Raman and Surface-Enhanced Raman
Spectroscopy Applied in Art Conservation
Name: Amber Kamman Student ID: 12931977 MSc. Literature Project
Daily supervisor: Iris Groeneveld
Examiners: prof. dr. ing. Maarten van Bommel and dr. Freek Ariese Date: 20-02-2021
2
Abstract
This literature review shows the value of invasive Raman spectroscopy techniques determination of colourants and some of their degradation products in works of art. Raman can identify the chemical composition of colourants in complex samples and identify which painting technique is used. This knowledge is of great value for proper restoration. Both organic and inorganic components can be identified and quantified with various Raman spectroscopy techniques.
The identification is done with a limited but growing amount of reference spectra. Surface-enhanced Raman spectroscopy (SERS) is a powerful technique to measure components at low concentrations. There are currently developments in the field of Raman spectroscopy such as micro-Raman, micrometre-scale spatially offset Raman spectroscopy (micro-SORS) and ultraviolet-surface enhanced Raman spectroscopy (UV-SERS).
Ultimately, it could be concluded that Raman spectroscopic techniques are of great added value in the field of art conservation. However, the Raman spectroscopy techniques cannot complete the complete picture alone. Other techniques such as high-performance liquid chromatography (HPLC), X-ray fluorescence (XRF) and infrared spectroscopy are needed to fill in the gaps. A complete picture can be generated by combining multiple techniques with multivariate data analysis.
Keywords: Raman, Surface-Enhanced Raman Spectroscopy, SERS, Art, Pigments, Dye, Colorants, Conservation
3
Acknowledgements
I would like to thank Iris Groeneveld and Maarten van Bommel for guiding me through writing this literature study, for their patience and helpful feedback. Furthermore, I would like to thank Freek Ariese for the enthusiastic introduction to SERS. His lectures sparked my interest to learn more.
Abbreviations
AFM Atomic Force Microscopy
ATR Attenuated total reflection
CCD Charge-coupled device
FT-IR Fourier-transformed Infrared spectroscopy
GC-MS Gas chromatography-mass spectrometry
HPLC High-performance liquid chromatography
LOD Limit of detection
LSP Localized Surface Plasmon
Micro-SORS Micrometre-scale spatially offset Raman spectroscopy
MVDA Multivariate data analysis
NIR Near-Infrared
RRS Resonance Raman Spectroscopy
SEM-EDX Scanning electron microscopy energy-dispersive X-ray spectroscopy
SERS Surface-Enhanced Raman Spectroscopy
SERSS Surface-Enhanced Resonance Raman Spectroscopy
TERS Tip-Enhanced Raman Spectroscopy
UV-SERS Ultraviolet-surface enhanced Raman spectroscopy
UV-VIS Ultraviolet-visible
W&N Winsor and Newton
XRD X-ray diffraction
4
Table of content
Abstract ...2 Acknowledgements ...3 Abbreviations ...3 1. Introduction ...5 2. Method...73. Art, degradation and components of interest ...8
3.1 Art and degradation ...8
3.2 Dyes ...8
3.3 Pigments ...9
3.4 Surfaces, binders and mordant...9
4. Raman spectroscopy applied in art conservation ... 10
4.1 Theory ... 10
4.2 Equipment ... 12
4.3 Resonance Raman Spectroscopy ... 14
4.4 Applications for measuring pigments ... 15
4.5 Critical review of Raman ... 18
5. Surface-Enhanced Raman Spectroscopy applied in art conservation ... 19
5.1 The technique... 19
5.2 Surface Enhanced Resonance Raman Spectroscopy ... 20
5.3 Decision of substrates for SERS ... 20
5.4 Applications for paint, pigments and other organic components ... 21
5.5 Critical review of SERS ... 24
6. Other Raman techniques applied in art conservation ... 25
6.1 Tip-Enhanced Raman Spectroscopy ... 25
6.2 Alternative Raman Spectroscopy techniques ... 26
7. Conclusions ... 28
7.1 Recommendation for further research ... 28
Literature ... 29
5
1. Introduction
Art is a big part of our cultural heritage, it is a symbol of beauty, religion, status and more. Colours are fundamental to obtain a deeper understanding and appreciation of a culture and society’s history. More knowledge can be gathered by studying the pigments and dyes in these objects. It helps with authenticating art, by comparing it with work from the same artist. Furthermore, it can render the initial colours of a faded art piece, identify painting techniques and estimate the date of creation [1]. The fading of art is caused by degradation processes. Degradation through time and other factors can hinder the correct interpretation of the artist’s intention with the art pieces. This can be prevented by learning more about the original state of the art pieces.
Figure 1: ‘For to Be a Farmer’s Boy’ by Winslow Homer 1887. (A) The current appearance of the painting. (B) Digital simulation of the original colour. Source [1].
The watercolour painting in Figure 1 shows what it can yield to research the degradation of paint pigments. The top one (A) is the current state and the bottom one (B) is a digital simulation after recreating the original colours with the acquired chemical knowledge [1]. This can give the painting a completely different 'feeling' and thus meaning.
Detecting components at a very low concentration from a limited sample size is essential with art preservation in mind [1]. Techniques such as Raman spectroscopy and Surface Enhanced Raman Spectroscopy (SERS) are promising for the analysis of both organic and inorganic
6 materials in art. Art objects are especially challenging, sampling is restricted and sometimes not even admissible. This can be challenging since art exists in different forms; a multitude of materials and layers makes the sample rather complex. Degradation of the materials used makes identification even more problematic due to the presence of many unknown compounds.
This literature review aims to give a state-of-the-art overview of Raman and SERS analysis of samples taken from different types of objects. A review will be given on the possibilities and limitations of these techniques in the research of a multitude of art objects. Primarily concentrated on organic materials such as dyes, organic pigments and binding mediums. The focus is on the application of the techniques in art and applied art. Non-invasive analysis directly on the object is not part of this literature study.
This literature report starts with background information about art, degradation and the components of interest (chapter 3). Then the theory and equipment of Raman spectroscopy are explained (chapter 4). This chapter also includes resonance Raman spectroscopy. At the end of this chapter, two case studies on the application of Raman in art research are discussed and critically reviewed. Chapter 5 follows the same structure as chapter 4 but focuses on SERS. The decision between substrates for SERS are explained in this chapter. Chapter 6 briefly explains and discusses other techniques, including Tip-Enhanced Raman Spectroscopy. In chapter 7, a conclusion to the aim of this research thesis is given, and suggestions for a follow-up study are made.
7
2. Method
This literature study is not comprehensive but provides an overview of several Raman applications, their advantaged and disadvantages, and further prospects in research on art objects1. The studies referenced and used for this literature review were selected based on
the following criteria:
▪ The focus is on the application of Raman spectroscopy techniques in art and applied art.
▪ The methodologies for measuring (degradation of) cultural heritage were successful with Raman techniques. When additional techniques were applied, clear argumentation and motivation are given.
▪ Sample preparation and measurement methodology were described meticulously. ▪ The advantages and possible disadvantages of the Raman techniques used were
discussed.
▪ Non-invasive analysis directly on the object is not included in this literature study.
8
3. Art, degradation and components of interest
This chapter discusses the impact of colours on art objects and how this could change due to degradation over time. Several causes of degradation and processes are elaborated. In the second part of this chapter, more information will be given about the components of interest and their complexity.
3.1 Art and degradation
Works of art connect and inspire people and teach us about history. Visual art exists in a lot of different forms, such as ceramics, drawings or paintings. An important aspect in visual art, that has a big influence on its interpretation, is colour. Over time, the colours of a piece of art can change as can be seen from the artwork of Van Gogh in Figure 2. In his painting ‘de slaapkamer’, Van Gogh has used different pigments that have changed over time. The most striking change is the colour of the walls. For this colour, Van Gogh used a blue pigment and cochineal red pigment. The cochineal red pigment is sensitive to light and overtime only the more stable blue pigment remained [2].
Figure 2: Vincent van Gogh's painting 'de slaapkamer'. Current condition (left) and s digital reconstruction of the initial colours before degradation (right). Source [3].
Factors that can cause these changes include, among others, biodegradation [4], photochemical degradation (non-radiative processes, radiative processes, photooxidation reactions, etc.) [5], indoor climate [6], a poor choice of the raw material used or preparation conditions [7]. Depending on the environment in which the art object is made, where located and/or stored, the different processes may have played a larger or smaller role.
3.2 Dyes
Natural dyes were used all over the world for centuries. These dyes are often used because of their brightness, non-toxic and soothing nature. They naturally occur in bark, fruits, plants, roots and seeds. These contain organic compounds that absorb visible light [8]. Natural dyes are gradually used less and less since William Henry Perkin made the first synthetic dyes in 1856 [9]. Dyes are soluble, thus they can be applied as solutions, which is different from pigments which are by definition insoluble in the medium in which they are applied.
9
3.3 Pigments
Pigments are compounds that are used to give material a colour [9], based on the period certain components would be used. They are applied as solid particles, finely ground, mixed with a liquid. These particles have a size between 10-20 micrometre. Pigments may be organic or inorganic. Inorganic pigments are often brighter and tend to last longer. For centuries, biological sources (organic pigments) were used. Other natural sources such as minerals could be used as both organic and inorganic pigments. Currently, mostly inorganic or synthetic organic pigments are used. A synthetic organic pigment is derived from petrochemicals. Inorganic pigments are naturally found on earth or produced through chemical reactions, such as oxidation.
More research is necessary on pigments in paint due to their complexity. Pigments have been prepared with various methods throughout history. The formulations change differently through degradation or interactions with other components in the paint [10]. The identification of the pigments (and binder) employed may contribute to assess valuable information for proper restoration and/or conservation treatment [11].
3.4 Surfaces, binders and mordant
Art is an extensive concept and exists in many forms. The colours used in art are applied to pottery, linen canvases, murals, glazes and so on. To be able to apply colours, some kind of binders could be required. Examples of binders are beeswax, glues, plant gums and oils [12]. Without these binders, there could be insufficient adhesion to the surface.
A mordant is a substance that is used to bind dyes on fabrics. Mordants that are frequently used mainly include alumina [12]. The dye and mordant form an inorganic substrate that can attach to a surface. These inorganic substrates are also called lake pigments.
10
4. Raman spectroscopy applied in art conservation
As mentioned previously, research on art objects can be challenging. High-performance liquid chromatography (HPLC) is often used for the determination of organic dyes. The main limitation of this technique is the required big sample size. For pigments, in which HPLC cannot be used, ultraviolet-visible (UV-VIS) spectroscopy is applied. However, this technique often has insufficient fingerprint capacity to determine colourants in complex matrices [13].
Spectroscopic techniques like Raman are promising for the analysis of both organic and inorganic materials in art. In this chapter, the theory and equipment of Raman spectroscopy and resonance Raman spectroscopy (RRS) are explained. At the end of this chapter, two case studies on the application of Raman in art research are discussed and critically reviewed.
4.1 Theory
By irradiating a molecule with a light source, energy can be scattered or absorbed. There are essentially two types of scattering occurring in Raman spectroscopy: the common elastic Rayleigh scattering, and the less common inelastic Raman scattering (divided into Stokes and Anti-Stokes scattering). This scattering is at most 0.001% of the intensity of the source. Hence it may initially seem that Raman scattering more difficult to detect than the IR vibrational bands [14]. Moreover, Raman radiation is in the visible or near-infrared regions for which more sensitive detectors are available [15].In Figure 3, the different types of scattering that occur in Raman are schematically shown with the corresponding energy diagram.
Figure 3: Schematic overview of the different types of scattering occurring in Raman spectroscopy and corresponding energy diagrams. Source: [16].
11 When an electron in the ground state (g0 in Figure 3) is excited to a virtual excited state with an energy of hv0 and falls back to the same ground-state (g0), no difference in the energy of the photon is emitted, and nothing changes the frequency of the light. This phenomenon is called Rayleigh scattering and is an example of elastic scattering because it does not change frequency. Rayleigh scattering has by far the highest relative intensity in the Raman spectrum and says nothing about the chemical composition of a sample [15, 16, 17].
The types of scattering that are considered during Raman are the inelastic Stokes and Anti-Stokes scattering. The incoming energy hv0 is not the same as the photon that is released
during the relaxation. In Stokes scattering, an electron in the ground state (g0 in Figure 3) is excited to a virtual excited state with an energy of hv0 and then falls back to a higher vibration
level (g1 in Figure 3). The energy of the emitted photon is then defined as hv0-hvs and has a
lower frequency than the radiation of the source. Because the frequency changes with respect to the source radiation, it is called inelastic scattering. Stokes scattering is lower in relative intensity than Rayleigh scattering, but higher than Anti-Stokes scattering [15, 16, 17].
Anti-Stokes scattering occurs if the electron is already at a higher vibrational level (g1 in Figure 7), then is excited to a virtual excited state having an energy of hv0+hvs and then falls back to
the ground state (g0 in Figure 3). The energy of the emitted photon is then defined as hv0 +
hvs and has a higher frequency than the radiation of the source. Again, because the frequency
changes with respect to the source radiation, this is also called inelastic scattering. Since initially (at ambient temperature) relatively few molecules are already in an excited state, the intensity of the Anti-Stokes scattering is relatively the lowest on a Raman spectrum. However, the intensity of the Anti-Stokes scattering can be increased by increasing the temperature during the measurement in order to get more molecules in this excited state as follows from the Boltzmann equation [15, 16, 17].
The scattering of the light is therefore expressed as the difference in wavenumber between absorbed light and the emitted radiation. This shift is completely independent of the wavelength of the light source due to the presence of virtual energy levels [15].
In essence, a Raman spectrum is mirrored with the same peaks to the left and right of the Rayleigh shift, see Figure 4. Furthermore, the example spectrum also clearly shows the difference in the relative intensity of the peaks.
Figure 4: Example of a Raman spectrum with the Rayleigh, Stokes and Anti-Stokes shifts. A spectrum was measured with a 532 nm laser of CCL4 [18].
When performing Raman analyses, the Stokes shifts are often considered instead of the Stokes shifts due to the higher intensity of the Stokes shift peaks. In some cases, the Anti-Stokes shifts should be used, especially if the sample to be analysed shows signs of
12 fluorescence. Fluorescence uses the same principles as Raman spectroscopy. Mostly fluorescence is observed after radiation with a light source, due to higher efficiency. This results in a lot of interference. Furthermore, the signals from fluorescence can cause disturbances in the Stokes shift portion of the spectrum. Anti-Stokes shifts do not experience interference from the fluorescence and can be used in this situation [16, 17].
4.2 Equipment
Figure 5 shows a schematic construction of a basic Raman spectrometer. It is important that the scattering is measured at an angle of 90 degrees from the laser to avoid interference from the laser source. The schematic representation clearly shows that the scattering perpendicular to the direction of the laser is picked up by the wavelength selector, after which the radiation is converted with a detector into data that can be read by a computer system.
Figure 5: Raman spectrometer, source [17].
The choice of the diffraction grating, or wavelength selector, depends on the form of Raman spectroscopy used. Traditionally, dispersive Raman used double or even triple grating monochromators, but today they are often replaced with a single monochromator in combination with holographic interference filters or "notch filters". These notch filters and interference filters are necessary to filter the Rayleigh scattering as much as possible to accurately measure the relatively weak Raman scattering [17]. A schematic representation of a dispersive Raman spectrophotometer is shown in Figure 6.
Figure 6: Dispersive Raman spectrophotometer, source [19].
The light that passes over the grating and/or filters is eventually collected on a detector. This detector is often a Charge-Coupled Device (CCD). This detector is a silicon multichannel array detector capable of detecting UV, visible and near-infrared light [20]. The sensitivity and high
13 quality are of great importance to be able to capture the relatively weak Raman signal in a representative manner [20].
CCDs can generate thousands to millions of individual detector units (or pixels) under the influence of light, which can be converted into a spectrum through a processing system. High-grade CCD detectors require a degree of cooling to use during the measurements often using Peltier cooling systems (and possibly liquid nitrogen cooling) [20]. The CCD is often exposed to light diffracted by a grating, which is shown schematically in Figure 7.
Figure 7: CCD detector, source [20].
Another way of Raman spectroscopy is Fourier-Transform Raman (FT-Raman) and uses a Michelson-Interferometer. This interferometer can measure the entire spectrum at once in the form of an interferogram by varying the position of the mirror in the interferometer. This contrasts with dispersing the spectrum into separate wavelengths. The position of the mirror is monitored very accurately with the aid of a HeNe reference laser [19]. A schematic representation of the interferometer in an FT-Raman spectrophotometer is shown in Figure 8.
Figure 8: Michelson-Interferometer, source [19].
The use of FT-Raman has advantages, especially the wide spectral window that can be obtained at high resolution and the high wavelength accuracy due to the calibration with the reference laser. Furthermore, multivariate models can be more reliably applied to FT data and the data is more consistent between instruments. A noteworthy advantage of FT-Raman is the use of a 1064 nm excitation laser which very effectively eliminates fluorescence, but the high laser energy of Raman lasers is often too much and can damage (organic) samples [19]. Recently, Fourier-Transform instruments have been overshadowed by advances in dispersive Raman. This is mainly due to the possibility to use lasers with another wavelength then
14 1064nm. Multiple laser wavelength options make optimisation for a wider range of compounds possible [21]. Additionally, the rapidly evolving optoelectronics market, while FT-Raman continues to use the ‘mature’ FT-IR platforms and Michelson-Interferometers. Furthermore, the application of ‘low-noise’ CCD detectors is quicker and often good enough for Raman measurements without the need for an FT-IR platform [19].
Raman scattering always occurs. A laser is necessary, it has a high enough intensity to produce a sufficient amount of scattering for measurements and good signal-to-noise ratios [17]. An overview of commonly used laser wavelengths and sources used in art application is given in Table 1.
Table 1: Common laser wavelengths and sources used in Raman spectroscopy in art applications. Source: [14].
Wavelength (nm) Source 413.1 Ar+ ions 457.9 Ar+ ions 473.1 Doubled Nd:YAG 488 Ar+ ions 514.5 Ar+ ions 532 Doubled Nd:YAG 632.8 He-Ne 647.1 Kr+ ions 780-785 GaAlAs diode 830 GaAiAs diode 1064 Nd:YAG
Fluorescence is the main limitation when measuring samples. In art, varnish layers are often strongly fluorescent. There are a few possibilities to overcome this problem. Firstly, selecting the best laser is essential. It is possible to overcome this fluorescence by selecting a laser of lower energy (thus higher wavelength). The energy could be so low that less or no fluorescence occurs [17]. Secondly, instead of Stokes scattering, anti-Stokes scattering could be used for observations of the Raman bands. The intensity of anti-Stokes bands are much lower, however, fluorescence does not occur in this region.
4.3 Resonance Raman Spectroscopy
As already discussed, Raman scattering is a relatively weak process. Through resonance enhancement, selectivity and sensitivity can be increased. This is achieved by utilizing a laser wavelength that is close to the molecule absorption wavelength. The lines from the most symmetric vibrations are enhancement up to a factor of 106 [17]. Thus, Resonance Raman
Spectroscopy (RRS) can be used to selectively pick out and positively identify a molecule in a matrix, see Figure 9. Electronic information about a molecule can be obtained from the intensities of the bands found in resonance, from the energy separations in overtone progressions, and from the overtone patterns that can be obtained [22].
15 Figure 9: Raman spectra that illustrate the resonance Raman effect. The spectra are of crocoite (a) and carotenoids (b) in corals, measured with a 473.1 and 632.8 nm laser. Resonance occurs when using the 632.8
nm laser in (a) and the 473.1 nm laser in (b). Source: [14] .
Another clear difference between normal Raman scattering and resonance Raman scattering is that in both cases the intensity is dependent on the fourth power of the frequency, but in resonance Raman scattering the intensity is also strongly dependent on how close the frequency of the used excitation is to that of the frequency of an allowed electronic transition [22]. See Table 2 for an overview of the main differences between Raman scattering and RRS. Table 2: The main differences between RS and RRS, reproduced from: [22].
Raman scattering Resonance Raman scattering
No overtones Overtones common
More modes observed in the spectrum Some modes selectively enhanced
No electronic information Electronic information present
Weak scattering Stronger scattering
Absorption will occur due to the close match between laser wavelength and absorption wavelength. Depending on the material, fluorescence and decomposition can occur. This is important to keep in mind while performing experiments. Decomposition could have a negative impact on further research of the same sample [22].
4.4 Applications for measuring pigments
Raman spectroscopy knows many applications in a wide variety of fields. Due to the underlying theory, it can be useful in measurements of various materials (organic and inorganic) in the gas, liquid and solid phase. Below, two applications of Raman spectroscopy in measuring pigments in art are elaborated. These two applications have been selected because they represent the possibilities and limitations of the technique. More applications will be mentioned afterwards. A critical review follows in the next section.
Roman wall paintings in The Satyr Domus [23]
Cerrato et al. have studied ancient Roman wall paintings in The Satyr Domus. The wall was made of mortar, a mixture of lime and sand. The surface of the wall paintings consists of a mixture of lime, powdered marble and powdered limestone. This mixture served as a white base layer, upon which the pigments were applied. A picture of the wall paintings is provided in the Appendix.
16 Figure 10: The ten fragments studied with various colourants. Source [23].
Ten fragments were used to identify the various colourants present, see Figure 10. These fragments cover a wide variety of colours. Raman spectra were used to identify Egyptian blue, goethite yellow, hematite red, pink (diluted red), green earth (glauconite and celadonite) and calcite white. Table 3 shows an overview of the pigments, its formula and peaks used for identification.
Table 3: Overview of identified pigments on Roman wall paintings in The Satyr Domus, source [23].
Name pigment Colour Formula Main peaks (cm-1)
Weaker peaks (cm-1)
Egyptian blue Blue CaCuSi4O10 430, 1085 1012, 989, 787, 765, 569,
462, 402, 375, 231, 192 164
Goethite yellow Yellow α-FeOOH 392 244, 299, 480, 554, 683
Hematite red Red α-Fe2O3 1320 220, 290, 410
Hematite red (diluted)
Pink α-Fe2O3 1320 220, 290, 410 &
1085, 712, 278, 152 Glauconite
green
Green (K,Na)(Fe3+,Al,Mg)
2(Si,Al)4O10(OH)2 590 270
Celadonite green
Green K(Mg,Fe2+)(Fe3+, Al)(Si
4O10)(OH)2 550 264
Calcite white White CaCO3 1085 712, 276, 152
A combination of multiple analytical techniques was used to identify chemical components and mineralogical phases. Scanning electron microscopy-energy dispersive X-ray spectroscopy (SEM-EDX) was used to examine the composition of the different layers. X-ray diffraction (XRD) and spectra databases were used to confirm the composition of the measured pigments. Minor contaminations found in the pigments revealed the preparation method used. UV-VIS spectroscopy and gas chromatography-mass spectrometry (GC-MS) were used to identify organic compounds. Components and decomposition products were found of bee wax, a component used as a pigment binder.
Late medieval objects from Norwegian churches [7]
Platania et al. have studied objects from the 15th and 16th-century form Norwegian churches. Three painted micro-samples from various locations were studied. The first sample came from a shrine in Bygland, the second from an altarpiece in Skjervøy and the last from a (now lost) shrine in Røldal. A picture of the medieval objects is provided in the Appendix.
A micro-sample was taken with a scalpel from the edge of already damaged parts of the paintings. The micro-samples were embedded in a resin and mounted in sample holders. Two Raman instruments were used during this research. One instrument was a confocal Raman
17 micro-spectrometer working in micro/single configuration. The laser yielded a second harmonic generation light, which had a wavelength of 370nm. Furthermore, a second Raman spectrometer with a 785 nm diode laser was adopted.
Figure 11: Raman spectra of the green layers, cross-sections of three various layers Excitation line is at 785nm. Source [7].
This research mainly focused on the green pigments found in the investigated layers. An umbrella term for these green copper-based pigments is Verdigris. These pigments show Raman bands at 946, 1050, 1350, 1442, 1598, 2855 and 2935 cm−1, see Figure 11. These bands were found in each sample. Table 4 shows an overview of the collected Raman bands, the assigned formulas and attribution.
Table 4: Characteristic bands collected from the Raman spectra measuring the green paint-layers of sample A, B and C. The strength of the peaks is assigned as follows: vs (very strong), s (strong), m (medium), w (weak), vw (very weak). Source [7].
Sample A (cm-1) Sample B (cm-1) Sample C (cm-1)
Assigned formula Attribution
127vs 127vs 127vs ν(Pb-O) lattice mode Lead tin yellow
196m 196m 196m Not assigned
457w 457w 457w Not assigned
946vw 947vw 946w ν(Ch2/ρ(CH2) Copper acetate monohydrate /
Copper carboxylates
1050m 1050m 1050s ν(CH3)/CO32- Copper acetate monohydrate /
lead white?
1087w 1087w 1088w ν(CC) Copper oleates
1130w ν(CC) Copper acetates / lead
palmitates / stearates
1190w δ(CH2) Lead oleates
1298m 1298m 1298m δ(CH2) Copper oleates
1309m δ(CH3) Copper acetate monohydrate
1442m 1442m 1442m ν(COO-)/ δ(CH
3) /
δ(CH2)
Copper acetate monohydrate / copper oleates / lead oleates
18
1458sh Not assigned
1540m νas(COO-) Lead carboxylates
1568sh Not assigned
1598m 1599m 1598m ν(COO-) Copper acetate monohydrate /
Copper carboxylates
1653sh 1650sh ν(C=C)/ ν(C=CH2) /
δ(OH)
Copper oleates / basic copper chloride
1740vw 1740vw 1740vw ν(C=O) Linseed oil
2720w ν(C-H) CH2 Not assigned
2811w Not assigned
2855sh 2856w 2855m ν(C-H) CH2 Copper acetate monohydrate
2935w 2934w 2935m ν(C-H) CH3 Not assigned
3525vw Basic copper acetates
3636vw Not assigned
The wings were analysed with complementary techniques including UV-VIS, SEM-EDX, ATR-FTIR and GC-MS. This strategy was chosen to identify both binders and pigments in the complex paint samples. The rationale for this approach will be further discussed in the critical review below.
4.5 Critical review of Raman
Complete picture
Raman makes it possible to identify which painting technique the artist used. The colours, material and painting technique used can provide more information about the time from which the artwork comes. Other equipment is needed to complete the picture. For example, XRF could be used to identify/verify the chemical composition of the pigments and mineralogical phases present [1]. FTIR could also be used for the identification of the pigment origin (preparation) or painting technique used [7].
Limitations
Raman spectroscopy is based on the interaction between photons and matter. This often implies that a sample needs to be taken, which can be as small as a few microns. The main advantage of Raman is that it is a non-destructive technique, thus the sample could be reused for additional experiments. The main downside of spectroscopy is that it is based on a weak process. Interference of fluorescence could mask the wanted peaks. Fluorescence is less of a problem when measuring inorganic and synthetic dyes. Organic and natural dyes are challenging. Since synthetic dyes were first made in the mid-19th century, conducting experiments on objects before this time could be challenging [1].
Developments
Developments in the Raman field include new hardware (e.g. fibre optics), software improvements (e.g. automated shifted baseline subtraction), new insights (e.g. micro-SORS) and new approaches (e.g. SERS). Considering the remarkable signs of progress made over the past years, new steps are set paving the way for more novel developments in this research domain [14]. There are also developments in the field of non-invasive Raman spectroscopy. Non-invasive Raman makes it possible to take measurements without taking samples, which could be advantageous in art research.
19
5. Surface-Enhanced Raman Spectroscopy applied in art
conservation
As described above, fluorescence can be a limiting factor in paint and pigments research. A technique such as Surface-Enhanced Raman Spectroscopy (SERS) could offer a solution here. This chapter elaborates on SERS, SERRS and pointers for a substrate, and then an application is discussed. At the end of this chapter, there is a critical review of the technique concerning art research.
5.1 The technique
A development in Raman spectroscopy is Surface-Enhanced Raman Spectroscopy. SERS is a technique based on a study carried out by Jeanmarie and Van Duyne in 1977 [24]. SERS involves obtaining Raman spectra in the usual way on samples that are adsorbed on the surface of colloidal metal particles (usually silver, gold, or copper) or roughened surfaces of pieces of these metals, see Figure 12. Because Raman signals are relatively weak, applying SERS is a way to amplify these weak Raman signals.
The total enhancement originates from two mechanisms. The electromagnetic (EM) effect and the electrochemical (EC) effect. The EM enhancement relies on a local field interaction between a metallic surface and a Localize Surface Plasmons (LSP) forming a strong electromagnetic field [25]. This occurs when the substrate is excited by light and causes an enormous increase in the size of the induced dipole. This consecutively causes a strong increase in inelastic scattering [26]. The EM enhancement decreases rapidly with distance. The EC enhancement is weaker and requires that the molecule is chemically adsorbed on the surface. The interaction between the metallic surface reduces the change in polarizability. The orientation of the molecule and distance to the surface influence the degree of enhancement. Overall, the EM effect has a contribution up to 1010 and the EC effect has a contribution up to
102.
Figure 12: Illustration of particles on a silver substrate. The occurring Raman scattering is enhanced. Source [27] SERS sees application due to the high sensitivity and the possibility for very selective molecular identification. The Raman lines of the adsorbed molecule are often enhanced by a factor of 103 to 106. When surface enhancement is combined with resonance enhancement
technique, the net increase in signal intensity is roughly the product of the intensity produced by each of the techniques. Consequently, detection limits in the range of 10-9 to 10-12 M have
been observed [17].
Art materials often show strong fluorescent backgrounds in de near-infrared (NIR) and visible region. The metal substrates used in SERS act like quenchers for the occurring fluorescence. Thus, identifications of analytes in art materials, such as pigments, is possible [28].
20
5.2 Surface Enhanced Resonance Raman Spectroscopy
Surface Enhanced Resonance Raman Spectroscopy (SERRS) combines surface enhancement with molecular resonance enhancement as described in 4.3 Resonance Raman Spectroscopy. Enhancement up to 1014 have been reported [22]. The adsorption of the
pigments on the metal substrates are efficient fluorescent quenchers. The higher enhancement makes it possible to use a lower laser power and shorter accumulation time. Which reduces the chance that photodegradation occurs.
5.3 Decision of substrates for SERS
The decision on which substrates to use for SERS measurements is important. The substrates are responsible for the success of SERS, because of their interaction between the substrate and the adsorbed molecules. The morphology of the surface is determinative, a smooth surface results in no signal enhancement and a rough surface results in a good signal. Various materials can be used, each with advantages and disadvantages in the search for an optimal measurement result.
Figure 13: Estimate wavelength ranges (nm) of Ag, Au, and Cu substrates to measure SERS. Figure is reproduced from [28].
Primarily, gold (Au) and silver (Ag) are used as SERS substrates, because these materials are non-reactive with oxygen molecules. Copper (Cu) is subordinate due to its reactivity i.e. instability in air. All three metals have localized surface plasmon resonances (LSPRs) that cover most of the visible and NIR wavelength range, where most Raman measurements occur, also making them convenient to use, see Figure 13. Further enhancement can be achieved through research about different shapes and new plasmonic materials.
Other metals besides silver and gold have been explored. These include the alkali metals, aluminium, gallium, indium, platinum, rhodium and metal alloys that have been applied as plasmonic substrate options for SERS. The main problem of some of these materials is their reactivity with air [28]. When substrates are developed that can overcome the reactivity with air, new pathways for SERS substrates would be possible.
Recently, SERS substrates of new materials like graphene [29] and quantum dots [30] are reported. These materials do not fit the traditional definition of SERS substrates and have not yet been applied in research on colourants in art, but could be worthy of further investigation.
21
5.4 Applications for paint, pigments and other organic components
Winslow Homer’s ‘For to be a farmer’s boy’ [13]
Winslow Homer made the watercolour painting ‘For to Be a Farmer’s boy’ in 1887, see Figure 14. He was known for his colour choices that maximize the visual impact of the painting. Due to photochemical degradation, some colours changed, which could lead the viewer to perceive something different from the initial intention of the painting.
Figure 14: Winslow Homer watercolour painting ‘For to Be a Farmer’s boy’. The current condition (left), after digital reconstruction (right). Source: [13].
The red box in the upper left corner was examined with a high magnification stereomicroscope, see Figure 14. The pigment grains were visualized, see Figure 15 (b), and their colours were documented. Between 5-15 pigment grains were taken from the sample region, (a) in Figure 15, with a tungsten needle. A colloid paste was applied to the grains. SERS spectrum was taken from seven grains, three of these grains are highlighted in Figure 15 (b). A 632.8 nm He-Ne laser was used.
Homer was known for using watercolour washes from Winsor and Newton (W&N). A catalogue published in 1887 served as a source for making SERS reference spectra. Some of these reference spectra turned out too complicated to use in this research. Only the spectra for Indian purple, madder carmine, purple madder and burnt carmine were used. Brosseau et al. observed that pH values influenced the interpretation of the spectra [13]. To eliminate this variable, extra experiments at different pH values need to be performed.
Figure 15: (a) a close-up of the sample region. (b) picture obtained with the stereomicroscope. Pigment grains a, b and c are highlighted. Source: [13].
Grain a has a red/purple colour as observed with the stereomicroscope. The spectra of pigment grain a, cochineal, W&N burnt carmine and W&N Indian purple are shown in Figure 16. After interpretation, Brosseau et al. concluded that pigment grain a was mostly likely Indian purple.
22 Figure 16: SERS spectra of pigment grain a, Cochineal, W&N burnt carmine and W&N Indian purple. The asterisk indicate
bands due to citrate. Source: [13].
South Netherlandish tapestry [31]
Leona et al. studied a South Netherlandish tapestry form 1495-1505. The tapestry was mainly made of wool and some silk, silver and gilt wefts. A single red wool fibre was used for the measurements. The fibre was first treated with hydrogen fluoride vapour and then with potassium nitrate and Ag colloid. A picture of the tapestry is provided in the Appendix.
Figure 17: The tapestry fibre sample, reflected dots are the Ag nanoparticles, the scale is 100 µm. Source: [31].
The laser beam was focused on the silver-covered spot to obtain the SERS spectra, see Figure 17. For the excitation wavelength, a 785 nm laser was adopted. An objective of 20x was used on the sample.
23 Figure 18: Obtained spectrum from the tapestry fibre sample (UNI) and a reference spectrum of alizarin (AZ).
Source: [31].
Figure 18 shows the obtained spectrum from the tapestry fibre sample (UNI) and a reference spectrum of alizarin (AZ). There is a clear agreement between the sample and reference spectra. Furthermore, Leona et al. mentioned that the unidentified peaks may be caused by purpurin. At certain pH values, purpurin can be identified in the presence of alizarin. To confirm or rule this out, more experiments with other pH values are needed. An overview of the characteristic bands of alizarin and purpurin are given in Table 5
Table 5: Characteristic SERS bands of alizarin in a Raman spectrum measuring with a 785 nm excitation laser. The strength of the peaks is assigned as follows: vs (very strong), s (strong), m (medium), w (weak), sh (shoulder). Reproduced from: [31].
Dye source Structure SERS bands wavenumber/cm-1 (at 785 nm excitation)
Root of Madder Rubia tinctorum L. and other plants
1628.0 m, 1603.2 m, 1553.5 m, 1508.5 w, 1479.0 w, 1458.9 m, 1451.1 m, 1424.7 s, 1406.1 sh, 1323.8 s, 1288.2 s, 1275.7 sh, 1209.0 m, 1188.8 m, 1162.5 m, 1052.3 m, 1018.1 w, 903.3 w, 819.5 w, 763.6 w, 720.2 w, 684.5 w, 664.5 w, 663.3 m, 582.1 w, 506.04 m, 476.6 m, 451.7 m, 419.1 w, 399.0 m, 343.1 m, 312.1 w. - 1606.3 m, 1558.2 m, 1505.4 sh, 1475.9 m, 1424.7 sh, 1389.0 s, 1320.8 vs 1297.5 vs 1288.2 vs 1266.4 sh, 1212.1 s, 1157.8 m, 1066.2 m, 1032.1 w, 976.2 m, 908.0 w, 650.4 m, 610.0 w, 537.8 s, 464.1 m, 428.4 m, 385.0 m, 340.0 w
24
5.5 Critical review of SERS
Scope
Adsorption of pigment molecules onto colloidal silver or gold is quite an efficient process. Once the pigment is adsorbed onto the colloidal layer, the pigments may show a strong surface enhancement of scattering. SERS has advantages over Raman, in particular, the limit of detection (LOD) is much better. The peaks are easier to distinguish due to the enhanced Raman signals and the fluorescence quenching effect of the noble-metal surface [13]. After assigning the molecule, information about isomerization, orientation and degree of aggregation could be obtained [32].
Limitations
Considering that SERS is a surface method, it is highly sensitive to matrix effects and contaminations [33]. The vulnerability of the substrates and sample to get contaminated by molecules in the air is constrictive. It could distort results and can make identification more complicated. Besides, other random adsorbed molecules can broaden the measured Raman signals. Hence, it can be difficult to distinguish between which Raman signals are from the target molecule and which could be caused by contaminations [27].
Both Leona et al. and Brosseau et al. mention that pH values can influence the interpretation of the spectra. It may be necessary to run experiments at different pH values to do a correct interpretation [13, 31] Furthermore, SERS is an invasive method, thus application in art research might be limited when sampling is restricted. When taking a sample is restricted, Raman can be used instead.
25
6. Other Raman techniques applied in art conservation
6.1 Tip-Enhanced Raman Spectroscopy
Tip-Enhanced Raman Spectroscopy (TERS) uses the same underlying theory as SERS and combines the surface analysis technique Atomic Force Microscopy (AFM) with the molecular identification of Surface-Enhanced Raman, see Figure 19. The tip of the AFM probe is coated with precious metal, such as gold, to enhance the induced dipole [34]. Measurements with the probe are done directly on the surface of the sample. The probe is withdrawn from the surface without leaving behind residue [35]. The probe is connected to a spectrometer to accumulate the obtained data. The data is used to determine the chemical composition of the sample and to form an image of the surface [34].
Figure 19: Schematic depiction of TERS. Source: [28].
In SERS spectroscopy, the obtained signal is accumulated from a group of molecules. With TERS, single-molecule detection is possible. The spectra are obtained from the molecules that are within a short distance (few angstroms to several nanometres) of the tip. This makes it possible to tune the wavelength of the laser exactly to the corresponding frequency of the plasmon [36]. A very small sample surfaces is sufficient.
Application
Various methods have been developed that can measure electromagnetic (long-range, several nanometres) and chemical (short-range, few angstroms) enhancement regimes [36]. For example, TERS was used to identify indigo and iron gall ink on dyes paper. Measurements on both reference and historic sample were possible [35]. Another example is the application of TERS to detect an ink, brilliant cresyl blue, on a glass [36]. An increase of more than thirty times in Raman signal was visible when the top was in contact with the brilliant cresyl blue ink, see Figure 20. Both these applications show a proof-of-concept. TERS could be applied for the identification of colourants (such as brilliant cresyl blue, indigo and iron gall) in cultural heritage.
26 Figure 20: TERS spectra of brilliant cresyl blue. The Raman spectra were measured with (a) the tip retracted and
(b) the tip in contact with the sample surface. Source [36].
6.2 Alternative Raman Spectroscopy techniques
Micro-Raman spectroscopy [37]Micro-Raman spectroscopy is a technique that uses a microscope lens. This allows analysis of samples as small as one micrometre in diameter. Which is an advantage over Raman spectroscopy when measuring inhomogeneous samples. The smaller measurement area limits the amount of possible measured molecules. An example of an application is the analysis of pigment sample from an artist’s paintbox. The pigment samples were powdered. Figure 21 shows the spectrum that was used for the interpretation of white-2 pigment, zinc oxide.
Figure 21: Spectrum of the powered white-2 pigment from the paintbox, produced by a micro-Raman setup. Source: [37].
Micrometre-scale spatially offset Raman spectroscopy
Micrometre-scale spatially offset Raman spectroscopy (micro-SORS) is a technique that can determine the chemical composition of several layers below the surface. The principle is based on measuring the Raman signal from a point that is displaced from the point of excitation. This
27 is done on the sample surface or further down on the z-axis of illumination. Then it is possible to separate the spectral contributions of the different layers [38].
Figure 22: (a) sample of the red mantle of the terracotta sculptures 'Christ's disciples', white square indicates the analyzed area. (b) close-up image and stratigraphy. Source: [39].
Conti et al. applied this technique to a sample of the terracotta sculptures ‘Christ’s disciples’ [39]. This sculpture has been repainted many times, resulting in many overlapping layers of red pigments, see Figure 22. The Raman spectra in Figure 23 show that the composition of the pigment changes depending on the depth of the measurement. The uppermost late is mainly red lead, deeper layers reveal peaks of lead white and cinnabar.
Figure 23: Spectra for various distances. The lined markers emphasize the change of pigment composition. The red lead and cinnabar reference spectra were obtained with conventional Raman spectroscopy. Source: [39].
Ultraviolet-surface enhanced Raman spectroscopy [28]
Ultraviolet-surface enhanced Raman spectroscopy (UV-SERS) is an application that could be desirable for art applications. Measuring in the ultraviolet range enables resonance detection of a multitude of molecules. These molecules include protein residues.
UV-SERS comes with a few challenges. The first is finding a substrate material that supports the surface enhancement in the UV-region. Materials including Aluminium, Cobalt, Lead/Platinum, Rhodium, Ruthenium were explored. Enhancement factors of these materials only reach ~102, which is low compared to Ag and Au in the visible region.
The other challenges would be avoiding photodegradation and the need for highly efficient optical elements. UV-SERS could potentially broaden the scope of SERS if these challenges are overcome.
28
7. Conclusions
This literature review showed the value of Raman spectroscopy techniques in the analysis of pigments and some of its degradation products in works of art. Raman can identify the chemical composition of different coloured pigments in complex samples. Additionally, Raman makes it possible to identify which painting technique is used [1]. The colours, material and painting technique used can give an indication about the time from which the art object comes, whether it is authentic, and it can help with conservation and if necessary, restoration [1]. The application of Raman has been further developed over the years, including additions such as the CCD detector, new fibre optics, and increasing knowledge about signal treatment and statistical data analytics [14, 19].Both organic and inorganic components can be identified and quantified. These components can be labelled using reference spectra in databases, but these databases are mostly limited to references from pure components. The number of reference spectra of degradation products is growing but currently very minimal [1].
SERS is a powerful technique to measure components at low concentrations. It can provide rich structural information because of the high sensitivity of the technique. The SERS enhancement depends on the substrate preparation and varies across the sample due to inhomogeneity. This is a limit to its applicability and possibility of quantitative measurements. Techniques such as TERS can overcome these shortcomings with single-molecule measurements by replacing the substrates with a probe [36]. There are currently developments in the field of Raman spectroscopy such as micro-Raman [37], micro-SORS [39] and UV-SERS [28].
Combinations of multiple analytical techniques are necessary to obtain a complete picture. A SERS or Raman spectrum can only detect so much. Various techniques have been applied for the aforementioned intend, such as XRF and SEM-EDX for elemental analysis, micro-Raman spectroscopy for identification and IR spectroscopy for characterization [7]. MS-based techniques are used for the identification of proteinaceous binders, necessary because of the complexity of the samples [11].
Techniques that could be used for research on colour changes and ageing processes include infrared spectroscopy Fourier-transformed (FTIR) in the attenuated total reflection (ATR) mode, UV–Vis spectroscopy, gloss and colourimetric measurements. Multivariate data analysis (MVDA) could be applied to a combination of Raman data with other techniques. This could provide insight into the most complex samples [11].
7.1 Recommendation for further research
Future experiments could involve artificial weathering of paint reconstructions to, for example, understand the mechanisms of delamination [7] and the further development/expansion of databases. The spectra in these databases could be used as reference spectra of components and degradation products used in art [1]. Furthermore, this research could be extended to include experiments with non-invasive Raman options, such as portable/handheld Raman to research on location [40, 41]. It may be interesting to investigate the new developments in SERS substrates [29] and the combination of analysis techniques using MVDA.
29
Literature
[1] S. Lin, “Investigation into the Use of Surface-Enhanced Raman Spectroscopy (SERS) for Organic Dye Analysis,” Department of Materials Science and Engineering, Massachusetts Institute of Technology, 2015.
[2] Research project REVIGO, “Original colours of Van Gogh's paintings,” [Online]. Available: https://www.vangoghmuseum.nl/en/about/knowledge-and-research/completed-research-projects/revigo/research-results-revigo-paintings. [Accessed 02 December 2020].
[3] V. Bozhulich, “The Chemistry of Pigments and How Scientists Prevent Color Degradation,” InChemistry, 23 June 2020. [Online]. Available:
https://inchemistry.acs.org/content/inchemistry/en/atomic-news/chemistry-of-pigments.html. [Accessed 3 December 2020].
[4] H. Ravikumar, S. S. Rao and C. Karigar, “Biodegradation of paints: a current status,” Indian
Journal of Science and Technology, vol. 5, no. 1, pp. 1977-1987, 2012.
[5] R.-M. Ion, A. Nuta, A.-A. Sorescu and L. Iancu, “Photochemical Degradation Processes of Painting Materials from Cultural Heritage,” in Photochemistry and Photophysics - Fundamentals
to Applications, IntechOpen, 2018, pp. 161-178.
[6] K. Keune, R. P. Kramer, Z. Huijbregts, H. L. Schellen, M. H. Stappers and M. H. van Eikema Hommes, “Pigment Degradation in Oil Paint Induced by Indoor Climate: Comparison of Visua and Computational Backscattered Electron Images,” Microscopy and Microanalysis, vol. 22, pp. 448-457, 2016.
[7] E. Platania, N. L. Streeton, A. Lluveras-Tenorio, A. Vila, D. Buti, F. Caruso, H. Kutzkee, A. Karlsson, M. P. Colombini and E. Uggerud, “Identification of green pigments and binders in late medieval painted wings from Norwegian churches,” Microchemical Journal, no. 156, 2020.
[8] S. Sunder Sharma, K. Sharma, R. Singh, S. Srivastava, K. Bihari Rana and R. Singhal, “Natrual pigments: Origin and applications in dye sensitized solar cells,” Materialstoday: Proceedings, no. 2019, pp. 1-5, 2020.
[9] H. Bhajan Singh and K. Avinash Bharati, “Introduction,” in Handbook of Natural Dyes and
Pigments, Woodhead Publishing India, 2014, pp. 1-3.
[10] B. van Driel, K. van den Berg, M. Smout, N. Dekker, P. Kooyman and J. Dik, “Investigating the effect of artists’ paint formulation on degradation rates of TiO2-based oil paints,” Heritage
Science, vol. 21, no. 6, 2018.
[11] C. Calvano, E. Rigante, R. Picca, T. Cataldi and L. Sabbatini, “An easily transferable protocol for in-situ quasi-non-invasive analysis of protein binders in works of art,” Talanta, no. 215, p.
120882, 2020.
[12] F. Casadio, M. Leona, J. R. Lombardi and R. Van Duyne, “Identification of Organic Colorants in Fibers, Paints, and Glazes by Surface Enhanced Raman Spectroscopy,” Accounts of Chemical
Research, vol. 43, no. 6, pp. 782-791, 2010.
[13] C. L. Brosseau, F. Casadio and R. P. Van Duyne, “Revealing the invisible: using surface-enhanced Raman spectroscopy to identify minute remnants of color in Winslow Homer's colorless skies,” Journal of Raman Spectroscopy, no. 42, pp. 1305-1310, 2011.
[14] D. Bersani, C. Conti, P. Matousek, F. Pozzi and P. Vandenabeele, “Methodological evolutions of Raman spectroscopy in art and archaeology,” Analytical Methods, no. 8, pp. 8395-8409, 2016.
30
[15] J. R. Ferraro, K. Nakamoto and C. W. Brown, Introductory Raman Spectroscopy - Second Edition, Amsterdam: Academic Press, 2003.
[16] G. S. Bumbrah and R. M. Sharma, “Raman spectroscopy- Basic principle, instrumentation and selected application of characterization of drugs abuse,” Egyptian Journal of Forensic Sciences, no. 6, pp. 209-215, 2016.
[17] D. A. Skoog, J. F. Holler and S. R. Crouch, “Chapter Eighteen - Raman Spectroscopy,” in
Principles of Instrumental Analysis, Sixth Edition, Brooks/Cole, Cengage Learning, 2007, pp.
481-495.
[18] Edinburgh Instruments, “What is Raman Spectroscopy,” General Microtechnology & photonics. [19] F. LaPlant, “Lasers, Spectrographs, and Detectors,” in Emerging Raman Applications and
Techniques in Biomedical and Pharmaceutical Fields , Heidelberg, Springer, 2009, pp. 1-24.
[20] Horiba, “What is a CCD detector?,” [Online]. Available:
http://www.horiba.com/scientific/products/raman-spectroscopy/raman-academy/raman-faqs/what-is-a-ccd-detector/. [Accessed 25 November 2020].
[21] Horiba Scientific, “In practice What is the difference between dispersive Raman and FT-Raman,” Horiba Scientific, [Online]. Available: https://www.horiba.com/uk/scientific/products/raman-spectroscopy/raman-academy/raman-faqs/difference-between-dispersive-raman-and-ft-raman/. [Accessed 07 January 2021].
[22] E. Smith and G. Dent, Modern Raman Spectroscopy - a practical approach, Chichester, West Sussex: John Wiley & Sons, Ltd, 2005.
[23] E. J. Cerrato, D. Cosano, D. Esquivel, R. Otero, C. Jimémez-Sanchidrián and J. R. Ruiz, “A multi-analytical study of a wall painting in the Satyr domus in Córdoba, Spain,” Spectrochimica
Acta Part A: Molecular and Biomolecular Spectroscopy, no. 232, 2020.
[24] D. L. Jeanmarie and R. P. Van Duyne, “Surface Raman spectroelectrochemistry: Part I.
Heterocyclic, aromatic, and aliphatic amines adsorbed on the anodized silver electrode,” Journal
of Electroanalytical Chemistry and Interfacial Electrochemistry, vol. 84, no. 1, pp. 1-10, 1977.
[25] E. Le Ru and P. Etchegoin, “4. SERS Enhancement Factors,” in Principles of Surface Enhanced
Raman Spectroscopy and related plasmonic effects, Amsterdam, Elsevier, 2008, pp. 187-188.
[26] C. L. Haynes, A. D. McFarland and R. P. Van Duyne, “Surface-Enhanced Raman Spectroscopy,”
Anal. Chem., vol. 77, no. 17, pp. 338 A-346 A, 2005.
[27] M. Nasimuzzaman Mishuk, “Optical Mapping of Tip-Enhanced Raman Spectroscopy (TERS),” BRAC University - Department of Electrical and Electronic Engineering, Dhaka.
[28] B. Sharma, R. R. Frontiera, A.-I. Henry, E. Ringe and R. P. Van Duyne, “SERS: Materials, applications, and future,” Materialstoday, vol. 15, no. 1-2, pp. 16-25, 2012.
[29] X. Ling, X. Liming, Y. Fang, H. Xu, H. Zhang, J. Kong, M. S. Dresselhaus, J. Zhang and Z. Liu, “Can Graphene be used as a substrate for Raman Enhancement?,” NanoLetters, no. 10, pp. 553-561, 2009.
[30] R. Livingstone, X. Zhou, M. C. Tamargo and J. R. Lombardi, “Surface Enhanced Raman Spectroscopy of Pyridine on CdSe/ZnBeSe Quantum Dots Grown by Molecular Beam Epitaxy,”
J. Phys. Chem., no. 114, pp. 17460-17464, 2010.
[31] M. Leona, J. Stenger and E. Ferloni, “Application of surface-enhanced Raman scattering techniques to the ultrasensitive identification of natural dyes in works of art,” Journal of Raman
31
[32] P. Lee and D. Meisel, “Adsorption and Surface-Enhanced Raman of Dyes on Silver and Golds Sols,” The Journal of Physical Chemistry, vol. 86, no. 17, pp. 3391-3395, 1982.
[33] A. V. Whitney, R. P. Van Duyne and F. Casadio, “An innovative surface-enhanced Raman spectroscopy (SERS) method for the identification of six historical red lakes and dyestuffs,”
Journal of Raman Spectroscopy, no. 37, pp. 993-1002, 2006.
[34] Horiba, “Tip Enhanced Raman Scattering (TERS) or nano-Raman,” [Online]. Available: http://www.horiba.com/scientific/products/raman-spectroscopy/raman-afm-and-nano-raman/more-about-tip-enhanced-raman-ters/. [Accessed 25 November 2020].
[35] D. Kurouski, S. Zaleski, F. Casadio, R. P. Van Duyne and N. C. Shah, “Tip-Enhanced Raman Spectroscopy (TERS) for in Situ Identification of Indigo and Iron Gall Ink on Paper,” Journal of
the American Chemical Society, no. 136, pp. 8677-8684, 2014.
[36] R. M. Stöckle, Y. D. Suh, V. Deckert and R. Zenobi, “Nanoscale chemical analysis by tip-enhanced Raman spectroscopy,” Chemical Physics Letters, no. 318, pp. 131-136, 2000.
[37] S. Akyuz, T. Akyuz, G. Emre, A. Gulec and S. Basaran, “Pigment analyses of a portrait and paint box of Turkish artist Feyhaman Duran (1886-1970): The EDXRF, FT-IR and micro Raman spectroscopic studies,” Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, no. 89, pp. 74-81, 2012.
[38] F. Casadio, C. Daher and L. Bellot-Gurlet, “Raman Spectroscopy of cultural heritage Materials: Overview of Applications and New Frontiers in Instrumentation, Sampling Modalities, and Data Processing,” Topics in Current Chemistry, vol. 62, no. 374, 2016.
[39] C. Conti, C. Colombo, M. Realini and P. Matousek, “Subsurface analysis of painted sculptures and plasters using micrometre-scale spatially offset Raman spectroscopy (micro-SORS),”
Journal of Raman Spectroscopy, no. 46, pp. 476-482, 2015.
[40] M. Vagnini, F. Gabrieli, A. Daveri and D. Sali, “Handheld new technology Raman and portable FT-IR spectrometers as complementary tools for the in situ identification of organic materials in modern art,” Spectrochimica Actaa Part Al Molecular and Biomolecular Spectroscopy, no. 176, pp. 174-182, 2017.
[41] P. Vandenabeele, K. Lambert, S. Matthys, W. Schudel, A. Bergmans and L. Moens, “In situ analysis of mediaeval wall paintings a challenge for mobile Raman spectroscopy,” Anal Bioanal
Chem, no. 383, pp. 707-712, 2005.
[42] The Metropolitan Museum of Art, “The Unicorn Surrenders to a Maiden (from the Unicorn Tapestries),” The Metropolitan Museum of Art, [Online]. Available:
32
Appendix
1 Picture of the Roman wall paintings housed in the Satyr Domus in Córdoba, Spain. Source: [23].
33 3 Picture of south Netherlandish tapestry and object details. Source: [42].
