Characterization and discrimination of some gem silicate minerals adopting LIBS, FTIR, and Raman spectroscopic techniques

Gems range in quality from semi-precious to precious, and they have been extracted from geological sites in antiquity by different cultures and civilizations, including the period of Pharaohs. From a scientific perspective, gemstones/gems are minerals with a value that are either natural or synthetic. Sometimes, it is difficult to distinguish natural gemstones from the synthetic ones, but using advanced analytical techniques, it is possible to identify and discriminate them precisely. Natural gemstones, as natural resources, occur in different geological environments, and their spectrum of quality is greatly variable, ranging from jewelry to showcase, depending on some properties, such as color, transparency, and hardness (durability). Contributions of spectroscopic techniques, namely laser induced breakdown spectroscopy (LIBS), Fourier transform infrared (FTIR), and Raman, in the study of gems furnish a solid ground for the identification, classification, and evaluation of minerals, particularly in the past five decades. According to Fabre,¹ the advances in the LIBS technique and its applications in geosciences are not just helpful in the identification, imaging, or quantification of wide-spectrum elements, but they are also very helpful in the characterization of homogeneous/heterogeneous mineral phases. This includes detection and measurement of light elements, C, O, and even organic matter.¹ In addition, LIBS applications in different varieties of samples and mineralogy, in particular, offer rapid chemical analysis in real time with or without little preparation.^2–4 

Moreover, FTIR spectroscopy is very useful in the field of mineralogy because it helps both qualitatively and quantitatively.⁵ This means mineral species can be identified, distinguished, and discriminated. Mixtures/admixtures of minerals can be identified by extraction from their FTIR spectrum, and this is attributed to the fact that their fundamental molecular vibration modes in the mid-IR (4000–400 cm⁻¹) and the absorbance bands (e.g., Al–O, Al–OH, and Si–O) are proportional to the pure mineral spectrum fulfilling the aspects of the so-called Beer’s law.^6,7 Raman spectroscopy is extremely helpful in the field of mineralogy because one can identify and characterize a certain mineral phase/species based on the resultant spectrum, particularly in the case of “polymorphic” minerals that have the same chemical formula but crystallize in different coordination schemes, i.e., the crystallographic systems.⁸ Therefore, Raman spectroscopy is a powerful experimental technique that permits the identification of mineral species without time-consuming and expensive preparation of the sample. In addition, it is a precise, simple, and rapid technique and necessitates no special sample preparation. For example, sheet-structured silicate minerals, such as serpentine, and clay minerals can be distinguished by Raman spectroscopy on the basis of their micro-structural characteristics.^7,8

In the present study, we present the first detailed micro-analysis and characterization of some gems from Egypt, in addition to a single sample from Saudi Arabia for comparison. Our study aims to contribute to the characterization of gemstones (peridot and emerald) and semi-gemstones (beryl, amazonite, and amethyst) in order to identify the source and to show the differences in the gemological, chemical, and spectral signatures. This would be helpful for gem characterization, tracing ancient routes of trades, and possible further exploration.

Geographically, the four studied silicate gems are sporadically scattered in the Eastern Desert of Egypt and the western part of the Arabian Peninsula in Saudi Arabia as shown in Fig. 1. Figure 2 displays the colored natural silicate gemstones. This area is occupied by the so-called Arabian-Nubian Shield that formed in the Neoproterozoic Era, some ∼1000 to 543 × 10⁶ years ago. Mineralogical, geochemical, and archaeological studies in the past focused on emerald only.^9–12 These studies concerned with the genesis of minerals with a very little attention to their gemological characteristics based on the characterization by the modern spectroscopic techniques, which is the main goal of the present study.¹¹ 

The experimental setup for LIBS analysis uses a Q switched Nd:YAG laser at its fundamental wavelength of 1064 nm (Brio, Quantel, France). A plano–convex lens of 5 cm focal length was used to focus the laser beam onto the target gem samples. The energy of the laser beam was optimized to be 45 mJ per pulse with a pulse duration of 5 ns (FWHM) at a 20 Hz repetition rate and was monitored by using a joulemeter (Scientech, Model AC5001, Boulder, CO, USA). The light from the produced plasma was collected via a 2 m fused silica optical fiber with a core diameter of 600 μm and fed to the entrance slit of an Echelle spectrometer (Mechelle 7500, multichannel, Sweden). The spectrometer was coupled to an ICCD camera (DiCam-Pro, PCO Computer Optics, Germany). The time resolved emission has been optimized at a delay time of 1500 ns and a gate width of 2500 ns. The collected LIBS spectrum was the average of three laser pulses that were shot at five different locations on the target surface. Finally, LIBS⁺⁺ software was used in the analytical process for the collected LIBS spectra.

Raman analyses were conducted using a German-made confocal Raman microscope (Model WITec alpha300R). The micro-Raman system worked under the conditions of 473/532/633 nm laser excitation, z-focus, and a software-controlled x–y sample stage for line scanning and mapping. The obtained spectra were possible at an excitation wavelength as high as 785 nm.

The FTIR analysis was conducted using a Japanese-made Model 4100 Jasco spectrometer working in the vibrational range of 400–4000 cm⁻¹ wavenumber with the aid of potassium bromide (KBr) as a reference , which has a transmittance of 100%. This machine is equipped with a Globar SiC source, a Ge-coated KBr beam splitter, and a liquid N₂-cooled HgCdTe detector. The FTIR spectra were collected under nitrogen purging at 2 cm⁻¹ pre-selected resolution at ∼64 scans.

In this study, we present for the first time LIBS micro-analyses of natural mineral gemstones from Egypt. The potentiality of LIBS is to investigate and discriminate between samples of similar elements and provide authentication of samples’ provenance. The results obtained from the analyses provide useful information about dominant elemental composition and the trace elements of the different varieties of gemstones. Figure 3 illustrates the characterizing elements for each group of gems as captured by LIBS. The ranges of wavelengths for each group have been selected according to the abundant elements that have a higher intensity with respect to each other. Amethyst (SiO₂) is a transparent, coarse-grained variety of the silica mineral quartz that is valued as a semi-precious gem for its violet color. Its physical properties are those of quartz, and the characterizing element is silicon, which has a higher intensity in comparison with other types of gems as it is clear in Fig. 3(a). However, the trace element responsible for the purple color of amethyst is iron. The presence of Fe as a chromophore has been proved by the detection of elemental lines of Fe in the range from 260 to 277 nm but not shown here.

Peridot [(Mg,Fe)₂SiO₄, a nesosilicate] is one of the unique gemstones that come in a single green color. However, depending on the level of iron content, the depth of green may vary from light yellowish green to olive and dark brownish green. Figure 3(b) depicts the presence of Fe with a higher content in both Egypt and KSA when compared to the other gems.

In Fig. 3(a), LIBS spectrum of amazonite [microcline (KAlSi3O8)] indicates the presence of Si as illustrated but with a lower content than that of both peridot and beryl. Moreover, the existence of Ca with a lower intensity than in peridot [Fig. 3(c)] as a divalent cation in the octahedral site indicates the couples’ cationic substitutions of alkalis in the amazonite tectosilicate. One can notice that no characterizing element has a higher intensity for amazonite among other gems.

The beryl group (Be₃Al₂Si₆O₁₈) from Wadi Ghazala and Um Sleimat, indeed, has a higher content of Ti. The common or pure color of beryl is white; thus, the LIBS analysis reveals the higher content of titanium as shown in Fig. 3(d). Moreover, the presence of Ti may suggest that it is an important chromophore in the beryl group that results in the green color, side by side with other transitional elements.

Another significant study describing the gemstones of the same name as beryl and peridot is shown in Figs. 4(a) and 4(b). Figure 4(a) reveals a comparison between the olivine peridot from Red Sea Egypt and that from Harrat Kishb (KSA); they have almost the same amount of Fe lines. The three emerald (beryl) gems are studied at different ranges of spectrum, as shown in Figs. 4(b) and 4(c). The emerald from Wadi Ghazala and Sinai Egypt shows higher intensity lines for silicon [Fig. 4(b)] compared to the other two emeralds from Um Sleimat and Um Addebbaa. Moreover, the Ti lines have a higher intensity for the emeralds from both Wadi Ghazala and Um Sleimat than that from Addebbaa, as shown in Fig. 4(c). The LIBS spectra of the dark bluish green emeralds (Um Addebbaa) reveal that the Cu lines at 313 nm imply the important roles of Cu²⁺ in the bluish coloration of emeralds, not shown here. The appearance of Ti spectral lines implies that the presence of Ti⁴⁺ in fluid inclusions in emeralds results in the appearance of a C peak at 274.5 nm.

Finally, these results conclude that fingerprint elemental lines for each group of gemstones may exist; however, there are more sharing elements between the targeted samples as expected from gemstones collected from different environments. These gemstones sometimes have a similar composition but may differ in the amount of elements (concentration of elements).

The FTIR results are demonstrated in Figs. 5 and 6. Figure 5 shows the FTIR spectra of emerald and beryl crystals, whereas Fig. 6 displays those of peridot, amazonite, and amethyst. The FTIR spectra of light green beryl crystals are distinguishable from those of dark green and bluish green emeralds. The obtained spectra enabled us to distinguish the gem quality of emeralds (dark green) that were collected from Um Sleimat and Um Addebbaa [Figs. 5(a) and 5(b)] from the low-quality pale green beryl crystals that were collected from Wadi Sikait and Wadi Ghazala [Figs. 5(c) and 5(d)]. Similarly, the FTIR spectra of amazonite and amethyst [Figs. 6(a) and 6(b)] were distinctive from those of peridot from the Saudi Arabian and Egyptian localities [Figs. 6(c) and 6(d)]. It is evident that some chromophore cations in the functional group (the octahedral sites), such as Cr³⁺, Fe^2+,3+, and Cu²⁺, play a significant role in the coloration of the investigated gems. Chromophores in natural minerals of definite inorganic chemical formulas are molecules that absorb particular wavelengths of visible light and transmit/reflect others causing coloration of the material. The concept of chromophore as a coloring agent is accepted by researchers from different scientific disciplines.¹³ Chromophore ions have a key role in the coloration of emerald from Ancient Roman mines in the Eastern Desert of Egypt.^11,14 Different research studies gave evidence of the coloration between Cu²⁺ and some transitional elements (Co, Cr, and V) using the electron paramagnetic (EPR) spectra. There is still a need for ultraviolet–visible–near-infrared (UV–Vis–IR) investigation to enhance such evidence as has been performed for some international gems, particularly beryls/emeralds.¹⁵ The visible chromophores are important even if their concentration in the beryl/emerald is low.

Figure 5 suggests that the discrimination between low-gem quality beryl (Wadi Sikait and Wadi Ghazala) and proper emeralds (Um Sleimat and Um Addebbaa) is possible both in the fingerprint region (400–1020 cm⁻¹) and in the functional group region (1020–4000 cm⁻¹). In the latter range, a band assigned for CO₂ around 2350–2470 cm⁻¹ is very characteristic for the dark green emerald from Um Sleimat and Um Addebbaa [Figs. 5(a) and 5(b)]. This is sometimes overlapped by stretching deuterated H₂O. The fingerprint region shows the common Si–O stretching (symmetrical and asymmetrical) and bending vibrations for the four crystals (Fig. 5).^15–17 At ∼1200 cm⁻¹, the effect of chromophore elements might appear, most probably due to Cu²⁺ based on the EPR spectra and trace element composition by the conventional XRF method.^14,18 This Cu²⁺ band can be seen in the FTIR spectrum of beryl from the Wadi Ghazala in which the trace of copper is responsible for the bluish tint of this light green beryl. In addition, the absence of the ∼2290 cm⁻¹ band of structural water in all of the studied beryls/emeralds is logical because they are natural and such a band can only be found in synthetic hydrothermal emeralds.^19,20 Such infrared data of the investigated Egyptian beryls/emeralds are consistent with the classification of the important gemmy mineral.^21,22 The infrared spectra of low-gem quality [Figs. 5(a) and 5(b)] are identical to the FTIR spectra of low-grade emeralds of some Indian beryl deposits at the Odisha Province.²³ Regardless of the magnitude of the transmitted band, the spectra of amazonite and amethyst are similar because they are structurally classified as tectosilicates [Figs. 6(a) and 6(b)]. The FTIR spectrum of amazonite [Fig. 6(a)] is typical of a feldspar structure in which the characteristic Si–O–Si and Al–O–Si stretching and bending are distinct in the functional group region at 640–650 and 500–510 cm⁻¹, respectively. No indication of a chromophore can be seen here. The Wadi Sikait amazonite from the ancient Roman sites is characterized by a broad and intense peak at ∼900 to 1200 cm⁻¹ that results from Si–O bonds and the vibrational modes of Si–O–Si and Si–O–Al. This makes amazonite distinguishable from those found in some cultural heritage sites in Mexico and Jordan. On the other hand, the FTIR spectrum of amethyst shows a more pronounced (metal, namely Fe^2+,3+) M–O bending at 1020–1050 cm⁻¹ even though it overlaps those of the S–O bending at 1020 cm⁻¹ and ∼700 cm⁻¹ reflections.²⁴ In fact, the iron species in amethyst is rare (Fe⁴⁺), resulting from O²⁻ charge transfer because of natural irradiation, and this gives the diagnostic violet-purple color of this SiO₂ variety.²⁵ In the fingerprint region of our amethyst, the O–H stretching vibrations at 3400–3750 cm⁻¹ are absent, indicating neither OH⁻ nor H₂O in the structure.⁵ Similar to peridot, the spectrum of crystals from the Kishb locality in Saudi Arabia shows a more or less flat fingerprint region from 1000 to 4000 cm⁻¹ [Fig. 6(c)]. However, it still indicates a typical olivine nesosilicate structure similar to that from the Egyptian locality of the Zabargad Isle [Fig. 6(d)]. It seems that the clarity (transparency) and strength of the coordination of metal (dominated by Mg over Fe²⁺) with oxygen play a role in the magnitude of bending reflections. Therefore, the ∼890 to 1005 cm⁻¹ M–O stretching or bending is more pronounced in the spectrum of peridot (brownish olive green) from the Zabargad Isle than in that of peridot from the Kishb locality (transparent olive green). Probably, Ni as a transitional element has a role here although it is still puzzling as the Kishb peridot contains up to 0.35 wt. % NiO as measured by Surour.²⁶ The spectra of crystals from the two localities [Figs. 6(c) and 6(d)] suggest the presence of C–H–O stretching bands at ∼2350 to 2360 cm⁻¹ due to the presence of CH₄ and CO₂ in the fluid inclusions hosted in olivine, particularly in peridot xenocrysts brought by the volcanic lava from a mantle source.^27–29 

The Raman results (in the range of 200–3000 rel. 1/cm) look very promising because the resultant spectra of the four studied silicate minerals are distinctly different as shown in Figs. 7(a)–7(d). Obviously, the Raman spectra of peridot, amazonite, and beryl are not identical. It is interesting to note that the Raman spectrum of the low-quality beryl from the Wadi Ghazala locality in Sinai is somehow similar that of amethyst [Figs. 7(b) and 7(d)]. Figure 7(a) illustrates the Raman spectra of transparent olive-green gemmy peridot from the Kishb locality. We could observe a very distinctive characteristic Raman shift at 820–860 cm⁻¹. Using the Raman peak position of Kuebler et al.³⁰ and the equation of Breitenfield et al.³¹ that uses this shift to predict the chemistry of peridot in terms of Mg/Fe of the forsterite (Fo) component, the obtained range of Fo is ∼90–92, which is consistent with the value from the electron microscope analysis (EMPA) of the same crystal.²⁶ The Raman spectra of the light green beryl from the Wadi Ghazala locality [Fig. 7(b)] showed some significant Raman shifts at ∼270, 320–325, and 565–570 cm⁻¹. The ∼270 cm⁻¹ is attributed to tiny hematite inclusions, whereas 565–570 cm⁻¹ possibly resulted from the deformation of Si₄O₁₁ in the ring structure.³² Previous studies^23,33 demonstrated that the 565–570 cm⁻¹ frequency might indicate the presence of Cr³⁺ chromophore. On the other hand, the peak at 320–325 cm⁻¹ indicates the ring vibration mode as the classical fingerprint of any beryl variety.

The Raman shifts of the investigated amazonite [Fig. 7(c)] are unique and characterized by a broad and intense peak at ∼900 to 1200 cm⁻¹ that results from Si–O bonds and the vibrational modes of Si–O–Si and Si–O–Al. Thus, we can say that our Egyptian amazonite from the ancient Roman sites in the Wadi Sikait area is distinguishable from those from other world localities, e.g., amazonite beads from cultural heritage sites in Mexico and Jordan.^34,35 In the latter cases, the given vibrational modes occur at 507–5012 and ∼1100 cm⁻¹ frequencies as a function of the tetrahedral (SiO₄)⁴⁻ stretching mode, respectively.³⁶ The Raman shifts of the Aswan amethyst [Fig. 7(d)] are characteristic as well, particularly the low frequency shifts at ∼125 and ∼205 cm⁻¹ that indicate the displacive transition due to the motion of Si and O around the triad crystallographic axis of the trigonal system in which the mineral crystallizes.³⁷ The most distinctive peak appears at ∼460 cm⁻¹, and it indicates either the Si–O–Si linkage or the v₂ symmetrical bending of the (SiO₄)⁴⁻ tetrahedra.³⁸ 

The results in the present work have proved the usefulness of elemental and molecular spectroscopic analysis techniques, such as LIBS, Raman, and FTIR, to study four groups of silicate gems of precious and natural gemstones that are collected from the “Arabian-Nubian Shield.” Two olivine peridot, two beryl (emerald) gemstones, amazonite, and amethyst have been selected for investigation and evaluation this study. Hydrous Cu carbonate and phosphate (e.g., malachite and turquoise, respectively) have been widely used in antiquity, such as in the time of Pharaohs in Egypt some 3000–7000 years ago, with lots of disputes concerning the source of gems and their validity for use in ancient jewelry. In order to trace all chromophores (Ti and transitional elements, such as Fe, Cr, V, and Co) in beryls/emeralds, LIBS analysis in a high range (200–900 nm) is considered. LIBS analysis, for the first time, has been adopted to detect the dominant elemental composition and the trace elements for the studied gemstones. As amethyst (SiO₂), for example, is valued as a semiprecious gem for its violet color, its physical properties are those of quartz and the trace element responsible for the purple color in amethyst is iron. Vibrational techniques, such as FTIR and Raman, have been demonstrated for the identification and exploration of the fingerprints and functional groups, in addition to their ability to discriminate between the natural and synthetic ones. The effect of Cu²⁺ as a chromophore causing bluish tint can be identified by the ∼1200 cm⁻¹ band. Furthermore, the absence of the ∼2290 cm⁻¹ band of structural water in all of the studied beryls/emeralds is because such a band can only be found in synthetic hydrothermal emeralds. This could be used as a criterion for distinguishing between natural and synthetic gemstones. Even minerals that have the same chemical formula but different crystallographic orientations can be characterized by Raman spectroscopy. The low frequency Raman shift of amethyst at ∼125 and ∼205 cm⁻¹ results from a displacive transition due to the motion of Si and O around the threefold axis in the trigonal symmetry. The position of the peaks in the Raman spectra of gems from the Kishb locality and the calculation of Mg/Fe (Fo content) based on the shift at 820–860 cm⁻¹ yield ∼90–92 Fo, which is similar to that obtained using EMPA in the literature.

Finally, this study would be helpful to build up a database for characterization, authentication, tracing of ancient routes of trades, and possible further exploration of gemstones.

Future studies should explore the gemstones available from the sites of cultural heritage and that are of considerable gemological value. This could be performed by adopting other precise analytical techniques, such as energy-dispersive x-ray analysis (EDAX) and XRD, to provide complementary information to differentiate natural and synthetic gemstones.

The authors have no conflicts to disclose.

Amal Abdelfattah. Khedr: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Resources (equal); Software (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Adel A. Surour: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Resources (equal); Software (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Ahmed El-Hussein: Conceptualization (equal); Investigation (equal); Methodology (equal); Validation (equal); Writing – review & editing (equal). Mahmoud Abdelhamid: Data curation (equal); Methodology (equal); Software (equal); Validation (equal); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.