2.1.

Selection of Volunteers

This study was approved by the Ethics and Research Committee of Universidade Anhembi Morumbi (Process No. CAEE 69573917.9.0000.5492) in conformity with the Resolution 466/2012 of the Brazilian National Health Council from the Brazilian Ministry of Health.³⁰ Volunteers who agreed to participate signed an informed consent form.

A group of volunteers (nine women and five men) was enrolled for the repellent testing, and another group (seven women and seven men) was enrolled for the sunscreen testing. The skin phototype ranged from I to IV (phototype classification according to Fitzpatrick scale³¹). Inclusion criteria were healthy people, aged between 18 and 60 years, without any skin wound and absence of any known allergic reaction to the ingredients of the products. Exclusion criteria were allergic reaction to products during the experiment period and painful sensibility to the Raman laser beam.

2.2.

Samples

The study used two samples of insect repellents of a market leading brand presented as spray and cream (identified as SPR and CR) and four samples of sunscreens, all presented as cream, of two market leading brands (identified as CEN and SD) with SPF of 15 and 30. The repellent products used in the study had a composition according to Table 1 and the sunscreen products had a composition according to Table 2.

Table 1

Composition of the insect repellent products used in the study per the information provided on the product labels. An asterisk (*) highlights the active ingredients.

Table 2

Composition of the sunscreen products used in the study per the information provided on the product labels. An asterisk (*) highlights the active ingredients.

2.3.

Application of the Products on the Skin

Initially, the sites where the products were applied were sanitized with an alcohol-soaked cloth (ethanol 95%) for removal of contaminants. For the insect repellents, volunteers received topical administration of spray and cream products on two circular sites of the anteromedial region of both forearms, each circular site with 25 mm in diameter, identified as SPR (spray) and CR (cream) [Fig. 1(a)]. A third circular site, located in the anteromedial region of left forearm and identified as CTR (control), was used as control and did not receive any product [Fig. 1(a)]. Spray insect repellent was applied by a single spray, 10 cm far from the volunteer’s skin, at the SPR site. Cream repellent was applied throughout the marked area with the help of a microspatula, using a standard amount of cream (about 9.4 mg), on the site identified as CR located at the left forearm. The quantities of each product used in this study are compatible with recommendations from the repellent manufacturers as prescribed in the respective labels.

Fig. 1

(a) and (b) Forearm markings for repellent and sunscreen application in circular sites, respectively. (c) Schematic diagram of the dispersive Raman spectrometer used for spectral data acquisition. The Raman probe possesses a conical aluminum tip to promote the focal distance between the probe tip and the skin surface.

Sunscreen products were applied topically at the anteromedial region of each volunteer’s right forearm. The application was performed on four circular sites with 25 mm diameter, identified with numbers from 1 to 4 [Fig. 1(b)]. In each region, a standard amount of sunscreen (9.4 mg) of the brands CEN (SPF 15 and 30, sites 1 and 2) and SD (SPF 15 and 30, sites 3 and 4) was applied, with the help of a microspatula throughout the marked area. Site 5 was kept as a control region, identified as CTR [Fig. 1(b)]. The quantity of the sunscreen applied on the marked area was according to the recommendation by the ANVISA for SPF determination tests, corresponding to $2.0\mathrm{mg}/{\mathrm{cm}}^2$ (as preconized by ISO⁹ and the European Cosmetic and Perfumery Association—COLIPA³²). Volunteers waited for 30 min before Raman spectra acquisition.

2.4.

Collection of the Raman Spectra

Initially, Raman spectra of the commercial products were acquired, placed on an aluminum sample holder, identified as REF_SPR and REF_CR (repellents spray and cream), and REF_CEN15, REF_CEN30, REF_SD15, and REF_SD30 (sunscreens of brands CEN and SD, with SPF 15 and 30, respectively). Then, spectra were acquired from each site shown in Fig. 1 in triplicate. Table 3 shows the number of spectra obtained in each site and the number of acquisitions per group. Some spectra were withdrawn due to very strong fluorescence background, particularly in the groups SD15 and SD30, which caused saturation of the detector in the Raman spectrometer and degradation of the Raman signal due to fluorescence shot noise.

Table 3

Application sites, number of spectral acquisitions, and name of groups used in the study.

Raman spectra of each region were acquired using a dispersive Raman spectrometer (model Dimension P-1, Lambda Solution Inc., Massachusetts), with a fiber optic Raman probe (model Vector Probe, Lambda Solutions Inc., Massachusetts) schematically presented in Fig. 1(c). The spectrometer uses a diode laser (830 nm, 350-mW max power) coupled to a Raman probe to irradiate the repellent/sunscreen sample or the skin. The probe captures the light scattered by the sample, which is dispersed by the spectrograph and directed to a high sensitivity, deep-cooled CCD camera connected to a notebook computer for acquisition and storage. Spectrometer wavenumber calibration was checked through the Raman bands of naphthalene; the correction of the spectrometer’s spectral response (intensity calibration) was applied at the time of spectrum storage. The Raman spectrometer has a resolution of about $4{\mathrm{cm}}^{-1}$ in the spectral range of 400 to $1800{\mathrm{cm}}^{-1}$.

A conical aluminum tip was used at the probe’s distal tip, to standardize the focal distance between the probe tip and skin, allowing for repeatability of excitation and collection geometry during spectra acquisition. Three spectra in each site area were acquired, each spectrum with acquisition time of 30 s and laser power on probe’s tip set to 250 mW.

2.5.

Spectra Processing and Statistical Analysis

Raman spectra from products placed on an aluminum sample holder and topically to skin were preprocessed the same way as detailed by Silveira et al.³³ First, high-intensity outliers from cosmic rays were manually removed; following, removal of the Raman background (most fluorescence emission) was done by fitting and subtracting a seventh-order polynomial to the baseline of each spectrum according to the procedure described by Lieber and Mahadevan-Jansen;³⁴ finally, in order to minimize errors derived from eventual subtle differences between measurement conditions (laser power, tissue absorption, artifact from probe movements, etc.) and to improve the comparison of the spectra from different sites, each spectrum was normalized by the area under the curve for discrete signals (1-norm).³⁵ After preprocessing, the mean spectrum of each group was calculated.

Using the recent literature related to Raman spectra of chemical components, the peaks corresponding to spectra of the active ingredients listed on repellent and sunscreen product labels (Tables 1 and 2, respectively) were identified within the spectra of the undiluted products. Then, the peaks of these ingredients were identified on the spectra of the skin that received the products compared to the spectra of untreated skin (control). These peaks were evaluated to verify if there was a statistically significant difference between the peak intensities of the sites that received the products compared to controls. The statistical analysis consisted of the Kolmogorov–Smirnov test to verify data normality and one-way analysis of variance (ANOVA) test followed by Tukey–Kramer post hoc test. The null hypothesis for the ANOVA was the equality in the mean of the peaks in the groups without (control) and with products, and the significance level to reject the null hypothesis was considered to be 5% ($p<0.05$).³⁶

3.1.

Insect Repellent

3.1.1.

Raman spectra and most significative peaks

Figure 2 shows the Raman spectra of repellent products in spray and cream (undiluted products, groups CR, and SPR) placed on an aluminum sample holder, with the peaks at 526, 690, 1003, 1295, 1458, and $1608{\mathrm{cm}}^{-1}$ present in both formulations. Table 4 lists the active ingredients indicated on the product’s description label, the Raman peaks of the active ingredients found in the literature, and the indication whether these peaks are the present or not in the spectra of the products. Figure 2 shows peaks that match those of active ingredient DEET present in the repellent formulation according to Table 4.

Fig. 2

Raman spectra of repellents in cream (REF_CR) and spray (REF_SPR) formulations. The labeled peaks match the ones of the active ingredient DEET as presented in the reference spectrum of $N,N$-diethyl-m-toluamide in Chemical Book.³⁷

Table 4

Active ingredient present in the product’s description (label) of the repellents, characteristic Raman peaks of the active ingredient as identified in the literature and the presence on these peaks in the sample’s spectra.

Active ingredient	Presence in label		Characteristic Raman peaks
	SPR	CR	From the literature (cm−1)	From Fig. 2 (cm−1)
DEET	Yes	Yes	527, 690, 712, 1003, 1293, 1457, 1472, and 160837	526, 690, 1003, 1295, 1458, and 1608

Note: CR, cream; SPR, spray.

Figure 3 presents the mean Raman spectra of the skin from volunteers with the applied repellent products in spray (SPR) and cream (CR) as well as the mean spectra of the skin without repellent (CTR). The Raman peaks labeled in Fig. 3(a) correspond to the most relevant peaks of the active ingredient DEET. In fact, all these peaks are in positions close and overlapped to the peaks of the CTR skin. Figures 4(b)–4(d) magnify the peaks at 526, 690, and $1003{\mathrm{cm}}^{-1}$: at $526{\mathrm{cm}}^{-1}$ there is an overlap with the skin’s Raman peak at around $540{\mathrm{cm}}^{-1}$, attributed to the stretching vibration of the S–S disulphide bridge in skin proteins (actin, collagen, and elastin);²⁸ at $690{\mathrm{cm}}^{-1}$ the DEET peak is in a valley between the skin’s Raman peaks at 640 and $720{\mathrm{cm}}^{-1}$, attributed to proteins (C–C twisting of phenylalanine and tyrosine and C–S stretching of the proteins, respectively, the last one presenting contribution from the C–N vibration at $720{\mathrm{cm}}^{-1}$ from the choline of phospholipids);²⁹ and at $1003{\mathrm{cm}}^{-1}$ there is an overlap with the skin’s peak at $1004{\mathrm{cm}}^{-1}$ from proteins (aromatic ring vibration of phenylalanine and tyrosine).

Fig. 3

(a) Mean Raman spectra of the skin from volunteers with the applied repellent products in spray (SPR) and cream (CR) and the skin without repellent (CTR). Plots of the peaks at (b) $527{\mathrm{cm}}^{-1}$, (c) $690{\mathrm{cm}}^{-1}$, and (d) $1003{\mathrm{cm}}^{-1}$ to allow the observation of the overlap of DEET peaks on SPR and CR with the peaks of skin from controls (CTR).

Fig. 4

Plot of the mean intensity and standard deviation of the Raman peaks of the active ingredient DEET found in the spectra of the skin of the groups CTR, SPR, and CR: (a) peak $527{\mathrm{cm}}^{-1}$, (b) peak $690{\mathrm{cm}}^{-1}$, and (c) peak $1003{\mathrm{cm}}^{-1}$. The symbol “*” indicates a statistically significant difference between the groups SPR versus CTR and CR versus CTR ($p<0.05$, ANOVA, and Tukey post hoc tests).

3.2.

Sunscreen

3.2.1.

Raman spectra and most significative peaks

Figure 5 shows the Raman spectra of the sunscreen products (undiluted) of the brands CEN (SPF 15 and 30) and SD (SPF 15 and 30) placed on aluminum sample holder, with the peaks at 1003, 1177, 1288, 1310, 1564, 1605, and $1611{\mathrm{cm}}^{-1}$. The correspondence of the observed peaks with the active ingredients identified in the product’s description (label) is indicated in Table 5. The spectra of products show peaks in the same positions but differ in the intensity of the peaks. These differences are related to different concentrations of the active ingredients between the products of the two brands and SPFs. Also, some peaks are present in one brand and absent in another.

Fig. 5

Raman spectra of sunscreen products of the two bands CEN and SD in two different SPFs (SPF 15 and 30). The labeled peaks represent the most significative peaks.

Table 5

Active ingredients present in the product’s description (label) of the sunscreens, characteristic Raman peaks of the active ingredients as identified in the literature.

Active ingredient	Presence in label		Characteristic Raman peaks
	CEN15 and CEN30	SD15 and SD30	From literature (cm−1)	From Fig. 5 (cm−1)
Ethylhexyl methoxycinnamate (octinoxate)	Yes	No	1170 and 161338	1177 and 1605
Benzophenone-3	Yes	No	1000 and 128039	1003 and 1288
Octocrylene	Yes	Yes	156040	1564
Bis-octoxyphenol methoxyphenyl triazine	Yes	Yes	*	—
Ethylhexyl triazone	No	Yes	*	—
Butyl methoxydibenzoylmethane (avobenzone)	No	Yes	160541	1611

*Compounds with spectra not found in the literature

Figure 6 shows the mean Raman spectra of the groups with sunscreen products applied on the skin: CEN15, CEN30, SD15, and SD30 and the control group CTR. The labeled Raman peaks in Fig. 6(a): at 1003, 1177, 1288, and $1611{\mathrm{cm}}^{-1}$, are those identified in the products that correspond to the active ingredients as listed in Table 5. As occurred with the repellents, some of the peaks of the active ingredients overlap with the peaks of the CTR skin. Figures 6(b)–6(f) magnify the peaks at 1003, 1177, 1288, 1564, and $1611{\mathrm{cm}}^{-1}$, respectively. The spectra shown in Fig. 6(a) do not show the peak at $1564{\mathrm{cm}}^{-1}$, related to octocrylene; instead, Fig. 6(e) shows that there is an overlap of this peak with the peak from skin at around $1562{\mathrm{cm}}^{-1}$, associated to nucleic acids, proteins, and hemoglobin.²⁹ In addition, the proximity to the high-intensity peak in $1611{\mathrm{cm}}^{-1}$, which is associated to octinoxate and avobenzone, with the low intensity peaks from skin at $1613{\mathrm{cm}}^{-1}$, assigned to nucleic acids, proteins, amino acids (phenylalanine, tyrosine, and tryptophan), and hemoglobin,²⁹ allows the identification of this peak and it is used as reference for the identification of both octinoxate and avobenzone. Concerning the peak at $1003{\mathrm{cm}}^{-1}$, there is an overlap with the peak at $1004{\mathrm{cm}}^{-1}$, attributed to proteins (aromatic ring vibration of phenylalanine and tyrosine),²⁹ as occurred with the repellents.

Fig. 6

(a) Mean Raman spectra of the skin from volunteers with applied sunscreen products (CEN15, CEN30, SD15, and SD30) and from control (CTR). The plots of the peaks at (b) $1003{\mathrm{cm}}^{-1}$, (c) $1177{\mathrm{cm}}^{-1}$, (d) $1288{\mathrm{cm}}^{-1}$, (e) $1564{\mathrm{cm}}^{-1}$, and (f) $1605/1611{\mathrm{cm}}^{-1}$ allow the observation of the overlap of the peaks from CTR, CEN15, CEN30, SD15, and SD30 with the peaks of CTR skin.

3.2.2.

Statistical analysis of the peaks of sunscreen’s active ingredients identified on the skin

Figure 7 shows the mean intensity and standard deviation of the Raman peaks related to the active ingredients of sunscreens in the skin spectra in the groups CEN15, CEN30, SD15, SD30, and CTR. The results of ANOVA and Tukey post hoc tests applied to these intensities of the five groups are also shown. As the peaks at 1003 and $1288{\mathrm{cm}}^{-1}$ and peaks 1177 and $1611{\mathrm{cm}}^{-1}$ presented a statistically significant difference of the CEN15 versus CTR and CEN30 versus CTR, it is possible to state that these peaks indicate the presence of the sunscreen CEN applied to volunteer’s skin due to the peaks of the active ingredients benzophenone-3 (related to peaks 1003 and $1288{\mathrm{cm}}^{-1}$ according to Table 5) and octinoxate (related to peaks at 1177 and $1605{\mathrm{cm}}^{-1}$ according to Table 5). Similarly, as the peak at $1611{\mathrm{cm}}^{-1}$ presented a statistically significant difference of the SD15 and SD30 versus CTR, this peak indicates the presence of the sunscreen SD due to the peak of avobenzone (Table 5).

Fig. 7

Plot of the mean intensity and standard deviation of the Raman peaks of the active ingredients of the sunscreen products found in the spectra of the skin: (a) peak $1003{\mathrm{cm}}^{-1}$, (b) peak $1177{\mathrm{cm}}^{-1}$, (c) peak $1288{\mathrm{cm}}^{-1}$, and (d) peak $1611{\mathrm{cm}}^{-1}$. The symbol “*” indicates a statistically significant difference between the groups CEN15, CEN30, SD15, and SD30 versus CTR when indicated ($p<0.05$, ANOVA, and Tukey post hoc tests).

3.3.

Identification of the Active Ingredients Using Raman Spectroscopy

This work is the first to show the presence of the active ingredients of insect repellent and sunscreen products topically applied to the skin. Chrit et al.²⁵ investigated the effects of active ingredients for skin hydration using confocal Raman, showing that Raman spectroscopy was capable to show the hydration enhancement effect brought by the active ingredients. Mélot et al.⁴² investigated the effect of compounds used as helpers on retinol transportation through skin layers with confocal Raman, showing that Raman spectroscopy was effective to measure efficiency of formulation on transporting of desired molecules through the skin.

The use of the peaks from DEET at 527 and $1003{\mathrm{cm}}^{-1}$ permits us to detect the presence of insect repellents applied topically to the skin, besides the existing overlap with the Raman peaks of the skin. In fact, the differences in the positions of these Raman bands from the positions of the bands found in the skin, at $540{\mathrm{cm}}^{-1}$ (attributed to disulphide bridge of proteins) and at $1004{\mathrm{cm}}^{-1}$ (attributed to the aromatic ring vibration of phenylalanine and tyrosine), makes possible the identification of these peaks in skin of nontreated and treated volunteers (SPR and CR versus CTR) with statistically significant differences. The other peaks of the DEET are not significant due to their low intensity and overlap with the peaks of the skin.

An interesting application of this study is the detection of possible degradation of DEET under sunlight exposure, since Bório et al.⁴³ demonstrated that peaks at 524 and $1003{\mathrm{cm}}^{-1}$ from DEET are unstable when the repellent is irradiated by ultraviolet light (UVA and UVB, $7.0\mathrm{mW}/{\mathrm{cm}}^2$ for 8 h), which could potentially reduce its topical effectiveness after prolonged sun exposure. The study can also be applied in assessing the required amount of product to promote the desired repellent effectiveness since the Raman spectrum can detect the presence of the active ingredient quantitatively through its peak intensity.

The assessment of the active ingredients in the spectra of sunscreens indicated that Raman spectroscopy was able to identify the differences in the composition of the products under test. The presence of the peaks at 1003 and $1288{\mathrm{cm}}^{-1}$ (related to benzophenone-3) and peaks at 1177 and $1605{\mathrm{cm}}^{-1}$ (related to octinoxate) for the CEN brand, and the peak at $1611{\mathrm{cm}}^{-1}$ (related to avobenzone) for the SD brand, evidenced of the presence of related active ingredients in these products, thus suggesting the capability of discriminative (qualitative) and quantitative analyses of Raman spectroscopy. This capability is also demonstrated when the products are applied topically to the skin, including identifying the brand.

An interesting finding of the study was the difference between the spectra of sunscreens comparing the two different SPFs of the same brand. This difference is clear in the nonnormalized spectra of undiluted products (not shown). Comparing the intensity of the undiluted products, it is possible to check that the higher SPF results in higher intensities of peaks related to the active ingredients of the formula, since the brands keep the same formulation, changing only the concentration.

Raman spectroscopy provides qualitative information regarding the composition of the sample under analysis as well as quantitative information related to the presence of the compound, which can be useful to manufacturers to detect and quantify the active ingredients of pharmaceutical formulations topically applied to the skin and may be used for quality control during the production process, thus attending GMP and PAT technology processes.^17–19 As the manufacturers know the concentration of the active ingredients and excipients of the formulation, it is possible to observe the presence of each active ingredient in the skin of an individual following an application protocol, or even to perform tests for degradation due to environmental conditions or simulated situations of use.

The Raman spectra showed peaks of the active ingredients that can be used to detect and identify the presence of insect repellents and sunscreens topically applied to the skin. Raman spectroscopy is a noninvasive and nondestructive technique that can be applied to measure the composition of the samples in situ and in vivo and has shown a reliable and quick means for the identification of these topical products in the skin, helping the identification of these compounds in protocols of efficacy evaluation. Finally, the reason for the high fluorescence background in some volunteers after product application could be related to interactions of the product with the stratum corneum, which could be exploited in a future study.