Analytical Methods Committee, AMCTB No. 106
First published on 11th June 2021
Fourier transform infrared (FTIR) spectroscopy is ideally suited to the cultural heritage sector due to the ability to apply it minimally or non-destructively with limited sample preparation, fast analysis times (spectra can be obtained in a matter of minutes), relatively low cost, and relative ease of use. FTIR has been applied to answer a range of archaeological research questions through analysis of both organic and inorganic materials. Examples include determining the firing temperatures of archaeological clays or identifying types of textile.
Fig. 1 Organic materials excavated at the Bronze Age site of Must Farm (East Anglia) displayed exceptional levels of preservation. Left: an expanse of timbers, part of the collapsed settlement; right, top: yarn wound around dowels; right, bottom: a complete axe haft. Images from Knight et al., 2019, reproduced under a creative commons license.4 |
IR instruments with Fourier transform processing (FTIR) are the most commonly used in heritage applications, as the improved signal-to-noise ratio allows the analysis of subtle changes in spectra as well as increased speed of analysis.1 This means that changes in the chemistry and composition of organic archaeological materials caused by decay can be determined through assessment of the characteristic absorption peaks. By assigning peaks to individual components of a material, its relative composition can be measured semi-quantitatively through comparison of the relative heights or areas of these peaks. Peak assignment is typically done with the aid of in-house or commercially available databases.2 Changes in the position or shape of the peaks can also be related to specific changes driven by decay. Comparison with modern reference samples gives an indication of the extent to which the material has degraded. Decay levels in different samples can also be compared, allowing an assessment of differences in preservation between, for example, samples that have been stored under different conditions or undergoing different conservation methods.
Fig. 2 Schematic showing the operation of a typical FTIR-ATR instrument, illustrating how damage to a sample could be caused by the application of pressure. |
A range of bench top FTIR instruments are available. These are increasingly small and transportable, and can operate outside dedicated labs, enabling in-field use and the rapid gathering of data. Whilst many are connected to a desktop or laptop computer in order to operate the instrument and gather spectra, some can be controlled via a built-in computer.
FTIR is an attractive method of analysis for several reasons including its speed, low cost, and limited sample preparation requirements. It can be used in a minimally destructive manner by taking only very small amounts of sample (several milligrams for ATR), or completely non-destructively if the nature of the sample allows. FTIR analyses only a very tiny fraction of an object, which may not be representative of the bulk, particularly in cases where heterogeneous decay has occurred. However, the speed of analysis allows this to be circumvented by collecting multiple measurements. Whilst FTIR instruments are relatively straightforward in terms of operation, data interpretation still demands expert knowledge, in particular requiring the correct assignment of spectral peaks to maximise the value of the data. Methods of data analysis may also vary between studies, for example in terms of how data is manipulated prior to interpretation, meaning that comparison between studies is not necessarily straightforward. Furthermore, FTIR data does not offer the detailed structural characterisation that analysis by, for example, nuclear magnetic resonance (NMR) or analytical pyrolysis would allow (see AMC Technical Brief 85).3
Chemometric methods (such as partial least squares modelling or principal component analysis (see AMC Technical Brief 100))3 are increasingly applied to FTIR data. These methods have the potential to improve the sensitivity of detection of changes in different materials, as several variations in the spectra can be accounted for simultaneously.
Wood is composed of three major biopolymers: cellulose, hemicelluloses, and lignin. These are closely bound within the wood structure alongside small amounts of non-structural components (e.g. tannins, resins and oils). In an FTIR spectrum of wood, certain peaks can be solely attributed to one of these components. Others contain contributions from more than one polymer and must be interpreted with caution. The relative heights or areas of these identified peaks are typically used to derive a lignin:cellulose (L:C) ratio (Fig. 3). Due to the greater vulnerability of cellulose to biological decay, this ratio is assumed to increase with increasing wood deterioration.
Other indicators of decay such as the oxidation of CO groups, cellulose hydrolysis, and demethylation of lignin can be identified by changes in the position or shape of the peaks related to these functional groups (Fig. 3).
FTIR analysis of bone yields information on both the organic and inorganic components (Fig. 4). Hydroxyapatite content is represented by peaks at 1410 cm−1 relating to the carbonate component and at 1020 cm−1 relating to phosphate. Hydroxyapatite deterioration is indicated by an increase in carbonate content, determined by comparing both peaks. Comparison with the amide stretch at 1640 cm−1, attributed to collagen, provides a measure of the organic:inorganic ratio.
Hydroxyapatite is thought to increase in crystallinity with increasing deterioration. This can be assessed by analysis of the doublet at 560 and 610 cm−1 which becomes increasingly split with increasing crystallinity (known as the infrared splitting factor (IRSF); illustrated in Fig. 4).
FTIR can alternatively be used to identify textiles based on their chemical composition. This allows the analysis of even decayed samples and may yield additional information such as the identification of dyes and treatment processes. There are marked differences between the spectra from plant-based fibres which are primarily cellulose, and animal-based fibres which are mainly protein. It is therefore relatively easy to distinguish between the two, even when chemical or biological deterioration has occurred. Through more careful examination of the ratios between characteristic peaks, it may also be possible to distinguish between different cellulose-based fibres (e.g. cotton which contains no lignin, and jute which contains around 11% lignin will have quite different spectra). Statistical analysis such as principal component analysis (PCA) may further help distinguish between fibre types when differences in spectra are subtle. Although animal-based fibres have different relative amino acid compositions, FTIR is unlikely to be able to distinguish between types (for example wool vs. silk). Methods such as pyrolysis gas chromatography mass spectrometry (py-GC/MS) or amino acid analysis by high performance liquid chromatography (HPLC) may be more suitable.
Ref. 5–8 provide further information on the use of FTIR in the analysis of archeological materials.
Dr Kirsty High, NERC Knowledge Exchange Fellow, Department of Chemistry, University of York
This Technical Brief was prepared for the Analytical Methods Committee (AMC), with contributions from members of the AMC Heritage Science Expert Working Group, and approved by the AMC on 25 th January 2021.
This journal is © The Royal Society of Chemistry 2021 |