Raman Spectra and Ancient Life: Vibrational ID Profiles of Fossilized (Bone) Tissues

Abstract

Raman micro-spectroscopy is a non-destructive and non-contact analytical technique that combines microscopy and spectroscopy, thus providing a potential for non-invasive and in situ molecular identification, even over heterogeneous and rare samples such as fossilized tissues. Recently, chemical imaging techniques have become an increasingly popular tool for characterizing trace elements, isotopic information, and organic markers in fossils. Raman spectroscopy also shows a growing potential in understanding bone microstructure, chemical composition, and mineral assemblance affected by diagenetic processes. In our lab, we have investigated a wide range of different fossil tissues, mainly of Mesozoic vertebrates (from Jurassic through Cretaceous). Besides standard spectra of sedimentary rocks, including pigment contamination, our Raman spectra also exhibit interesting spectral features in the 1200-1800 cm^-1 spectral range, where Raman bands of proteins, nucleic acids, and other organic molecules can be identified. In the present study, we discuss both a possible origin of the observed bands of ancient organic residues and difficulties with definition of the specific spectral markers in fossilized soft and hard tissues.

Keywords: bone; diagenesis; fossilized tissue; micro-Raman spectroscopy and imaging; spectral markers.

Figure 1

Representative Raman microspectroscopy of the fossilized bone samples (GSIM DB2; cross-sectional fossil reptile bone fragment collected by the Slovak-Iranian Paleontological Expedition in 2017; southern Laurasia/northern Gondwana (Iran), Mesozoic (Jurassic)). (a) Single-point Raman analysis of the sedimentary rock section (1) revealed common minerals including calcite and quartz. Two other Raman spectra characterized the bone thin section: (b) the iron oxide red pigment, hematite, was identified in the dark-red bone regions (2), and (c) undefined carbonate compounds were recorded in the whitish areas (3). Spectra were recorded using a 785 nm-laser excitation and are depicted directly (without any additional post-measurement processing) as an average of the spectra recorded at different points of the corresponding sections ((1) sedimentary rock; (2) and (3) bone) of the analyzed fragment.

Figure 2

(I) Representative micro-Raman spectra acquired on bone fragments of Rhynchotus (a representative of the only extant ratites that can fly, Tinamiformes, Brazil) (a), and Tinamus (Tinamiformes, Brazil) (b); subfossilized bone fragments (Mammalia, non-recent Holocene, Slovakia) (c,d); Deinocheirus (Ornithomimosauria, Upper Cretaceous, Mongolia) (e); and cross-sectional bone fragment of an extinct reptile (GSIM DB2, Jurassic, Iran) (f). Spectra were acquired using a 785 nm-laser excitation. Each illustrated spectrum represents an average of the spectra recorded at different points of the compact bone selected throughout the entire bone surface; no spectral treatment (except normalization) was applied to the recorded data. Spectra were normalized for clarity of presentation to the bands at ~1280 and 430 cm⁻¹. (II) Typical Raman spectrum of bone tissue showing the major bands and the corresponding assignments: most bands could be assigned to mineral phosphate, carbonate (below 1100 cm⁻¹), or the organic phase, i.e., matrix collagen (above 1100 cm⁻¹). The presented spectrum is an average spectrum of the Figure 2I(a,b) spectra. Finally, the spectrum was background corrected for clarity of presentation. (ν: stretching mode; δ: deformation-bending mode; Phe: phenylalanine).

Figure 3

(I) Micro-Raman spectra of the selected subfossil and fossilized fragments of bones, an eggshell, and a tooth. For more details, see Table 1. Spectra were acquired using a 785 nm-laser excitation. All spectra were baseline-corrected and normalized to the band at ~1280 cm⁻¹. Bands marked with an asterisk correspond to the phosphate and carbonate bands (Ft and Fes, respectively). (II) Curve fitting analysis of the 1150–1800 cm⁻¹ Raman region characteristic for the fossilized reptile bone sample GSIM DB2 (Fb3). The red curve represents the resulting fitted curve displayed together with the original data (dotted-wine curve) and the Raman spectrum of the recent bone (dashed-orange curve). The peak characteristics of individual components are summarized together with tentative assignments in Table 2. The most intense/relevant peaks are highlighted and numbered. Green lines mark frequency positions of two commonly broad bands from carbonaceous material.

Figure 4

(a) Raman spectra of unknown carbonate compounds recorded at different positions of the fossilized tibia of Deinocheirus mirificus. The inset visualizes the variation of intensity of the band at ~1280 cm⁻¹ as a function of the position. (b) Illustrative light microscopy images (LMI) and the corresponding Raman microscopy images (RMI) of the selected regions of the fossilized reptile bone sample GSIM DB2. Reconstructed false-color Raman mapping images show distribution of the discussed carbonate signal across the scanned areas, since the colors reflect the integrated intensities of a series of Raman bands recorded within the 1200–1800 cm⁻¹ spectral range. Thus, the brightest areas correspond to the most intense Raman signal. Circumferential mineral walls surrounding the outer border of each osteon, known as the ‘cement lines’, can also be recognized (indicated by the red arrow-heads). (c) False-color Raman mapping images of the cross-sectional tibia fossilized bone fragments of Yanornis reconstructed on base of two principal spectral components: (SPECTRUM 1—yellow spectrum) the Raman spectrum related to the discussed carbonate signal and (SPECTRUM 2—grey spectrum) the fluorescence background spectrum. No more Raman spectra (except the Raman spectra of the glass microscope slide and the epoxy resin) were identified in the analyzed samples. Spectra were acquired using a 785 nm-laser excitation.