surface of the pillar is maximum [15] and therefore, this allowed the collection of a signi cant amount of rust suitable for characterization by several techniques. Moreover, this is the region where the rust is the oldest as this portion of the pillar is inaccessible to the public. A portion of the rust sample collected from each location was rst ground into ne powder and mounted in between two thin polymer foils. A part of the same ground powder was also analyzed using FTIR. The polymer foil containing the DIP rust was used for MoÈ ssbauer spectroscopic analysis of the phases present. The polymer foils were mounted in a di ractometer such that the at foil was placed horizontally and the incident beam could sample the specimen through the whole range of 2y studied. Several di raction patterns were obtained using Cu K a radiation at di erent scan speeds and scanning angle steps. The present paper R. Balasubramaniam, A.V. Ramesh Kumar / Corrosion Science 42 (2000) 2085± Fig. 2. (a) Location from below the decorative bell capital from where the rust samples were collected and (b) same location photographed after nearly a year indicating that the rust is self-healing. 2090 R. Balasubramaniam, A.V. Ramesh Kumar / Corrosion Science 42 (2000) 2085±2101 reports the di raction pattern taken at a scan rate of 38 per min (i.e per s), with the scan being conducted between 2y ˆ 10±808 in steps of A small portion of the rust was nely ground and then pressed (in vacuum) in the form of a disc using spectroscopically pure dry KBr. The FTIR spectrum was recorded at room temperature using a Nicolet Magna 750 Series 2, USA FTIR system. In order to study the rust by MoÈ ssbauer spectroscopy, the polymer foils were mounted on a sample holder of a MoÈ ssbauer spectrometer in transmission geometry. The spectrometer controller model was S-600 of Austin Science Associates, USA with linear motor of ASA, USA which was coupled with a Norland Ortec Multichannel analyzer and proportional counter (Kr at 1 atm and 3% CO 2 ) manufactured by Ranger Corporation, USA and these were used to count the 14.4 KeV g-ray. The source used was 400 MBq 57 Co source in Palladium matrix (Amersham, UK). The MoÈ ssbauer spectrum was recorded in a y back mode at room temperature in 1024 channels, which gives a single spectrum. The MoÈ ssbauer data was tted to Lorentzian approximation using mm/ sirus v.2.7 MoÈ ssbauer data handling program [16]. The MoÈ ssbauer parameters reported are with respect to a-fe standard. 3. Results and discussion 3.1. X-ray di raction Phase identi cation The complete di raction pattern obtained from the rust is presented in Fig. 3. The results from the second rust sample were identical. The XRD data were analyzed with the corrosion products of iron, phosphorus, lead, chromium, nickel, copper, and tin using JCPDS di raction les [17]. The JCPDS data of oxides and oxyhydroxides of iron were compared carefully. The angles and the relative intensities (after subtracting the background radiation) are provided in Table 1. The peaks could only be unambiguously identi ed with iron hydrogen phosphates. The di raction peaks were carefully compared with the standard for iron Table 1 XRD analysis of Delhi iron pillar rust using Cu K a radiation 2y (expt) (8) I/I 0 (expt) (%) phase 2y (standard) (8) I/I 0 (standard) (%) FeH 3 P 2 O 8 4H 2 O FeH 3 P 2 O 8 4H 2 O FeH 3 P 2 O 8 4H 2 O FeH 3 P 2 O 8 4H 2 O FeH 3 P 2 O 8 4H 2 O R. Balasubramaniam, A.V. Ramesh Kumar / Corrosion Science 42 (2000) 2085± phosphate (FePO 4 2H 2 O Ð le number ) and several other iron hydrogen phosphate hydrates (FeH 3 P 2 O 8 4H 2 O Ð le number , FeH 6 P 3 O 9 H 2 OÐ le number , FeH 2 P 3 O 10 (3/2)H 2 O Ð le number and FeH 3 P 2 O 6 3H 2 O Ð le number ) [17]. All the peaks could be identi ed with the phase FeH 3 P 2 O 8 4H 2 O. The 2y values and corresponding intensity of the peaks from the standard di raction le of FeH 3 P 2 O 8 4H 2 O are also provided in Table 1. It is seen that the high intensity peak of the phase FeH 3 P 2 O 8 4H 2 O appears as a signi cant peak at 2y of The experimental peak occurring at 2y of is again a signi cant peak of the phase FeH 3 P 2 O 8 4H 2 O. The possible reasons for the experimentally determined peak at being the highest intensity peak rather than the peak at are (a) di erent degrees of hydration, (b) slight di erences in composition, and (c) orientation e ects in the DIP rust sample as compared to the standard. Contribution to the 100% experimental peak (at 2y of ) due to di raction from other planes is also a possibility as the standard pattern shows the presence of three di raction peaks clustered very near to this angle. A similar situation is obtained for the di raction peak occurring at 2y of , as there are two signi cant peaks (of theoretical intensities 19 and 18%) of the phase FeH 3 P 2 O 8 4H 2 O occurring near this di raction peak (Table 1). Therefore, there is no ambiguity in the identi cation of the iron hydrogen phosphate hydrate phase (FeH 3 P 2 O 8 4H 2 O) in the DIP rust as the experimentally Fig. 3. X-ray di raction pattern of the DIP rust showing the intensities in the region 2y ˆ : 2092 R. Balasubramaniam, A.V. Ramesh Kumar / Corrosion Science 42 (2000) 2085±2101 obtained relative intensities can also be matched to the standard JCPDS di raction data of this phase. As the rust obtained is very old, it is not surprising that the stable crystalline form of iron hydrogen phosphate hydrate [18] was identi ed. However, it should be borne in mind that the presence of the precursor to this phase, i.e. Fe 3 (PO 4 ) 2 in the rust could not be ruled out as it is amorphous [18] in nature and would not have provided di raction peaks. Its relative amount is not important as it is the precursor to the much more stable crystalline form of iron hydrogen phosphate hydrate, which has been unambiguously identi ed in the rust by XRD. The presence of FeH 3 P 2 O 8 4H 2 O is also theoretically favored as it is composed of FePO 4 and H 3 PO 4, whose standard free energies for formation is extremely negative at room temperature. At K and 1 bar pressure, the standard free energy of formation of FePO 4 2H 2 Ois kj/mol, while that for crystalline H 3 PO 4 is kj/mol [19]. In comparison, the standard free energies of formation of the oxides and oxyhydroxides of iron are much higher [19]. On the topic of thermodynamics, it is interesting to note the experimental result of Urasova et al. [20] who found that phosphate is stable even at P concentration as low as 0.24 wt% in the Fe-P-O system. Therefore, the identi cation of iron hydrogen phosphate hydrate is reliable as the laws of thermodynamics are exact and conclusions based on thermodynamics are sound and reliable. From the above analysis, it would apparently appear that the rust of the iron pillar is composed of only iron phosphates. This may not be true as very careful analysis of the XRD pattern revealed weak peak(s) in the region of 2y ˆ (Fig. 4). When the di raction data of the several oxides and oxyhydroxides of iron were analyzed, it was noticed that the maximum intensity peaks in the case of several iron oxides and oxyhydroxides are concentrated in the 2y range between 30 and 358. It, therefore, appears from the very weak di raction intensities obtained in this range that there should be a very small amount of crystalline iron oxides or oxyhydroxides present in the rust Comparison with earlier XRD studies Earlier XRD studies of rusts from the DIP are summarized below in order to compare the present results with them. The rust samples that were obtained in the present case were taken from the surface of the pillar just below the decorative bell capital and this rust must be the oldest rust on the pillar. Moreover, this rust has been undisturbed because this location of the pillar is inaccessible to the public. Therefore, this important aspect must be borne in mind while viewing the earlier rust analyses as all of them were conducted on rusts obtained from the lower regions of the pillar. The rust forming in the lower regions cannot be in the undisturbed condition because of the custom of the visitors who try to encircle their arms around the pillar with their backs against the pillar in the belief that if such an act can be successfully completed, it brings good luck. A minute sample of the rust scraped from the area about 5 feet from the ground was found essentially amorphous and contained clayey and siliceous matter [21]. X-ray uorescent analysis of the same rust con rmed that there was R. Balasubramaniam, A.V. Ramesh Kumar / Corrosion Science 42 (2000) 2085± no free iron present in the rust. Based on these results, Bardgett and Stanners [21] proposed that the coating might have risen from repeated handling. This conclusion is not surprising as the rust sample was taken from the location where contact with humans was frequent. Ghosh [22] analyzed the rust obtained from the buried underground portion of the DIP and from the portion above the ground. He found that both the samples were contaminated with sandy and clayey matter. The rust samples were separated with a magnet and it was found that the DIP underground rusts contained 99.0% of magnetic matter, while the rust sample from the exposed area contained 65.7±80.1% of magnetic matter. X- ray di raction analysis of the rusts (conducted at Bern University [22]) revealed that the magnetic portion contained g-feooh (lepidocrocite), a-feooh (goethite) and Fe 3 x O 4 (magnetite), while the non-magnetic portion contained g- FeOOH and quartz. Interestingly, although the presence of a-fe 2 O 3 (maghmite) was not con rmed directly, it was opined that there might be some a-fe 2 O 3 present based on the asymmetry of di raction lines of Fe 3 x O 4 [22]. It is well known that magnetite formation is favored in restricted oxygen environments, such as soil burial conditions. Therefore, the higher proportion of magnetic matter in the underground rust must be due to a higher magnetite content in them. The identi cation of the oxides/oxyhydroxides in the rust samples taken from above the ground are indicative of the state of the rust in the region where the pillar comes into contact with humans. This is not the stable rust layer present on Fig. 4. X-ray di raction pattern of the DIP rust showing the intensities in the region 2y ˆ : 2094 R. Balasubramaniam, A.V. Ramesh Kumar / Corrosion Science 42 (2000) 2085±2101 the pillar. However, the above observation provide information about the rust layer in the region that was in human contact. Interestingly, a sample was cut from the bottom exposed region of the pillar for metallographic analysis [22] and the rust formed on this freshly exposed surface after 1.5 years of exposure was studied by XRD by Lahiri et al. [23]. They identi ed g-feooh (lepidocrocite) and a-feooh (goethite) in the rust, with the major phase being g-feooh (lepidocrocite) based on the di raction intensities. While the above studies provide information about the state of the rust in the ground level region of the pillar, the XRD results provided in the present study provides information about the state of the stable oldest rust on the pillar. Nevertheless, the information from the above studies can be utilized to construct the sequence of events leading to the formation of the protective passive lm on the DIP [24] Fourier transform infrared spectroscopy It is established that IR spectroscopy can be utilized to understand the process of rusting of steels. Misawa et al. [5,6] studied the mechanism of atmospheric rusting and the protective amorphous rust on low allow steel by using IR absorption spectroscopy. It has been pointed out by them that the absorption band at higher wavenumber region is due to OH stretching and at lower wavenumber is because of Fe±O lattice vibration. Misawa
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks