Exploring the impact of fragrance molecular and skin properties on the evaporation profile of fragrances

Abstract

An evaporation study was conducted on the skin of several volunteers to examine whether skin properties influence fragrance molecules’ evaporation rates. The aim was to identify the observed variations and explore the responsible fragrance molecular and skin factors. To study the evaporation phenomenon, fragrance molecules evaporation was semi‐quantified from each volunteer's skin. This approach allowed a comparison of evaporation across individuals and provided insights into how the fragrance molecules interacted differently depending on skin types. Skin properties were subsequently measured to explain the observed differences in evaporation between individuals. Statistical analysis was performed to understand how both skin type and the intrinsic properties of fragrance molecules contribute to the observed variations in evaporation rates. This study offers promising results, advancing our understanding of the evaporation behaviour of fragrance molecules and its relationship with the physicochemical properties of the skin and the intrinsic characteristics of the fragrances.

Fragrance molecules release is influenced by skin type, with the effect varying depending on the specific fragrance compound. Two patterns of behaviour have been observed among fragrance molecules: the more volatile compounds whose evaporation increases with surface skin roughness and the less volatile and more lipophilic compounds influenced by skin hydration and transepidermal water loss.

INTRODUCTION

The skin is the body's heaviest and most extensive organ, performing numerous essential functions such as protection, thermoregulation, sensory perception, exchange, and metabolic activities. Structurally, the skin consists of three layers: the epidermis (from the Greek ‘epi’, meaning ‘top’, and ‘derma’, meaning ‘skin’), the dermis, which is the intermediate layer, and the hypodermis (from the Greek ‘hypo’, meaning ‘underneath’), the deepest layer. Depending on the individual, the skin is uniquely adapted to different environmental conditions, lifestyles, and genetic factors, resulting in variations in its structure, function, appearance, and needs. These differences manifest in various skin types, each with specific characteristics and care requirements. Understanding these variations is crucial for developing effective skincare routines and treatments tailored to individual needs.

The cosmetic market continues to grow, driven by increasing consumer demand for more personalized, effective, and innovative products. However, a significant challenge remains: how can we meet this demand when so much about how skin interacts with cosmetic ingredients is still unknown? This gap in knowledge about the interaction between skin and cosmetic molecules poses a critical obstacle to creating highly personalized and effective products.

A significant amount of in vivo research has been conducted on various aspects such as toxicological evaluation [1], active ingredients’ penetration [2, 3] and activity [4], skin barrier function [5], and surface properties like the pH gradient across skin layers [6], as well as the composition of epidermal lipids, using techniques like confocal laser scanning microscopy (CLSM) [7] and the sampling of human skin volatile organic compounds (VOCs) [8]. However, the retention and evaporation of fragrance molecules on the skin remain relatively underexplored and not fully understood. Understanding how fragrance molecules interact with the skin, how long they persist, and how they evaporate differently depending on the skin type is a crucial area of research. This knowledge is essential for determining the longevity and effectiveness of cosmetic products, though it presents considerable challenges.

Few studies in the literature investigate fragrance evaporation on the skin, and even fewer attempt to explain the observed phenomena. For example, some studies describe the evaporation profiles of fragrance molecules using mathematical models [9], while others are conducted in vitro [10, 11] or in vivo [12, 13, 14]. These studies highlight the variation in evaporation rates depending on the fragrance molecule and the complexity of the matrix. However, the underlying reasons for these varying profiles remain largely unexplored. Mookherjee et al. [15] sought to deepen the understanding of these different profiles, suggesting that high volatility is not the sole factor responsible for rapid evaporation. Instead, it involves a combination of highly volatile, middle‐boiling molecules, as well as high‐molecular‐weight, high‐boiling compounds that possess a characteristic called high diffusivity. It is clear that the intrinsic properties of the molecules play an important role and need to be investigated in greater depth to better understand their varying retention.

In addition to the molecular properties of fragrances, the properties of the skin surface also contribute to their retention. Skin is not an inert surface; it interacts with the molecules and significantly slows their evaporation compared with relatively free evaporation from a glazed surface [12]. According to Schwarzenbach et al. [16], three parameters seem to influence fragrance diffusion: surface temperature, fragrance molecules concentration, and the flow of air across the surface. No significant influence was observed from transepidermal water loss (TEWL), sebum concentration, or surface topography. However, this study by Schwarzenbach et al. is not complete enough, as it is based on a single volunteer and lacks sufficient measurements to draw general conclusions.

Another study explored the influence of skin properties on fragrance molecules evaporation using solid phase microextraction (SPME) technology on a panel of nine volunteers of different ages, ethnicities, and genders. The researchers concluded that no significant differences in evaporation rates were observed for any molecule based on skin differences, except for the higher boiling components, where some notable variations were noted. However, the article does not provide statistical analysis to support these observations, nor does it provide enough information about the skin properties of each volunteer to establish correlations with the observed differences in evaporation rates. Indeed, it is clear that further studies are needed to truly understand the effect of skin properties on the evaporation rates of molecules. This research is essential for better anticipating evaporation, and consequently, the perception of fragrances on the user's skin.

This article aims to explore the evaporation of fragrances depending on individuals' skin types, while also examining the fragrance molecular and skin properties responsible for these variations. The evaporation process was studied on the skin of 10 volunteers and compared with total evaporation from an inert surface, highlighting potential interactions between fragrances and human skin. Through semi‐quantification analysis, individuals were classified into groups based on the variation in evaporation rates. Finally, statistical analysis identified the key fragrance molecular and skin properties influencing fragrance molecules evaporation.

MATERIALS AND METHODS

Fragrance mixture composition and preparation

As described in a previous article [14], the primary objective was to select a mixture of molecules varying in chemical structures and physical properties to represent the chemical complexity in a perfume‐type mixture. Table 1 provides a summary of the selection criteria. All chosen molecules are soluble in ethanol, the solvent used for this mixture, and differ in their hydrophobicity (log P; where P is the n‐octanol–water partition coefficient), chemical structure (ester, alcohol, or terpene), boiling point, vapour pressure, and retention indices (RI). The RI were measured experimentally by injecting a mixture of hydrocarbons relative to C8–C24 on a 5% phenyl–95% polydimethylsiloxane (PDMS) column.

TABLE 1.

Summary table of the physicochemical properties (RI, logP, boiling point, vapour pressure, and molecular weight) of the eight molecules composing the fragrance mixture, ranked according to their RI.

Molecule	Chemical structure	RI	logP [17]	Boiling point (°C) [17]	Vapour pressure (mmHg at 25°C) [17]	Molecular weight (g/mol) [17]
Ethyl butyrate (ester)		806	1.85	121	14	116.16
Myrcene (acyclic terpene hydrocarbon)		988	4.17	167	2.1	136.23
Limonene (cyclic terpene hydrocarbon)		1027	4.57	176	1.55	136.24
Ethyl heptanoate (ester)		1095	3.33	188	0.68	158.24
Ethyl octanoate (ester)		1194	3.84	208	0.020	172.26
Citronellol (terpene alcohol)		1219	3.91	225	0.020	156.27
Ethyl decanoate (ester)		1398	4.86	245	0.034	200.32
Hexyl cinnamaldehyde – Jasmonal H® (aromatic and unsaturated aldehyde)		1743	5.33	308	0.002	216.32

The detailed composition of the fragrance mixture is also provided in the previous article [14].

Each component was present at 1% (w/w) except for citronellol and hexyl cinnamaldehyde, which were at 2% (w/w), all in a 96% ethanol solution. This formulation was designed to reflect an ‘eau de toilette’, where fragrance ingredients, dissolved in the solvent, account for 10% of the total mixture. The fragrance mixture is stored at 4–6°C between and after each application. The mixture of volatile compounds complies with IFRA51th regulations, adhering to the authorized concentrations for alcoholic perfumery (cat 4 IFRA) suitable for skin application and the CE 1223/2009 regulation.

Device to study in vivo fragrance molecules evaporation

To study the phenomenon of in vivo evaporation, an innovative system was designed and developed, as detailed in Hadjiefstathiou et al. [14].

The SPME technique was used to extract and concentrate the volatile compounds from the headspace (HS) . After equilibration and trapping times on the fibre, the compounds were analysed using an Agilent 8860 GC with a flame ionization detector (FID). Peak areas are automatically integrated using the Agilent software ‘Agilent OpenLab Control Panel’, which facilitates the semi‐quantification of each compound in the fragrance mixture. The resulting chromatogram allows the quantification of the compounds released into the HS above the studied surface. It should be noted that the data were processed using a GC correction factor as well as an SPME deformation factor.

Study participants

The study protocol conformed to the principles set forth by the Declaration of Helsinki and was approved by the URCOM Laboratory Committee on Human Research at the University Le Havre Normandie. All participants provided written informed consent before participation. Ten participants were recruited for the study: four young females (mean age 28.8 ± 3.6 years), two middle‐aged females (mean age 59.0 ± 4.2 years), three young males (mean age 28.7 ± 0.6 years), and one older male (aged 51), with no apparent signs of skin disease. The code given to the subjects consists of the first letter indicating gender (W for women and M for men), followed by the first two digits indicating age. For two subjects with identical gender and age, an arbitrary number, 1 or 2, was added.

In vivo biometrological measurements

All the measurements were performed under standardized conditions after a 15‐min resting period [room temperature and relative humidity conditions were 20.1 ± 0.3°C and 57.4 ± 1.0 RH, respectively]. To reduce bias, all measurements were conducted by the same operator. To ensure reproducibility, measurements were taken on the same area of the forearm of the volunteers before each evaporation study. A Corneometer® CM825, a skin‐pH‐Meter® PH905, a temperature probe ST 500, a Tewameter® TM 300, a Sebumeter® SM 815, and the Visioscan VC® 20plus (Courage‐Khazaka electronic GmbH MPA580, Germany) were used to measure stratum corneum hydration, pH, temperature, transepidermal water loss (TEWL), lipid index, and surface roughness, respectively. Five values were recorded for each measurement, except for TEWL and lipid index, where three values were recorded, and for roughness, where 10 values were recorded.

SC hydration

The Corneometer® CM825 provides the hydration level of the skin surface based on the capacitance measurement of a dielectric medium. It measures the change in the dielectric constant due to skin surface hydration by assessing the capacitance differences of a precision capacitor.

Cutaneous pH

Skin‐pH‐Meter® PH905 determines the skin's pH using a combined electrode that houses both an H⁺ ion‐sensitive electrode and an additional reference electrode. The probe, which contains the measurement electronics, was calibrated before each use.

Skin temperature

The Skin‐Thermometer® ST 500 measures skin temperature based on relative infrared temperature detection, expressed in°C. Measurements were recorded once the temperature had stabilized after sufficient contact with the skin.

TEWL

The Tewameter® TM 300 measures TEWL in g/m²/h on different skin zones. It indirectly gauges the gradient of water evaporation from the skin using two pairs of sensors (temperature and relative humidity) inside a hollow cylinder. Measurements were performed over a 2‐min period.

Lipid index (LI)

The Sebumeter® SM 815 measures sebum levels on skin, based on grease spot photometry. The Sebumeter®'s opaque tape becomes transparent upon contact sebum. The transparency of the tape is then measured by a photocell, with light transmission reflecting the sebum content in μg/cm².

Roughness

The Visioscan VC® 20plus consists of a UVA light source and a charge‐coupled device camera, which produces grayscale images of a measurement area sized 6 × 8 mm. The principle of the measurement is based on the analysis of the grey levels (where 0 value is black and 255 is white). Based on the described analysis of the grey level distribution, the software calculates various skin topography parameters. These parameters include the surface evaluation of living skin (SELS) parameters: smoothness (SE_sm), roughness (SE_r), scaliness (SE_sc), and wrinkles (SE_w), as well as the skin topography volume, V _vc (mm²) and unfolded surface, S _sc (%). Additionally, the ISO parameters R _1=t, R _3=z and R _5=a, can be determined. The ISO parameters were measured along 180 circularly arranged lines (star roughness), compensating for the influence of wrinkles direction. SE_r expresses the ratio of dark grey levels to the roughness of the whole image. For the generally observed roughness, the value is interpreted inversely: the smaller SE_r, the less rough the skin [18]. The opposite parameter is SE_sc which calculates the area occupied by bright pixels compared with the entire image. The smaller the SE_sc, the less pronounced the desquamation of the stratum corneum. The idea of SE_sm is that when very smooth, skin shows a lower variety of different levels of grey; thus, the histogram over the grey level distribution is more homogeneous and less wide. Hence, the smaller the SE_sm, the smoother and more homogeneous the skin surface. SE_w determines the number of dark pixels corresponding to the wrinkles; the more visible the wrinkles (broad, deep wrinkles), the higher the value. The skin topography V _vc is the core void volume, and S _sc is the arithmetic mean of the main curvature of the peaks on the surface. The roughness parameter R _1=t corresponds to the vertical distance between the highest peak and the lowest valley over the profile evaluation length. The average roughness R _3=z is the sum of the highest peaks and lowest valleys along the profile evaluation length, while the mean absolute deviation of roughness irregularities from the mean line over a profile evaluation length is R _5=a.

Statistical analysis

The statistical analyses of collected data were performed using XLSTAT® software from Addinsoft (version 2012). All the measurements were carried out at least in duplicates, and then the average values, the standard deviations, and the coefficients of variation (ratio of the standard deviation to the mean) were calculated. An analysis of variance (ANOVA) followed by Fisher's pairwise comparison test was applied to the data to determine when a specific skin property was significantly different according to the skin type (p ≤ 0.05), partial least squares (PLS) to predict and model relationships between variables, and principal component analysis (PCA) to reduce the dimensionality of large datasets, simplifying data exploration and visualization.

RESULTS AND DISCUSSION

The objective of this study was to investigate the evaporation phenomenon on the skin of volunteers and to observe whether skin properties have any effect. If an effect is observed, we aim to distinguish and understand the differences. As a first step, the evaporation phenomenon was studied on each volunteer through the semi‐quantification of fragrance molecules evaporation on their skin. This allowed us to compare the total evaporation on each volunteer and see how each molecule of the fragrance mixture interacted differently depending on skin type. The skin properties were then measured to explain the differences observed in evaporation between individuals. Finally, an explanation is provided to understand how skin type as well as the intrinsic properties of the molecules can influence evaporation.

Theoretical evaporation of fragrances based on their intrinsic properties

Molecules composing the fragrance mixture were chosen for their chemical complexity and their similarity to those found in commercial perfumes. By focusing on the molecules’ intrinsic properties and simulating a nearly pure evaporation scenario without interactions, we can better evaluate their behaviour on a living surface, such as human skin, where additional factors affect evaporation.

First of all, it is interesting to focus on the order of elution of molecular release from the GC chromatographic column and the order of evaporation quantities for each molecule measured with the SPME technique.

The sequence of molecules eluted in GC with an apolar stationary phase and a specific temperature gradient corresponds to their boiling points. For instance, ethyl butyrate vaporizes the fastest with a boiling point of 121°C as it requires the lowest energy. Following it are myrcene (167°C), limonene (176°C), ethyl heptanoate (188°C), ethyl octanoate (208°C), ethyl decanoate (245°C), and hexyl cinnamaldehyde (308°C), which vaporizes the slowest.

In terms of evaporated quantities, under ideal conditions, each molecule should evaporate at similar rates (except for citronellol and hexyl cinnamaldehyde, for which the concentration was doubled). In reality, several factors can disrupt this ideal evaporation. Experimental conditions, such as SPME deformation, the fibre's affinity for different molecules, and the compromise of equilibration times (as small, volatile molecules reach equilibrium faster than larger, less volatile ones), all impact results. Additionally, each molecule's intrinsic properties also play a role in diverging from theoretical evaporation as well as interactions between molecules.

In a ‘pure’ evaporation scenario, the influence of logP is less significant, as logP primarily relates to the partitioning between polar and non‐polar phases (e.g. water and oil). Instead, evaporation in such cases is primarily influenced by molecular weight, vapour pressure (Table 1), and intermolecular forces, with the kinetics being highly dependent on temperature.

Regarding molecular weight, lighter molecules evaporate faster since they require less energy to escape from the liquid phase. For example, ethyl butyrate (116.16 g/mol) is the lightest molecule in the fragrance mixture and will evaporate the fastest, whereas hexyl cinnamaldehyde (216.32 g/mol), the heaviest molecule, will evaporate the slowest.

When considering vapour pressure, molecules with higher vapour pressures at room temperature evaporate faster because they have a greater tendency to escape into the vapour phase. Ethyl butyrate, with a vapour pressure of 14 mmHg, will evaporate the fastest. In contrast, hexyl cinnamaldehyde, with a vapour pressure of 0.002 mmHg, will evaporate extremely slowly. Myrcene (2.1 mmHg) and limonene (1.55 mmHg) have moderate vapour pressures, suggesting that they will evaporate faster than the ethyl esters and citronellol.

Moreover, molecules with stronger intermolecular forces require more energy to overcome these attractions and evaporate. For example, citronellol contains hydroxyl groups that can form hydrogen bonds, making it more resistant to evaporation compared with molecules like ethyl butyrate, which only experience weaker van der Waals forces.

In summary, molecules with higher vapour pressures, smaller molecular weights, and weaker intermolecular forces evaporate more quickly and reach higher concentrations in the vapour phase. Conversely, molecules with lower vapour pressures, larger molecular weights, and stronger intermolecular forces evaporate more slowly, resulting in lower concentrations in the vapour phase. Experimental conditions also play a role, potentially favouring the evaporation of specific molecules.

In vivo fragrance molecules evaporation study on 10 volunteers

To study the evaporation phenomena of the fragrance mixture after in vivo application, a test campaign was designed. A group of 10 volunteers, selected to represent a diverse cross‐section of society in terms of age, sex, and ethnicity, had the fragrance mixture applied directly to their skin. To investigate the interactions between human skin and the studied molecules, and thus their retention, we also examined evaporation phenomena on a chemically inert glass surface to simulate evaporation without retention, as a reference. Using the set‐up described in the Materials and Methods section, we quantified the evaporation of each compound in the fragrance mixture, as well as the total quantity evaporated in vitro and in vivo for each volunteer. The objective was to assess the behaviour of each molecule on living skin compared with an inert surface, while simultaneously comparing the total evaporation across different skin types.

First, let's focus on the differences in evaporation rates of each molecule individually, both in vitro and in vivo. Figure 1 shows the difference in the mean area of each evaporated molecule between the average skin surface and the glass surface. Subtracting the two values provides insight into the impact of the skin on compound retention and whether the skin surface promotes or hinders the retention of molecules. A positive value indicates greater evaporation on the skin than on the glass, while a negative value indicates retention on the skin. What stands out is that the quantity of hexyl cinnamaldehyde does not vary significantly between the skin and glass surfaces, with a slight increase of 4 UA in evaporation on the skin. Limonene and myrcene, with evaporation rates of 59 and 62 UA, respectively, show more evaporation on the skin than on the glass surface. The remaining five molecules evaporated less on the skin than on the glass, with a progressive decrease: ethyl butyrate, ethyl heptanoate, ethyl decanoate, citronellol, and ethyl octanoate, showing 91, 454, 473, 1018, and 1182 UA, respectively, more evaporation on the glass surface than on the skin.

FIGURE 1.

Difference in the area of each molecule evaporated (A_skin‐Aglass); A positive value indicates greater evaporation on the skin than on the glass, while a negative value indicates retention on the skin.

To explain these observed differences, it is essential to consider the intrinsic properties of each molecule, as previously detailed, as well as their potential interactions with the skin. Limonene and myrcene exhibit higher evaporation rates on the skin than on a glass surface. Their volatility allows them to escape into the air, with minimal retention by the skin. One possible explanation is that the skin has a certain level of humidity due to the sweat it produces. This slightly humid environment may enhance the evaporation of volatile molecules, as water vapour can interact with the molecules, allowing them to move more freely into the air.

In contrast, the minimal difference in evaporation rates of hexyl cinnamaldehyde between skin and glass suggests that this compound does not interact more with one surface than the other. As a relatively heavy molecule, this intrinsic property appears to have a greater influence.

When comparing the evaporation of fragrance molecules (ethyl butyrate, ethyl heptanoate, ethyl octanoate, ethyl decanoate, citronellol) on the skin versus an inert surface like glass, several factors can account for the observed differences in retention, without initially considering specific values. Both esters and alcohols contain functional groups that can form hydrogen bonds or van der Waals interactions with skin‐polar lipids or proteins, such as keratin [19]. These interactions influence both the absorption and retention of fragrance molecules on the skin, which helps to explain why certain molecules evaporate at different rates in vivo compared with an inert surface like glass.

If we compare esters with alcohols, the higher polarity of alcohols allows them to form more hydrogen bonds with the skin's outermost layer, the stratum corneum, which is composed of keratin proteins and lipids. This interaction enables alcohol molecules to ‘stick’ to the skin more effectively, slowing their evaporation compared with esters. Esters, with their polar ester group (‐COOR), are generally less polar than alcohols. The ester group can form weaker interactions with the skin, but these interactions are not as strong as those formed by alcohols. As a result, esters tend to evaporate more quickly from the skin's surface and are less retained.

The four esters, ethyl butyrate, ethyl heptanoate, ethyl decanoate, and ethyl octanoate, show distinct evaporation behaviours in this study, with a clear trend indicating that they evaporate more on the glass surface than on the skin. Specifically, they exhibit progressively greater differences in evaporation rates, with ethyl octanoate showing the largest difference (1182 UA more evaporation on the glass surface), followed by ethyl decanoate (473 UA), ethyl heptanoate (454 UA), and ethyl butyrate (91 UA).

Larger molecules generally have lower volatility; therefore, they evaporate more slowly compared with smaller molecules. As the chain length increases, the intermolecular forces (such as van der Waals interactions) also become stronger. Ethyl octanoate and ethyl decanoate, with their longer carbon chains, have stronger intermolecular forces, which require more energy to overcome, making their evaporation slower. In contrast, ethyl butyrate, being smaller and more volatile, may evaporate more easily from the skin; hence, it shows a smaller difference in evaporation compared with the glass surface.

However, ethyl octanoate appears to behave differently from the other esters, as its evaporation rate does not follow the expected trend of progressively decreasing evaporation on skin with increasing chain length. This suggests that atypical environmental conditions may be influencing the evaporation dynamics. The reason ethyl octanoate does not follow the same trend as the other esters can be attributed to its intermediate volatility, moderate hydrophobicity, and significant interaction with skin lipids as well as its molecular structure and geometry [20].

It is evident that skin is not a typical surface, which is clearly reflected in the altered evaporation behaviour of the molecules. The following part of the article presents an original and complementary way of studying the various possible interactions at the skin surface by studying the release/retention of volatile probes. To deepen our understanding of these phenomena, the next step is to examine the evaporation profiles of each molecule across different skin types. We will investigate whether significant differences exist, and if so, explore the underlying causes and explanations for these variations.

Now, let's discuss the behaviour of each molecule on different skin types, which appears to vary. Results are shown in Figure 2 for each volunteer, with an increase in the quantities evaporated from left to right. Greater evaporation could lead to a more intense perception of the fragrance's odour. It is also possible to discern the distribution of each molecule for each volunteer, allowing for a better assessment of the evaporation profile of the complete mixture as a function of the subjects. At first glance, the molecules are fairly evenly distributed. When focusing now on the total evaporated quantities, five groups were observed according to ANOVA statistical tests (p ≤ 0.05), indicating a difference in the retention of fragrance molecules depending on the skin type.

FIGURE 2.

Quantities evaporated for each volunteer obtained by GC‐FID when applied to the forearm of each volunteer, with an increase in the quantities evaporated from left to right. (A–E) Values with a different letter for each subject indicate a significant difference (p ≤ 0.05).

Table 2 complements the results by presenting the coefficients of variation (CV) for the evaporated quantities of the studied molecule across different panellists. The CV is the ratio of the standard deviation to the mean value. A higher CV signifies a greater level of variability. In this context, a higher CV for a molecule suggests more variable evaporation rates across skin types, indicating distinct behaviours on various skin types. Depending on the molecule, this variability may be more or less pronounced. Molecules are ranked increasingly. For example, ethyl butyrate shows minimal fluctuation across skin types, whereas hexyl cinnamaldehyde exhibits greater variability.

TABLE 2.

Summary table of the molecules present in the fragrance mixture with their corresponding coefficients of variation, calculated according to the quantities evaporated on each subject, ranked increasingly from top to bottom.

Molecule	Coefficient of variation
Ethyl butyrate	5
Ethyl octanoate	11
Citronellol	14
Myrcene	17
Ethyl decanoate	19
Limonene	20
Ethyl heptanoate	22
Hexyl cinnamaldehyde	32

To better illustrate these variations and highlight the complex behaviour of molecules across skin types, radar charts have been generated for each molecule. These charts reveal three distinct evaporation patterns, providing additional insights that complement the sequence in Table 2. Myrcene, limonene, ethyl heptanoate, and ethyl octanoate (Figure 3) behave similarly across all volunteers, displaying a consistent trend. Citronellol, ethyl decanoate, and hexyl cinnamaldehyde (Figure 4) exhibit a different pattern. Finally, a third trend is observed with ethyl butyrate (Figure 5).

FIGURE 3.

Radar illustration of the evaporation profile of myrcene, limonene, ethyl heptanoate, and ethyl octanoate on each subject.

FIGURE 4.

Radar illustration of the evaporation profile of citronellol, hexyl cinnamaldehyde, and ethyl decanoate on each subject.

FIGURE 5.

Radar illustration of the evaporation profile of ethyl butyrate on each subject.

Focusing on ethyl butyrate (Figure 5), its behaviour appears consistent across all skin types. This suggests that skin properties do not significantly influence its evaporation, and that its evaporation is primarily governed by its intrinsic properties. This could be due to its small molecular size and high volatility, which favour evaporation over interaction with the skin.

In contrast, the differences in the behaviour of the compounds like myrcene, limonene, ethyl heptanoate, and ethyl octanoate (Figure 3) compared with citronellol, ethyl decanoate, and hexyl cinnamaldehyde (Figure 4) likely arise from variations in their chemical structure, molecular weight, functional groups, and volatility. These factors influence how each compound interacts with its environment, in this case, the skin. The first group, myrcene, limonene, ethyl octanoate, and ethyl heptanoate, consists of molecules with lower molecular weights and weaker intermolecular forces, making them more volatile. In contrast, citronellol, ethyl decanoate, and hexyl cinnamaldehyde are larger molecules with higher molecular weights, leading to stronger intermolecular forces, such as van der Waals forces and hydrogen bonding (in the case of citronellol), and consequently lower volatility. Additionally, the structures of these larger molecules, which allow for more freely rotating bonds, make them more prone to interactions. Furthermore, nonpolar compounds like myrcene and limonene tend to have a higher affinity for the lipidic compounds present on the skin surface. Conversely, more polar compounds like citronellol or hexyl cinnamaldehyde are likely to interact more strongly with polar solvents, such as water or ethanol.

With these fragrances now categorized based on their behaviour on different skin types, the next step is to understand the underlying causes of these trends. Specifically, why do evaporation rates vary among individuals, and how do skin types influence retention?

By closely examining Figure 2 to begin interpreting the relationship between skin type and evaporation quantities, we observe that the quantities generally increase from left to right, with a tendency for higher evaporation rates on male skin compared with female skin, except for volunteer M28. These differences may be explained by gender‐linked differences observed in skin properties. Indeed, based on gender, hormone metabolism, sweat rate, sebum production, or surface pH differs [21, 22]. However, these are just some preliminary hypotheses, and a larger sample of participants could help confirm these findings. Age is also a factor that may affect SC function, structure, and composition, as well as TEWL differences [23]. Indeed, the authors report that the TEWL rates generally decrease with age, indicating an overall improvement in SC barrier function. However, when comparing the evaporated quantities among volunteers of different ages, a clear difference is not observed, making it difficult to distinctly differentiate between the groups.

One other possible explanation for the differences observed between subjects could be their skin type. For instance, comparing skin phototypes, volunteer W27 exhibits a higher phototype. Given the functional differences between black and white human skin, it is reasonable to expect that their behaviour might differ. Indeed, Reinertson et al. [24] studied the SC lipid content on different abdominal skin of African and Caucasian skins and demonstrated higher values in African subjects than in Caucasian ones. A higher level of lipids can result in stronger fragrance molecules retention on the skin, for example. Furthermore, according to Wilson et al. [25], African skin has a significantly higher mean TEWL than Caucasian skin. A higher TEWL could potentially influence the HS concentration of water vapour distillable molecules but is not sufficient to explain the complexity of the observed phenomena.

Although there are five groups, no single criterion (age, gender, skin phototype) clearly explains these differences in evaporation rates. The differences observed in the quantities evaporated seem to stem from deeper reasons that could be elucidated with measurements of the skin properties of each of the subjects studied. The next step is to link the average evaporated quantity and the observed trends for each molecule, according to skin type, with the physicochemical properties of the skin, such as sebum levels, pH, temperature, TEWL, hydration, sebum, and skin roughness.

Skin properties measurements

As previously mentioned, statistical analysis identified five distinct groups of individuals based on the quantities evaporated. Additionally, it was shown that the evaporation profiles of fragrance molecules vary depending on the molecule and skin type. The next step involved collecting detailed skin properties for each volunteer, comparing our data with existing literature, and conducting a deeper investigation to determine whether these five groups can be further differentiated and to explore precisely how specific skin properties influence the evaporation of individual molecules.

Table 3 gathers all the skin data collected for each volunteer concerning pH, temperature, TEWL, hydration, lipid level, and roughness parameters (SE_R, SE_sc, SE_sm, SE_w, R ₁, R ₃, R ₅, V _vc, S _sc) measured on the forearm skin. Focusing first on the pH values collected, the average values for each gender were calculated to be 4.9 ± 0.4 for women and 5.1 ± 0.5 for men. When comparing our data with the literature for the same area, the forearm skin, we observe similar pH values (5.3 ± 0.5 for women and 4.5 ± 0.2 for men [6] in one study, and 5.6 ± 0.4 for women and 4.3 ± 0.4 for men [26] in another study). Comparing the pH values of women and men from the literature, we observe that men's skin surface is more acidic than women's. According to our results, this difference is not discernible (p > 0.05). Regarding temperature, the average temperature obtained from our panel is 30.0 ± 0.9°C, which is similar to the value found in the literature for the lower forearm, 29.2 ± 0.9 [27]. Another skin property collected was TEWL, with an average value of 7.3 ± 1.5. Comparing our mean value with the literature, one can see that the values are very close. The average forearm TEWL value of 150 female subjects aged from 18 to 80 years was 9.1 ± 0.4 [28], and in another study, it was 7.9 ± 1.4 for 20 females and 10.0 ± 2.6 for 20 males [5]. Hydration was also measured as a skin property. Analysing the collected data, we observe rather low hydration values in our subjects. Indeed, with an average value of 34.9 ± 5.4, the average moisture‐related skin type is classified as dry skin. Our study is based on a group of 10 volunteers, so the chances of having chosen a single skin type are higher compared with a study of 150 volunteers, where the average value obtained is closer to normal skin, 44.1 ± 3.0 [28]. Sebum was also measured on the forearm of each volunteer with values approaching zero. Indeed, the sebum level on the forearm area can be considered negligible, close to zero [21, 28].

TABLE 3.

Summary table of the skin parameters (pH, temperature, TEWL, hydration, lipid level, and roughness parameters (SE_R, SE_sc, SE_sm, SE_w, R ₁, R ₃, R ₅, V _vc, S _sc) measured for each volunteer.

Subject	pH	Temperature	TEWL	Hydration	Sebum	SEr	SEw	SEsc	SEsm	R 1	R 3	R 5	V vc	S sc
W62	4.8	29.4	5.7	43.4	0.5	1.7	89.4	0.5	297.8	102.8	78.1	23.7	84.1	618.6
W27	5.1	29.2	6.8	39.4	1.6	1.8	43.0	0.0	146.0	87.3	71.4	17.1	65.7	634.7
W56	4.0	29.0	6.1	37.4	0.0	2.2	91.1	0.1	269.7	95.7	71.4	22.0	72.1	556.3
W26	5.2	31.4	7.6	36.4	2.0	2.9	64.9	0.0	165.2	73.8	56.6	14.5	61.4	534.1
M28	5.3	29.0	6.5	35.4	0.8	2.4	58.5	0.0	207.7	97.4	73.5	19.6	74.2	664.8
W34	5.1	30.9	10.6	29.8	0.0	1.8	72.8	0.0	231.2	91.9	69.1	21.0	72.0	588.6
W28	4.9	29.7	7.0	40.0	2.2	1.7	62.4	0.1	230.1	100.2	75.0	22.1	83.5	676.6
M292	4.9	31.0	5.7	31.7	5.4	1.2	50.9	0.3	212.6	129.9	102.6	22.4	86.8	822.4
M291	4.8	30.6	8.8	26.9	0.0	2.7	60.4	0.2	186.7	89.9	68.8	17.9	71.3	569.3
M51	5.3	30.0	7.9	29.2	0.0	3.1	106.0	0.8	313.7	124.1	94.2	22.8	117.4	613.7
Average ± standard deviation	5.0 ± 0.4	30 ± 0.9	7.3 ± 1.5	34.9 ± 5.4	1.2 ± 1.7	2.2 ± 0.6	69.9 ± 19.8	0.2 ± 0.2	226.1 ± 54.8	99.3 ± 16.7	76.1 ± 13.2	20.3 ± 3.0	78.8 ± 15.8	627.9 ± 82.2
p value	0.097	0.063	0.032	0.011	0.002	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001

If we now focus on the roughness parameters obtained with the Visioscan probe and compare the data from the present study to those found in the literature for the same skin zone, we find that the SELS parameter values are very similar. As in literature, the values of SEr were higher for men than for women (3.90 ± 1.63 and 2.34 ± 0.86, respectively; p < 0.05 [18]) and in our study (2.4 ± 0.5 and 2.0 ± 0.4). Higher values of SEr are associated with a greater number of skin lines and furrows, indicating that men's skin image has more profound wrinkles. Similar implications followed from SEw values, though the differences are less pronounced: 68.9 ± 20.2 for men and 70.6 ± 18.1 for women. Comparing the values of the SEsc parameter from our study with those from Dabrowska et al. [18] study, our values obtained for the two genders are not gender‐discriminating, whereas they are in the literature (0.61 ± 0.33 for female and 0.56 ± 0.40 for men; p < 0.05), indicating drier skin and more visible desquamation of the stratum corneum in women. The last SELS‐parameter, SEsm, was lower for women than for men in both studies (108.52 ± 57.47 and 121.98 ± 40.63, respectively; p < 0.05) and in our case (223.3 ± 58.6 and 230.2 ± 65.4, respectively), demonstrating smoother and more homogeneous skin for women in comparison to the skin relief of men. The average age of this study being 37 ± 13.8 versus 24 ± 3 in Dabrowska et al. [18] study could explain the differences observed when comparing the values obtained.

Concerning the ISO roughness parameters, these have also been studied for the same area, the forearm skin, by Kottner et al. [29] and Trojan et al. [30]. Differences can be observed from one study to another, but this is not worrying as the values are in the same order of magnitude, taking standard deviations into account. These differences can be explained by various factors, such as the group of subjects studied (phototype, age, gender, number of subjects) or even the differences in the probe used for the measurements. When comparing the roughness parameters R ₁, R ₃, and R ₅, in line with those of the SELS parameters, we saw that men have rougher skin than women (R ₁: 92.0 ± 10.5 and 110.3 ± 11.7, respectively, R ₃: 70.3 ± 7.4 and 84.8 ± 8.3, respectively, R ₅: 20.1 ± 3.5 and 20.7 ± 3.9, respectively). Roughness is also influenced by age. We have divided people into two groups: those under 35 and those over 50. Younger skin (28.7 ± 2.6 years) is less rough than older skin (56.3 ± 5.5 years) (R ₁: 95.8 ± 17.3 and 107.5 ± 14.8, respectively, R ₃: 73.9 ± 14.0 and 81.2 ± 11.7, respectively, R ₅: 19.2 ± 2.9 and 22.8 ± 0.8, respectively). Regarding the V _vc and S _sc parameters introduced in the Materials and Methods section, those two parameters are in line with the R ₁, R ₃, and R ₅ parameters for men and women (V _vc: 73.1 ± 9.2 and 87.4 ± 10.3, respectively, S _sc: 601.5 ± 52.5 and 667.6 ± 58.3, respectively) and young and older subjects (V _vc: 73.5 ± 9.0 and 91.2 ± 23.5, respectively, S _sc: 641.5 ± 95.0 and 596.2 ± 34.6, respectively).

Now that the skin properties have been measured for each panellist, it remains to be understood whether the difference in evaporated quantities is related to these properties. This issue will be further investigated in the next section.

Skin properties influencing evaporation

The evaporation of fragrances on the skin is a complex phenomenon that depends on both the intrinsic properties of the molecules and the properties of the skin itself. Concerning the impact of the physicochemical properties of the skin, it is first possible to rank the properties according to their impact on the total quantity of fragrance molecules evaporated. For this purpose, a statistical analysis using PLS (partial least square) regression was performed on the total amount of evaporation observed on each individual (the dependent variable), to predict evaporation and identify the importance of the explanatory variables represented by the set of skin properties. Table 4 presents the scores for variable importance for the projection (VIP), which provide an estimate of the contribution of each predictor to the PLS regression model, along with the percentage of importance for each skin parameter.

TABLE 4.

Summary table of the variable important for the projection (VIP) scores and the corresponding % of importance for each skin parameter from the PLS regression.

Variable	VIP	% importance VIP
Hydration	1.9	15.1
V vc	1.3	10.9
R 1	1.3	10.8
R 3	1.2	10.1
TEWL	1.1	8.6
R 5	0.9	7.6
pH	0.9	7.5
S sc	0.9	7.3
Temperature	0.9	7.3
SEsc	0.8	6.6
SEr	0.5	3.8
SEsm	0.3	2.1
Sebum	0.2	2.0
SEw	0.0	0.4

Note: VIP scores greater than or equal to 1, presented in bold, were considered significant contributors.

To highlight the skin physicochemical properties that most influence the evaporation of fragrances, only variables with a VIP score greater than or equal to 1 were considered significant contributors [20]. The skin parameters that have the most significant impact on evaporation are thus hydration followed by roughness parameters (V _vc, R ₁, and R ₃), and TEWL.

Once the parameters that most significantly impact evaporation are identified, the next step is to propose hypotheses to understand their role. The first and most influential parameter is hydration. Skin hydration plays a crucial role in evaporation because the water content of the skin affects its permeability and overall barrier function. Well‐hydrated skin tends to have lower evaporation rates, as the moisture content helps retain the fragrance molecules on the skin for a longer period. In contrast, when the skin is dry, evaporation rates increase because there is less moisture to interact with the fragrance molecules.

When considering surface roughness parameters, three key factors, V _vc, R ₁, and R ₃, stand out as being particularly relevant to evaporation. These parameters appear to directly impact evaporation rates. One can suggest that rougher skin surfaces provide more area for fragrance molecules to interact with, potentially altering how they evaporate. A rough surface can slow down the fragrance molecules release due to the increased surface area that holds the fragrance molecules. However, focusing on this single parameter is not sufficient to explain the observed differences between subjects.

When considering TEWL, which measures the amount of water evaporating from the skin surface, it is also a key indicator of the skin's barrier function. High TEWL suggests that the skin is losing more moisture, which can accelerate the evaporation of fragrances, as water molecules carry water‐soluble fragrance components with them. Conversely, lower TEWL indicates a better skin barrier, which can slow down the evaporation process. Once again, this theoretical explanation cannot directly explain our results (Table 3) and should be considered through a multiparameter approach.

For this purpose, to relate these influencing parameters to the evaporation of each of the fragrance compounds, a PCA was carried out using all the data. Figure 6 shows the biplot of the PCA with the individuals/subjects represented by black dots and the variables (chromatographic areas of each compound in italic and skin parameters in a frame).

FIGURE 6.

PCA biplot with the different subjects as individuals and variables (both quantities evaporated for each fragrance molecule and more influencing skin parameters).

This analysis reveals two clusters of fragrance compounds' behaviour. In the first cluster, there are ethyl butyrate, myrcene, limonene, ethyl heptanoate, and ethyl octanoate. The evaporation of this cluster appears mainly correlated with the parameters R1, R3, and Vvc. For these molecules, the quantity evaporated is greater when skin roughness parameters increase. It is interesting to note that these compounds have the lowest RI in Table 1 due to their lower boiling points. The five fragrances that create this cluster also correspond to the previous groups based on evaporation profiles (Figure 3), with the addition of ethyl butyrate, which is the compound least influenced by skin type (Figure 5).

The second group of fragrance compounds includes citronellol, ethyl decanoate, and hexyl cinnamaldehyde. Their evaporation is correlated with transepidermal water loss (TEWL) and is inversely correlated with hydration. This cluster corresponds to molecules with a higher boiling point, which are eluted last in our chromatographic analysis (Table 1). These lipophilic, low‐volatility compounds appear to be strongly influenced by the dynamics of water evaporation and the skin's state of hydration. Again, the molecules of the cluster were previously regrouped in Figure 4. They have similar evaporation profiles.

Finally, one can look at the position of the subjects/individuals in Figure 6. It can be seen that some subjects tend to be in the lower part of the PCA (M51, M291, M292, and W34), while others tend to be in the upper part of the PCA (W62, W27, W56, and W26). We can relate this projection to the data in Figure 2. Individuals with the highest fragrance molecules evaporation in Figure 2 are positioned in the lower part of the PCA and individuals with lower fragrance molecules evaporation are positioned in the upper part. We can therefore assume that fragrance molecules evaporation will be higher for rough skin and/or a skin characterized by a high TEWL. Individuals with moisturized, less rough skin will tend to evaporate fragrances less. These results need to be verified with a larger group of subjects.

CONCLUSION

This article presents a study on the evaporation of fragrance molecules in vivo to explore the influence of skin properties on evaporation rates. It has been demonstrated that fragrance molecules behave differently depending on their intrinsic properties, and that skin properties significantly impact the release of fragrance molecules. To investigate the evaporation phenomenon, fragrance molecules evaporation was semi‐quantified on each volunteer's skin using HS analysis with SPME coupled with GC‐FID.

Results showed that fragrance molecules evaporation was influenced by skin type, particularly hydration, TEWL, and roughness parameters. We were also able to distinguish different clusters of fragrance compounds depending on their intrinsic properties. Two patterns of behaviour have been identified among fragrance compounds: the more volatile compounds whose evaporation increases with surface skin roughness and the less volatile and more lipophilic compounds influenced by skin hydration and TEWL. The findings from this study provide a deeper understanding of the evaporation process of fragrance molecules and their relationship with both fragrance molecular and physical skin properties. The work also demonstrates that fragrance molecules release varies depending on the molecule chemistry. This suggests that focusing solely on the physical parameters of the skin is not enough, as the chemical composition of the skin can also play a key role in explaining the evaporation of each molecule. Expand the study to include other areas of the skin (e.g., forehead, face, etc.), each with its own distinct chemical composition of the hydrolipidic film, and its specific microrelief, would be valuable. This could further strengthen the conclusions and shed light on other skin properties, such as surface energy, which also influence the release of fragrance molecules.

CONFLICT OF INTEREST STATEMENT

None.

Hadjiefstathiou E, Savary G, Malhiac C, Terescenco D, Picard C. Exploring the impact of fragrance molecular and skin properties on the evaporation profile of fragrances. Int J Cosmet Sci. 2025;47:981–995. 10.1111/ics.13085