Raman Spectroscopic Characterization of Carotenoid and Phenolic Co-Occurrence in Euphorbia pulcherrima Bracts and Inflorescences

Materials and Methods

Silver nitrate >99% (Sigma-Aldrich, AgNO₃), acetone 99.5% (Synth, C₃H₆O), ethanol 96% (Vetec, C₂H₅OH), methanol 99.8% (Vetec, CH₃OH), acetic acid 99.7% (Synth, C₂H₄O₂). All aqueous solutions were prepared using deionized water (Mili-Q SQ 2Series, Merck Millipore, Merck KGaA, Germany). All reagents were used without prior purification.

Plant Material

The specimens of Euphorbia pulcherrima Willd. ex Klotzsch (Figure 2), were collected between June and December 2023 in the city of Juiz de Fora, Minas Gerais, Brazil. The experiments were performed with leaves, bracts, nectar gland, pollen, male, and female flowers.^9,51

OPEN IN VIEWER

Figure 2.

Diverse Euphorbia pulcherrima bracts and flowers. (a) Pink, red, yellow, and white poinsettias. (b) Flowers, pollen, and nectar gland of red poinsettia. (c) Developing stage of red bract.

Extraction Methods

The green leaves, red, and yellow bracts of E. pulcherrima (Figure 2) were subjected to extraction with two different extracting media. To optimize the extraction of carotenoids, acetone (Ac) was used as the extracting agent, ⁵² while for the extraction of phenolic compounds, an alcoholic mixture of ethanol:methanol:water at 1:1:1 (Alc) volume ratio with the addition of 1.0 mL of acetic acid.^53,54 Briefly, 5 g of bracts were moistened with 50.0 mL (3x) of the extracting solvent at room temperature and then ground using a mixer instrument for approximately 10 minutes. Each crude extract labeled as Ac-G, Alc-G for green leaves, Ac-R, Alc-R for red bracts and, Ac-Y, Alc-Y for yellow was filtered, and the solvents were removed through filtration under reduced pressure. ⁵⁵ The masses of the leaves and bracts, as well as their resulting dry extracts, were: Ac-G: 3.2  g/0.32  g, Alc-G: 2.6  g/0.29  g; Ac-R: 5.44  g/0.23  g, Alc-R: 5.39  g/0.36  g; Ac-Y: 4.82  g/0.14  g; Alc-Y: 5.02  g/0.27  g. White and pink bracts were not extracted due to the limited sample availability.

Green Synthesis of the Silver Nanoparticles (AgNP) with E. pulcherrima Crude Extracts

For the green synthesis of AgNP, dried extracts of green (Ac-G and Alc-G) and red (Ac-R and Alc-R) E. pulcherrima bracts were used. Solutions of the E. pulcherrima extracts were prepared by dissolving 10.0  mg of the dried extracts in 10.0 mL of 70% v/v ethanol, followed by sonication for 10 minutes. In the synthesis step, 10.0 μL of the E. pulcherrima solution was mixed with 990.0 μL of 3.0 × 10^–3 mol L^–1 AgNO₃ aqueous solution at room temperature and atmospheric pressure. The mixtures were then subjected to sonication for 10 minutes. The solutions, initially green or red depending on the extract used, rapidly changed to a yellowish color, characteristic of AgNP formation.

Instrumentation and Spectroscopic Analyses

Spectra were recorded using a Bruker MultiRAM spectrometer (Bruker Optik GbmH, Germany), neodymium-doped yttrium aluminum garnet (Nd:YAG) laser operating at 1064  nm, using a germanium detector cooled with liquid nitrogen, equipped with mirror objective and video camera (R348-VAR) and a motorized x,y,z stage (R313). Macro-mode Raman spectroscopy was performed on three green leaves, two inflorescences, ten bracts (three red, two pink, three yellow, two white) as well as from dry crude extracts of green, red, and yellow bracts. The spectra were acquired at least in duplicate in each leaf, bract, and crude extracts. The experimental conditions were spectral window from 4000 to 50 cm⁻¹, spectral resolution of 4.0 cm⁻¹, laser power ranging from 70 to 150  mW, and 512 or 1024 cumulative scans. The 2D macro-mapping sampling: spectral window from 4000 to 50 cm^–1, spectral resolution of 4.0 cm⁻¹, 150  mW and 256 scans. Resonance Raman spectra from colored bracts were also analyzed using an excitation wavelength of 532  nm with a Bruker Senterra dispersive Raman spectrometer (Bruker Optik GbmH, Germany) equipped with a charge-coupled device (CCD) detector and coupled to an Olympus BX51 microscope (Olympus, Japan) and operating with a long working distance 50× magnification objective. Data were collected in the wavenumber range of 1700 to 300 cm⁻¹, with a spectral resolution of 3.0–5.0 cm⁻¹, 50×, 50  mW, and integration times of 5  s. Raman mapping of pollen grain was obtained with 785  nm exciting radiation (Bruker SENTERRA) as follows: spectral window from 1770 to 400 cm⁻¹, 50  mW, 50×, 10 accumulations, 10  s. Dry and aqueous solution crude extracts (Ac-G, Alc-G, Alc-R, Alc-R) were recorded with 632.8  nm in the spectral window from 1780 to 390 cm^–1, resolution of 3.0–5.0 cm⁻¹, 50×, 0.2, and 20  mW, 1–15  s. The intensity ratios between carotenoids and polyphenols (I_car/I_pol) were calculated after applying the baseline corrections of spectra taken from the in-situ samples and crude extracts using the 1064  nm laser line. Baseline correction was carried out using the second-derivative (zeroes) method, with zero-crossings of the second derivative (after adjacent-averaging smoothing) serving as anchor points. Data analysis and graphing were performed using OriginLab 8.0 software. ⁵⁶

The absorption spectra of crude extracts as well as the extinction spectra of the NPs synthesized from the bract extracts were obtained by using an Ocean Optics spectrometer model USB2000 + XR1-ES (Ocean Optics, USA), operating with an Ocean Optics DH-2000-BAL radiation source in the ultraviolet–visible–near infrared (UV–Vis-NIR) region, and quartz cuvettes with an optical pathlength of 1.0  cm. All electronic spectra were collected from 200 to 1100  nm, with a spectral resolution of 1.0  nm. UV–Vis spectra were recorded for the crude extracts using water and the corresponding solvent used for extraction: acetone (Ac-G, Ac-R, Ac-Y), and ethanol:methanol:water at a 1:1:1 ratio (Alc-G, Alc-R, Alc-Y).

The surface-enhanced Raman spectroscopy (SERS) spectra of pure AgNP, i.e., the SERS spectra of extracted molecules adsorbed on the AgNP surface after their synthesis, were obtained through direct analysis of the colloidal suspension using a dispersive Raman spectrometer operating at a wavelength of 632.8  nm.

In Situ Raman Analyses of E. pulcherrima Bracts

The brightly colored bracts of E. pulcherrima are modified leaves that surround the tiny flowers, or a cluster of cyathia, which make up the plant's inflorescence. Figure 3a shows the in situ Raman spectra of E. pulcherrima bracts with diverse coloration and distinct parts of the flower (shown in Figure 2). In the spectra obtained using an excitation line at 1064  nm (Figure 3), bands characteristic of two main classes of natural products were observed: namely, phenolic acids/polyphenols, with bands around 1630–1570 cm^–1 ν(C=C), 1370–1340 cm^–1 ν(C–O), 1360–1300 cm⁻¹ δ(C–OH), and 1270–1250 cm^–1 δ(O–H)/ν(C–O) (Table I);^17–19 and carotenoids, with bands in the range of 1530–1520 cm^–1 ν(C=C), 1158 cm^–1 ν(C–C), and 1005 cm^–1 ρ(C–CH₃), as summarized in Table I.^14,57–59 Raman spectra from the high wavenumber region showed bands around 3070 cm^–1 ν(C–H)_ar and 2950–2850 cm^–1 ν(C–H) that can be ascribed to a broad range of molecules (Figure S1, Supplemental Material). Significant changes in the spectral profile the mid wavenumber region were observed according to color variation. In the green leaves, bands assigned to carotenoids present a higher intensity (e.g., 1527, 1157, and 1005 cm^–1) when compared to the main bands ascribed to phenolic acids/polyphenols (1630–1570 cm^–1) (Figure 3a). The ratio between the band intensities of the ν(C=C) mode in carotenoids (∼1527 cm^–1) and ν(C=C)_ar polyphenols (∼1600 cm^–1) was I_car/I_pol = 1.89. Conversely, in the non-green colored bracts, the opposite is observed, with the phenolic acids/polyphenol bands being more intense (Figure 3a). The intensity of the carotenoid bands gradually decreases in the following order: green leaves (I_car/I_pol = 1.89), yellow (I_car/I_pol = 0.72), red (I_car/I_pol = 0.4), and pink bracts (I_car/I_pol = 0.3). In the white bracts, the characteristic bands of carotenoids (approximately 1530, 1159, 1005 cm^–1) were not detected, despite the presence of the low-intensity bands at 1513 and 1000 cm⁻¹, which could also be attributed to flavonoids.^15,18,60 White cultivars are known to contain a few prevailing anthocyanins also detected in the bracts of the originally red E. pulcherrima species. ⁶¹ The absence of carotenoid content in the white bracts was confirmed in the spectra performed with 532  nm exciting radiation (Figure S2, Supplemental Material). The signal enhancement effect known as resonant Raman scattering,^23,52 was observed in all colored bracts, except for the white one. Such a phenomenon is explained by the resonance of the electronic transitions of carotenoids, commonly observed in the region around 550  nm, with exciting radiation used.

OPEN IN VIEWER

Figure 3.

(a) In situ Raman spectra of green, red, pink, yellow, and white bracts from E. pulcherrima. (b) Raman spectra from female flowers (style), male flowers (pedicel/stamen), nectar gland, and pollen with exciting radiation at 1064  nm.

OPEN IN VIEWER

Table I.

Tentative Raman band assignments (cm^–1) for Euphorbia pulcherrima bracts using 1064  nm laser excitation.

Green	Red	Pink	Yellow	White	Assignment	References
–	1648 w	1651w	1651 w	1654 vw	ν(C=C)	62,63
1606 w	1610 s	1612 s	1614 s	1610 s	ν(C=C)flav	16–19 , 64
1630 sh	–	–	–	–	ν(C=C)flav	16–19 , 64
1568 vw	1568 m	1570 m	1568 m	1569 m	ν(C=C)flav	16–19 , 64
1527 s	1523 w	1527 w	1525 w	–	ν(C=C)car	59,65
–	–	–	–	1513 vw	ν(C=C)flav	59,65
–	–	1496 w	1500 vw	1500 w	ν(C=C)flav	17–20 , 65
1440 w	1442 w	1440 w	1442 w	1446 w	ρ(C–H)flav/car	17–20 , 59,64,65
–	1359 w	1365 w	1363 m	1361 w	ν(C–O)flav	16–19 , 64
1327w	1336 w	1340 w	–	–	ρ(OH)flav	16–19 , 64
–	1315 w	1307 w	1307 w	1311 w	ρ(C–H)	62,63
1190 w	1182 w	1176 vw	1182 w	–	ν(C–C)	62,63
1157 m	1158 w	1157 vw	1158 w	–	ν(C–C)car	59,65
1005 w	1005 w	1004 vw	1005 w	1000 vw	ρ(C–CH3)car	59,65

s: strong, m: medium, w: weak, vw: very weak. flav: flavonoids, car: carotenoids. ν: stretching, δ: angular deformation (scissoring), ρ: angular deformation (rocking).

In Situ Raman Analyses of Poinsettia Inflorescence

The poinsettia inflorescence consists of tiny cup-shaped clusters of yellowish-green male and female flowers termed cyathia, which are surrounded by yellow nectar glands. Raman spectra obtained from distinct parts of the flowers also presented carotenoids and phenolic acids/flavonoids with different spectral patterns (Figure 3b).

Spectra from the yellow nectar gland showed major bands of the both classes with predominance of carotenoids at 1526, 1159, and 1005 cm^–1 (Table S1, Supplemental Material), while an inverse correlation was observed with those from the dark-red male (pedicel/stamen) and female (style) flowers attributed to phenolic acids/flavonoids at 1610, 1569, 1362, 1308, and 1184 cm^–1 (Figure 3b).

Yellow pollen grains showed carotenoid bands at 1526, 1159, and 1005 cm^–1 and phenolic compounds at 1640, 1603, 1587, 1445, 1213, and 1169 cm^–1 (Table S1, Supplemental Material). This spectral feature may be interpreted as a mixture of sporopollenin and carotenoids.^66–68 Sporopollenin is a heteropolymer composed of ferulic and coumaric acids as major building blocks crosslinked by hydroxyl fatty acids present in the outer walls of spores and pollen grains.^66,67,69,70 The Raman mapping of a single pollen grain performed with 785  nm laser excitation confirmed the distribution of both carotenoids and sporopollenin throughout the exine (outer layer of pollen wall) (Figure S3, Supplemental Material).

Carotenoids appear to be the primary pigments responsible for the coloration of pollen grains and nectar glands. The identification of carotenoids in pollen from E. pulcherrima is in agreement with literature, which reports that entomophilous species (insect-pollinated) such as E. pulcherrima have highest amount of protein and carotenoid compared to anemophilous species (wind-pollinated). ⁶⁹

Raman Mapping from Developing Red Bract

Raman mapping of the developing red bract was used to assess the spatial distribution of compounds within the selected area (Figure 4a). Figure 4b displays the characteristic Raman spectra acquired across 25 distinct sites, labeled A through Y, corresponding to the specific locations on the bract where the data were collected. The green part of the bract (Figures 4b–A, B, C, F, G, H, K, L, P, U) displayed a spectral pattern similar to that of the green leaf (Figure 3a); the most intense Raman bands are attributed to carotenoids (yellow stripe, Figure 4b). As the bract develops to red, it can be observed a predominance of the flavonoid bands (pink stripes, Figures 4b–I, J, M, N, O, S, T, X, Y). The integration of Raman bands at selected wavenumbers is shown in Figure 4c and Figure 4d. The multi-trace map (Figure 4c) shows that both classes of compounds occur along the mapped area, with increasing red color gradient assigned to phenolic acids/flavonoids from the margin to the midrib of the bract. The integration of flavonoid (1645–1585/1580–1547 cm^–1) and carotenoid (1540–1500 cm^–1) marker bands was represented in the heat maps by the magnitude of the Raman intensity. The heat maps show hues of orange color representing the highest Raman intensity, while hues of blue represent the lowest. It can be observed an inverse correlation between the two classes of metabolites as the red color develops (Figure 4d).

OPEN IN VIEWER

Figure 4.

Raman mapping of the partially colored stage of red bract development. (a) Optical image displaying the mapped area. (b) Raman spectra displaying the wavenumber region between 1650–1400 cm^–1 of each selected spot (A–Y). (c) Multiple integration of marker bands in the range of 1630–1596 cm^–1 (red: flavonoids), 1540–1500 cm^–1 (orange: carotenoids) and a combination of both (light to dark brown). (d) The heat map color gradient shows the integration of flavonoid and carotenoid marker bands, the color scale indicates the magnitude of the Raman intensity, where yellow represent the highest Raman intensity while blue is the lowest.

The Raman mapping revealed how ubiquitous carotenoids and flavonoids are in the leaves or bracts, regardless of the developmental stage. Even in the fully colored bract carotenoids are synthesized. Carotenoids were selectively identified in the analysis of crude extracts performed with different solvents discussed below.

Raman Spectral Analyses of Crude Extracts

Two distinct extraction methods were used on the green leaves, yellow, and red bracts of E. pulcherrima to evaluate the carotenoid occurrence in green leaves and non-green bracts. The acetone crude extract of green leaves (Ac-G) yielded the best results, with characteristic Raman bands at 1529, 1159, and 1007 cm⁻¹, when compared to hydroalcoholic mixture (Alc-G) (Figure 5a, Figure 5b, and Table S2). Besides carotenoids, the Ac-G presented very weak bands at 1710 (C=O)_ester, 1640 (C=C)_allylic, 1618 (C=C)_ring, and 1570 cm^–1(C=C) assigned to phenolic compounds. On the other hand, the Alc-G showed a significant content of phenolic compounds indicated by broad bands observed at 1691 (C=O)_ester, 1630 (C=C)_allylic, 1606 (C =C)_ring, 1585 (C=C), 1570 (C=C), 1500 cm^–1 (C=C). As can be seen in each set of spectra (Figure 5a and Figure 5b), phenolic compounds presented distinct maximum Raman intensities for the stretching vibration of the aromatic rings at approximately 1620 (Ac) and 1610 cm^–1 (Alc). The occurrence of phenolic compounds was also revealed by the UV–Vis spectral analyses showing absorption bands at 230, 268, 320  nm (Ac-G), 227, 266, 346  nm (Ac-R), 225, 270, 345  nm (Alc-G), and 227, 266, 346, 510 (Alc-R). Besides, the crude extracts also showed bands of chlorophylls at 667  nm (Figure S4, Supplemental Material).

OPEN IN VIEWER

Figure 5.

Raman spectra of crude extracts from E. pulcherrima bracts obtained with (a) acetone and (b) hydroalcoholic mixture media obtained using 1064  nm wavelength excitation.

An expressive content of carotenoids was also observed in yellow bracts (Ac-Y: I_car/I_pol = 0.9), and minor amounts in red ones (Ac-R: I_car/I_pol = 0.37) (Fig. 6a). Both in situ and crude extract analyses show variations in Raman intensities (1608–1570 cm^–1), suggesting that yellow and red bracts possess distinct phenolic profiles (Figure 3a and Figure 5a). Unlike the green leaf hydroalcoholic extracts (Alc-G), Alc-Y and Alc-R showed no carotenoid signals, though they maintained a similar 10 cm⁻¹ blueshift in the aromatic region. These shifts in band position suggest that each solvent possesses a different extraction efficiency for specific classes of compounds. Decomposition of the broad bands between 1670–1485 cm^–1 in both types of crude extracts (Figure S5) confirmed the structural variability and selectivity of the solvent in recovering carotenoids and phenolic compounds. The acetone crude extracts were efficient for carotenoids (1527 cm^–1) as well as polyphenols (flavonoids/anthocyanidins) according to marker bands approximately 1650–1660, 1620–1600, 1580–1570 cm^–1 (Table S2, Supplemental Material), ^{60 71–73} while hydroalcoholic mixture favored phenolic acids/polyphenols (1700–1790, 1630–1640, 1610–1600, and 1570–1560 cm^–1).^{16,60,73–75}

OPEN IN VIEWER

Figure 6.

UV–Vis spectra of acetone (Ac) and hydroalcoholic (Alc) crude extracts from E. pulcherrima (Ep). Panels (a) and (b) represent green leaf extracts (Ac-G, Alc-G), while (c) and (d) represent red bract extracts (Ac-R, Alc-R). Solid lines indicate spectra from Ep/AgNP in aqueous solution; dotted lines represent the corresponding crude extracts in aqueous solution.

Gu et al. ¹¹ reported that green bracts extracted with methanol yielded phenolic acids and higher diversity of flavonoid compounds than in red turning bracts, corroborated by spectral results obtained in this study with hydroalcoholic mixture (Alc-G) (Figure 5b). In addition, a study performed with green, reddish and red bracts showed that cyanidin accumulation in red bracts was two-fold higher than in green leaves, while in reddish leaves was 1.4-fold higher than in green ones. ⁷⁶ The color transition is due to differential vacuolar anthocyanin accumulation, which is influenced by environmental factors such as temperature and photoperiod. ⁷⁶

According to the literature the red pigmentation of bracts is mainly due to cyanidin followed by pelargonidin, ⁶¹ the pink and white bracts being composed of the same anthocyanins but, in a much smaller amount (approximately three times less than the red one), ⁶¹ the orange-red bract is based on the prevalent presence of pelargonidin derivatives, ⁶ however, the characterization of the yellow bract have been overlooked.

The structural diversity of phenolic derivatives in poinsettia bracts, as well as in several other plants, may contribute to the co-pigmentation effect. ⁸ The co-occurrence of flavonoids, anthocyanins, and carotenoids in the yellow bract could be attributed to the concomitant accumulation during the pigmentation process and/or an intermolecular co-pigmentation. Several examples of co-pigmentation in flowers have been described in the literature,^77–81 mainly intermolecular association between anthocyanin and flavonoids; ⁸² they stabilize and modulate (in particular blue, violet, and red) colors in flowers and berries. Nonetheless, yellow floral pigments are commonly derived from aurones, chalcones, and flavonols. ⁸¹ To a lesser extent, carotenoids serve as the primary components, as described for the yellow petals of Rosa spp., ⁸³ Nicotiana rustica L., ⁷⁰ and Camellia nitidissima, ⁸⁴ where flavonoids and anthocyanins are present only in smaller amounts. Although E. pulcherrima’s bracts are histologically and morphologically distinct from petals, they share several metabolites. In this sense carotenoid content in the yellow bract of E. pulcherrima could be the major pigment responsible for the coloration.

Green Synthesis of AgNP Using E. pulcherrima (Ep) Crude Extracts

The crude extracts from the bracts of E. pulcherrima were employed in the synthesis of silver nanoparticles (AgNP) to evaluate the potential of Ep/AgNP in green synthesis. The characterization of the synthesized Ep/AgNP was performed by UV–Vis spectroscopy (Figure 6) and confirmed by the observation of surface-enhanced Raman spectroscopy (SERS) effect (Figure 7). The UV–Vis spectra of the complex Ep/AgNP showed a characteristic localized surface plasmon resonance (LSPR) band of silver nanostructures at approximately 400  nm (Figure 6). Bands of crude extracts and localized surface plasmon of AgNP did not overlap, confirming the reaction among polyphenols and Ag⁺ salt (Figure 6). The Ac-R was the most efficient in forming AgNP, followed by Alc-R, as evidenced by the high intensity of the LSPR band (Figure 6c). In contrast, Ac-G did not favor nanoparticle formation. According to the results, the AgNP formed are suggested to be small particles, since the LSPR band maxima appear in the range of 410–430  nm, similar to citrate-reduced syntheses (Figure S6, Supplemental Material). ⁸⁵ A large particle size distribution could be inferred for these effective syntheses, as indicated by the broad LSPR band profile and the extension of such bands to wavelengths above 500  nm. ⁸⁶

OPEN IN VIEWER

Figure 7.

SERS spectra of chemical species present in green synthesized AgNP using E. pulcherrima crude extract of green leaves (Ac-G, Alc-G) and red bract (Ac-R, Alc-R) using 632.8  nm wavelength excitation.

The plasmonic properties of Ep/AgNP were confirmed by the signal enhancement observed in SERS spectra. Figure 7 shows some vibrational bands observed at 1599, 1556 cm^–1 attributed to ν(C=C)_ar, 1483 δ(C–H), 1332 cm^–1 δ(C–O–H), δ(O–H), and 1018 δ(C–H),^19,73 which can be assigned to polyphenols species. Raman spectra from dry and aqueous solution crude extracts acquired at 632.8  nm showed only fluorescence backgrounds (Figure S7, Supplemental Material). This confirms that the signal enhancement observed in AgNP-containing samples is a result of the SERS effect provided by the nanostructured silver surface.

The Ac-R is rich in carotenoids and flavonoids/anthocyanidins, while the Alc-R is rich in phenolic acids/flavonoids/anthocyanidins. The higher efficiency of AgNP formation from the extracts of red bracts, especially in an acetone medium (Ac-R, Figure 6) may be due to the flavonoid and anthocyanin contents^39,87,88 consequently, their greater reducing efficiency compared to the extract of green leaves (Ac-G, Figure 6). Recent work demonstrated that aqueous crude extracts from E. pulcherrima were effective in reducing silver and silver chloride nanoparticles (EP-Ag/AgCl-NPs). ³⁴ According to Sajjadi et al., polyphenols identified in aqueous crude extracts from Euphorbia peplus were the main compounds responsible for the reduction of Ag^– ions; ³⁸ other authors showed that LSPR bands placed near 420  nm could be generated by tea plant (Camellia sinensis), standard polyphenols-mediated Ag⁺ reduction,^89,90 or carotene-stabilized AgNP. ⁹¹ Silver nanoparticles synthesized with β-carotene have also been reported, ⁹² so the combination of carotenoids and flavonoids may have a synergistic effect. Based on the aforementioned literature, it can be concluded that flavonoids and anthocyanins extracted via diverse methods from E. pulcherrima serve as effective natural reducing agents for the synthesis of silver nanoparticles. This green synthesis approach is not only environmentally friendly due to reduced waste generation but also economically advantageous, lowering manufacturing cost. Most importantly, the resulting nanoparticles offer enhanced biocompatibility for biomedical and pharmaceutical applications.^93,94