Functional Performance and Safety Evaluation of Optimized Plant-Based Dye Mixtures for Intense Hair Coloration

Abstract

This study aimed to develop a sustainable and safe alternative to chemical hair dyes by exploring the functional performance and safety evaluation of herbal mixtures. Natural dyes were extracted from Lawsonia inermis leaves, Clitoria ternatea flowers, and Indigofera tinctoria leaves using an eco-friendly extraction method with deionized water and ultrasonication. The ratios of these natural dyes were optimized using statistical tools, specifically Minitab, to determine the most effective formulation. The safety profiles and dyeing performance of individual dyes and their combinations were evaluated with a focus on color intensity, stability, and resistance to washing and light exposure. The optimal herbal mixture, with a ratio of 2:2:1:1 (L. inermis/C. ternatea/I. tinctoria/water), demonstrated the highest absorbance and lowest lightness, indicating the darkest color profile. When applied for 30 min, this herbal combination yielded a long-lasting and intense dark color. Further testing using the hen’s egg chorioallantoic membrane test confirmed the favorable safety profile, and examination under a scanning electron microscope showed no damage to the hair cuticle, indicating that the herbal formulation is safer than chemical hair dyes. Therefore, this herbal mixture showed promise as an alternative to synthetic dyes, and further formulation development was suggested.

1. Introduction

Hair color is a strong indicator of age, and in an ageist society, aging is often stigmatized [1]. Generally, half of all adults have gray hair by the age of 50 [2]. Although gray hair occurrence is related to aging, there are many factors that could induce premature gray hair, such as genetics, oxidation, metabolism, and stress [3]. Gray hair develops due to a decrease in melanin, which is a pigment produced by melanocytes located in the hair follicles [4,5]. Until now, there has been no solution to prevent or solve gray hair by dealing with the melanin production in the hair follicles. However, social and technological advancements have made hair dyeing a common practice for concealing gray hair, enabling individuals to mask signs of aging and appear younger [1,3].

Nowadays, synthetic dyes are widely used due to their lower cost of production and brighter colors, and they offer a broad spectrum of color options [6]. However, synthetic dyes pose significant risks to human health and the environment, including safety hazards for workers, such as cancer caused by toxic mordants, like ferrous sulfate [7,8]. In response to the negative effects of synthetic dyes, there has been increasing interest in natural plant dyes over the past few decades. Although natural plant dyes may not always be cost effective, they are preferred for their lower toxicity to humans [8]. Additionally, natural dyes are considered eco-friendly as they are sourced from renewable resources and are biodegradable [9].

In addition to the safety of natural hair dyes, other biological effects have been recognized as their significant advantages. Henna (Lawsonia inermis), a widely utilized natural hair dye, is acclaimed for its dual functionality in enhancing hair color while removing excess scalp sebum and providing conditioning effects [10]. Additionally, the butterfly pea flower (Clitoria ternatea), widely known as a rich source of anthocyanin pigments, has been reported to provide sebum control and demonstrate effectiveness similar to ketoconazole in alleviating dandruff [11]. Additionally, C. ternatea extract was found to promote the proliferation of human dermal papilla cells and stimulate initial hair growth in C57BL/6Mlac mice, similar to the effects of minoxidil [12]. On the other hand, Indigofera (Indigofera tinctoria) known for containing the economically significant indigo dye, which is commonly used as a hair colorant [13], has been reported to exhibit remarkable antimicrobial activity [14]. Although these natural dyes have been traditionally and recently used commercially for hair dyeing, a combination of various natural dyes is recommended to achieve a natural dark brown hair color. However, most existing research has not explored the optimization of natural dye combinations to enhance hair dyeing efficacy or addressed their long-term stability and resistance to external factors, like washing and light exposure. Additionally, while individual natural hair dyes have been studied, no comprehensive evaluation has been conducted to assess the safety and efficacy of the natural hair dye combination.

Therefore, this study aimed to provide a sustainable and safe alternative to chemical hair dyes, highlighting the potential benefits of herbal mixtures in cosmetic applications. Natural dyes from L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves were extracted and optimized for their ratios using statistical tools, such as Minitab, which played an instrumental role in efficiently developing and fine tuning natural herbal mixtures for optimal results. In addition, the safety profiles and hair dyeing performance of the individual dyes, as well as their combinations, were evaluated to determine their effectiveness and suitability for cosmetic use. The novelty of the current study lies not only in the integration of natural and sustainable resources but also in the optimization of natural dye combinations to enhance dyeing efficacy.

2. Materials and Methods

2.1. Plant Materials

L. inermis leaves were collected from the agricultural field of Chulabhorn Royal Pharmaceutical Manufacturing Facilities by Chulabhorn Royal Academy, Chon Buri, Thailand in November 2023. The geocoordinates of the harvest location were 12.66302° latitude and 100.97702° longitude. The leaves were separated from the branches and washed with reverse osmosis (RO) water, which is within the acceptance criteria of the RO water controlled by the Chulabhorn Royal Pharmaceutical Manufacturing Facilities by Chulabhorn Royal Academy following the water for pharmaceutical use in United States Pharmacopeia (USP) with the conductivity below 1.3 µS/cm at 25 °C [15]. Subsequently, the plant materials were dried in a hot air oven (Memmert GmbH+ Co. KG, Schwabach, Germany) set at a temperature of 50 °C overnight (12–16 h) until dryness. The dried leaves of L. inermis were ground into a fine powder using a stainless-steel grinding machine (HR-2200, Zhejiang Harui Industry and Trade Co., Ltd., Ningbo, China) and kept in a sealed aluminum bag protected from light until further used. On the other hand, dried C. ternatea flower and dried I. tinctoria leaf powders were purchased from the local market in Chiang Mai, Thailand. All dry plant materials were characterized and evaluated for particle sizes under a stereo microscope (Olympus, Tokyo, Japan). Additionally, the moisture contents of each dry plant material were assessed using an HC103 Moisture Analyzer (Mettler Toledo, Greifensee, Switzerland) [16]. Briefly, 1 ± 0.1 g samples were heated at 105 °C until stable weight was achieved within 50 s. The experiments were performed in triplicate.

2.2. Chemical Materials

Sodium chloride (NaCl), sodium hydroxide (NaOH), hydrochloric acid (HCl), and sodium lauryl sulfate (SLS) were of analytical grade and were purchased from Sigma-Aldrich (St. Louis, MO, USA). Chemical hair dye (Garnier Color Naturals, shade 1, natural black, L’Oreal (Thailand) Ltd., Bangkok, Thailand) was purchased from a supermarket in Chiang Mai, Thailand. The relevant chemicals in this commercial hair dye product include oxidative dye precursors, e.g., toluene-2,5-diamine, m-aminophenol, resorcinol, and N,N-bis(2-hydroxyethyl)-p-phenylenediamine sulfate, which are commonly used in dark hair dyes, particularly black or dark brown shades. Additional ingredients in the formulation that may act as irritants or sensitizers, potentially causing irritation or allergic reactions in some individuals, include ethanolamine (an alkalizing agent that opens the hair cuticle), ammonium hydroxide (used for pH adjustment), and hydrogen peroxide (an oxidizing agent).

2.3. Extraction of Natural Dyes

Various factors affecting natural dye extraction from L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves have been evaluated, including solvents, extraction methods, temperatures, and extraction durations. However, some factors have already been established based on previous studies, while milder conditions were evaluated to identify more sustainable methods, as shown in

Table 1. Extraction conditions for the dried powder of L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves.

Extraction Conditions	L. inermis Leaves	C. ternatea Flowers	I. tinctoria Leaves
Dried powder to solvent ratio (w/v)	1:20	1:20	1:20
Solvents	EtOH, water, and NaOH	Water	Water
Extraction methods	Orbital shaker and ultrasonication	Ultrasonication	Ultrasonication
Temperatures	RT and 50 °C	70 and 80 °C	50, 60, 70, and 80 °C
Extraction durations	30, 60, and 120 min	60 min	30 min

Note: EtOH = ethanol; NaOH = 1 M sodium hydroxide aqueous solution; RT = room temperature.

. The dried powders of each plant material were dispersed in the solvents and subjected to the orbital shaker (Innova^™2100 Eppendorf, Hamburg, Germany) or an ultrasonication bath (USC1040M, Digital Ultrasonic Cleaner, USC-D series, Bioevopeak Co., Ltd., Jinan, China), which had a capacity of 10 L, an ultrasonic power of 240 W, and operated at 40 kHz. After extraction, the resulting mixtures were filtered through Whatman No. 1 filter paper. The filtrates were then collected and pre-cooled to −20 °C on a 30 cm diameter freeze-drying tray and subjected to a freeze-dryer (Christ, Beta 2–8 LD-plus, Osterode am Harz, Germany), of which the condenser was pre-cooled to −40 °C for 30 min prior to the freeze-drying process. The frozen samples were placed in the drying chamber, and the vacuum pump was activated until the pressure reached 0.12 mbar. After 24 h of drying, the vacuum was gradually released until the pressure stabilized. The dried samples were collected and kept in a sealed aluminum bag protected from light until further use.

2.4. Characterization of Natural Dye Extracts by Ultraviolet–Visible (UV/VIS) Spectroscopy

The extracts of L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves were separately dissolved in DI water at 31.25 g/L and subsequently subjected to the UV/VIS spectrophotometer (UV-2450, Shimadzu, Duisburg, Germany) set at wavelengths ranging from 200 to 800 nm.

2.5. Development of Herbal Mixtures Containing L. inermis Leaves, C. ternatea Flowers, and I. tinctoria Leaves

A factorial design of experiments was employed to evaluate the herbal mixtures containing L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves. The purpose of the factorial design was to identify which extract had the most significant influence on dyeing performance, specifically in terms of high UV absorbance and low lightness (L*). The L* value indicated the brightness level of the sample, ranging from 0 (black) to 100 (white), and was measured using a colorimeter (Cortex Technology Aps, Plastvaeget 9, Hadsund, Denmark). A 3³ full factorial design using Minitab software (version 21.4.0, Minitab Inc., State College, PA, USA), comprising three factors each at three levels, was utilized to assess the factors and their interaction effects. The three factors included three types of natural dye extracts (L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves). The three levels of each factor were the maximum concentration of 33% w/v or 34% w/v, a mid-level concentration of 16.5% w/v or 17% w/v, and a low level of 0% w/v. This design resulted in 27 experimental runs (3³ = 27). However, the run with a 0:0:0 ratio was excluded, leaving a total of 26 runs for analysis. The data collected from these runs were then examined to determine the model’s coefficients, the normal probability plot, and the Pareto chart for UV absorbance and lightness. To minimize systematic errors, the order of the experiments was randomized. Data analysis was conducted using Minitab software (version 21.4.0, Minitab Inc., State College, PA, USA), allowing for the identification of the main effects and interactions among the factors.

2.6. Irritation Test by Hen’s Egg Chorioallantoic Membrane (HET-CAM) Test

The natural dye extracts and their selected combination were evaluated for the irritation potential using the HET-CAM test [17,18]. In brief, the fertilized hen eggs from the Pradu Hang Dam Chiang Mai breed, aged 7 to 9 days, were utilized in the test. Prior to the experiment, the hen eggshell was opened, and the CAM was moistened with a normal saline solution (9 g/L NaCl aqueous solution). Subsequently, 30 µL of the samples were applied to the CAM, and any signs of irritation on the vessels were observed under a stereo microscope (Carl Zeiss^™ Stemi 508 doc Stereomicroscope, Zeiss, Oberkochen, Germany) for 5 min at an ambient condition. Each individual natural dye extract, along with its combination, was evaluated in comparison with the chemical hair dye at the same concentration of 10 g/L. The observations at each time point were recorded in seconds and used to calculate the irritation score (IS) using the following equation:

IS = [(301−H) × 5]/300 + [(301−L) × 7]/300 + [(301−C) × 9]/300,

(1)

where H is the time point of the first observation of vascular hemorrhage, L is the time point of the first observation of vascular lysis, and C is the time point of the first observation of coagulation. Regarding the absence of hemorrhage, lysis, or coagulation within 5 min of the observation, the values were recorded as 0, indicating no irritation. This ensured that the absence of irritation signs did not artificially influence the IS. Observations for the calculation of the IS were conducted for a maximum of 5 min. The IS was classified into four categories, including no irritation (0.0–0.9), mild irritation (1.0–4.9), moderate irritation (5.0–8.9), and severe irritation (9.0–21.0). An aqueous solution of sodium lauryl sulfate (SLS) at a concentration of 10 g/L was used as a positive control, whereas a normal saline solution was used as a negative control. After the initial 5 min observation, the eggs were incubated again in the hatching chamber (Nanchang Howard Technology Co., Ltd., Nanchang, China) set at a temperature of 37.5 ± 0.5 °C and a humidity of 55 ± 7% for an additional 55 min. The signs of irritation were observed again under a stereo microscope (Carl Zeiss™ Stemi 508 doc Stereomicroscope, Zeiss, Oberkochen, Germany). The experiments were performed in triplicate.

2.7. Color Staining Performance Test

Human bleached hair tresses, obtained from a salon in Chiang Mai, Thailand, were from the random and blind collection process, so no specific donor information was identified. The hair was bleached by the salon using Lolane Pixxel Hair Bleaching Powder with 9% hydrogen peroxide (S.C. Sereechai Beauty Co., Ltd., Bangkok, Thailand). The obtained bleached hair was divided into small tresses, each 10 cm in length and weighing 1 g. Each hair tress was washed with approximately 1 g of shampoo, which was applied and massaged gently onto the hair for about 1 min. After rinsing with 100 mL of DI water until the shampoo foam was gone, the hair tress was gently patted dry with a towel and blow-dried using a hair dryer (HP8108/00, Philips Electronics (Thailand) Ltd., Bangkok, Thailand), followed by styling with a plastic comb. To prepare the natural dye and its mixture for the dyeing process, 5 g of freeze-dried extracts or their selected combination was dissolved in 5 mL of DI water to create a 50% w/v natural dye formulation. The bleached hair tresses were immersed in the formulation, and a brush was used to apply the natural dye formulation evenly over each hair tress. Afterward, the dyed hair tresses were placed in a Petri dish for various durations (5, 10, 15, or 30 min). Subsequently, the hair tresses were washed with shampoo and dried with a hair dryer as described previously. The color of the dyed hair tresses was measured in terms of lightness (L*) using a colorimeter (DSM II ColorMeter, Cortex Technology, Hadsund, Denmark).

2.8. Color Stability Test After Washing

The hair tresses dyed with 50% w/v natural dye extracts or their selected combinations were investigated for their color stability after washing. In brief, the hair tresses obtained from the previous color staining performance test were washed with shampoo and dried with a hair dryer. Subsequently, the color of the dyed hair tresses was measured in terms of lightness (L*) using a colorimeter (DSM II ColorMeter, Cortex Technology, Hadsund, Denmark). The stability of the hair color was investigated after washing up to 4 times.

2.9. Color Stability Test After Light Exposure

The hair tresses dyed with 50% w/v natural dye extracts or their selected combinations were investigated for color stability after light exposure. In brief, the hair tresses obtained from the previous color staining performance test were kept under conditions with or without light exposure and evaluated for color using a colorimeter (DSM II ColorMeter, Cortex Technology, Hadsund, Denmark) after 1, 3, 5, and 7 days. Under light exposure conditions, the hair tresses were placed in an environment with ambient natural light and temperature, which included indirect sunlight during the daytime through windows but excluded direct sunlight. In contrast, under light-protected conditions, the hair tresses were placed in a light-proof paper box, which was kept at ambient temperature.

2.10. Hair Morphology by SEM

The bleached hair tresses, along with those treated with natural hair dye extracts, including L. inermis, C. ternatea, I. tinctoria, and their mixture for 30 min, were evaluated for their morphology under SEM. Additionally, bleached hair tresses treated with chemical hair dye for 30 min were also evaluated. For natural hair dyes, the extract was mixed with DI water in a 1:1 ratio prior to the dyeing process. The DI water used in the current study had a conductivity not exceeding 1.3 µS/cm at 25 °C, which fell within the criteria of the US Pharmacopeia standard [15]. Since the conductivity of DI water serves as a key indicator of its purity, lower conductivity values reflect higher purity levels. For chemical hair dye, Garnier Color Naturals, shade 1, natural black (L’Oréal (Thailand) Ltd., Bangkok, Thailand) was used. Prior to the application of the chemical hair dye, the colorant was mixed with the developer in a 1:1 ratio, and after 30 min, it was rinsed off with water. After the hair tress was dried using a hair dryer, a hair fiber from each tress was randomly collected for SEM. In brief, the hair strands, each 1 cm in length, were adhered to a platform coated with carbon tape. Excess hair outside the carbon tape boundaries was trimmed. The samples were then introduced into a sputtering machine for coating with a 0.2 mm layer of gold (Au) using a current of 20 mA for 40 s, resulting in a coating thickness of 10 nm. The coating process was conducted under a pressure of 10⁻⁴ mbar using argon (Ar) gas. Following the coating, the hair samples were transferred to an SEM (Hitachi, SU3800 model, Tokyo, Japan) for imaging at magnifications of 800× and 2000×.

3. Results and Discussion

3.1. Dried Powder of Natural Dye Materials

The dried powder of L. inermis, C. ternatea, and I. tinctoria, as shown in Figure 1, exhibited particle sizes of approximately 305 ± 60, 168 ± 40, and 194 ± 69 µm, respectively, with moisture contents of 9.1 ± 0.1%, 6.5 ± 0.1%, and 5.2 ± 0.1%, respectively. The findings indicated that the moisture content of all plant materials in the present study fell within the standard criteria, as many medicinal plant materials are typically dried to a moisture content ranging between 10% and 20%, which is considered ideal for minimizing bacterial and fungal growth [19]. The results highlighted the lower moisture levels observed, which are beneficial in preventing the decomposition of bioactive components, either due to chemical changes or microbial contamination [19]. Excessive moisture can promote microbial growth, degrade active compounds, and reduce shelf life, emphasizing the need for proper drying and moisture control in herbal product processing [20]. Therefore, standardizing the moisture content of medicinal plant materials is essential for ensuring their quality, safety, and efficacy.

3.2. L. inermis Extracts

Various factors influencing the extraction of L. inermis leaves were evaluated, including solvents, extraction temperature, duration, and extraction methods. Different solvents had a distinct impact on the external appearance of the leaf extracts. The ethanolic extracts appeared greenish, the aqueous extracts yellowish, and the alkali extracts dark brown, as shown in Figure 2. The findings corresponded well with the previous research, indicating that dyeings produced with alkaline extracts of L. inermis leaves exhibited superior color strength compared to those obtained using DI water [21]. However, under alkaline conditions, the rate of autooxidation of certain chemicals is significantly enhanced, leading to the accelerated production of reactive oxygen species (ROS), which can interact with other compounds, potentially influencing the stability and intensity of the L. inermis stain. Aside from lawsone, L. inermis is rich in various bioactive compounds, including catecholic and aminophenolic derivatives, which are generally susceptible to autoxidation in aqueous solutions, and their autoxidation rates increase with increasing the pH value [22]. Therefore, the aqueous extracts were chosen over alkaline solutions to prevent the accelerated oxidation of bioactive compounds in L. inermis, ensuring the stability and effectiveness of its stain and beneficial properties. Additionally, the aqueous extracts would have a superior safety profile compared to alkaline solutions.

On the other hand, variations in extraction temperatures showed minimal significant effect visually, though differences were detectable in the visible light spectra. Notably, higher temperatures generally reduced the color intensity of the extracts, with the exception of the ethanolic extracts. The findings align with previous research, which indicated that higher extraction temperatures, particularly above 70 °C, reduce the color intensity of L. inermis extract due to the decreased stability of dye molecules at elevated temperatures [23]. Therefore, conducting extractions at room temperature is recommended to preserve the color intensity of the extracts while ensuring efficient extraction, particularly when using DI water, as higher temperatures tend to degrade the dye molecules.

The L. inermis leaves were hence further extracted using DI water as a solvent, with the assistance of a shaker and ultrasonication, for various extraction durations at room temperature. The appearance of the resulting L. inermis extracts, along with their visible light spectra shown in Figure 3, indicated that extraction durations influenced color intensity, with longer durations resulting in more intense colors. Extraction durations of 30 and 60 min were found to yield comparable color intensities, as indicated by similar visible light absorbance spectra. In contrast, an extraction duration of 120 min resulted in a significantly more intense color in both the extraction using an orbital shaker and ultrasonication. On the other hand, the L. inermis extracts obtained through ultrasonication exhibited higher visible light absorbance across all extraction periods when compared with the extraction using an orbital shaker. Therefore, it was evident that ultrasonication was more effective for extracting L. inermis leaves compared to the orbital shaker. Additionally, the extraction duration of 120 min was suggested.

3.3. C. ternatea Extracts

The C. ternatea extracts obtained using DI water and ultrasonication at 70 °C and 80 °C are presented in Figure 4. While no notable difference was observed in the visual appearance of the extracts, as their colors displayed similar intensities, their visible light spectra exhibited distinct differences. The extract obtained at 70 °C showed maximum absorbance at 540 nm, 573 nm, and 618 nm, whereas the extract derived at 80 °C exhibited peaks at 670 nm and 748 nm. Consequently, the extract obtained at 70 °C was selected for further investigations.

3.4. I. tinctoria Extracts

The extracts of I. tinctoria obtained from aqueous extraction with ultrasonication at temperatures ranging from 50 to 80 °C are presented in Figure 5. The extract colors varied significantly with extraction temperature, exhibiting a blue-green hue at lower temperatures, transitioning to greenish, and ultimately turning yellow as the temperature increased. The visible light spectra of each extract aligned with its visual appearance. Despite these color differences, the visible light spectra of all extracts were consistent, showing a maximum absorbance at around 664 nm. However, extraction at 50 °C is an energy-efficient approach that yields extracts with the most pronounced blue-green coloration, making it a practical and sustainable choice for further experiments.

3.5. Comparison of Selected Natural Dye Extracts

The natural hair dyes derived from the leaves of L. inermis, the flowers of C. ternatea, and the leaves of I. tinctoria exhibited distinct external appearances, as shown in Figure 6. The aqueous extract of L. inermis, obtained through ultrasonication for 120 min at room temperature, exhibited an orange color with maximum absorbance observed at 220 nm and 275 nm. The findings were in line with a previous study, which revealed that the L. inermis extract exhibited an orange color in visible light and a maximum absorption detected at 275 nm in the visible light spectrum [24]. Previous studies suggested that the peaks around 275 nm are attributed to the π-π* transition of lawsone [25]. Additionally, the phenolic compounds exhibited strong absorption at this wavelength [26]. However, although the L. inermis extract using alkali exhibited a darker color (Figure 2), the aqueous extract was selected for further investigation in this study due to its superior safety profile compared to the alkali. Additionally, higher temperatures negatively impacted the color of the extract, making room temperature preferable, while a longer duration of 120 min enhanced the color intensity of the L. inermis extract (Figure 3). These findings align with a previous study, which reported that the optimal extraction condition for the L. inermis aqueous extract was at the lowest investigated temperature [27]. As the temperature decreased, the concentration of lawsone (the dye in the leaves of L. inermis) in the extract increased, whereas higher temperatures, up to 90 °C, resulted in the instability and degradation of lawsone [27]. On the other hand, a previous study suggested an extraction time of 120 min, as extending the extraction time beyond 120 min resulted in a decrease in the lawsone content of the aqueous extract, and the lawsone compound was no longer detectable after an extraction period of 24 h [27].

On the other hand, the aqueous extract of C. ternatea flowers, obtained through ultrasonication for 60 min at 70 °C, exhibited a dark blue color with maximum absorbance observed at 265, 360, 575, and 622 nm. These findings were consistent with a previous study, which revealed that the absorbance around 265 nm in the C. ternatea flower aqueous extract indicates the presence of a benzene derivative, including anthocyanins [28,29]. In addition to the anthocyanins, the non-anthocyanin flavonoids were also detected as another band around 350 nm [30]. Additionally, the peak around 550–580 nm corresponded to the purple quinonoidal base, while the peak around 600–620 nm represented the blue anionic quinonoidal base [28]. In the case of C. ternatea flowers, an extraction period of 60 min was suggested to be sufficient for the ultrasound-assisted extraction method [31]. Jeyaraj et al. (2021) reported that a temperature of 50 °C for 60 min was found to be more efficient for the aqueous extraction method, as it required a shorter duration to achieve the optimum yield and total anthocyanin content [31]. Additionally, Jaafar et al. (2020) reported that no significant changes were observed in the total phenolic content or antioxidant activity during the 60 to 120 min of extraction [32]. Although some studies have indicated that temperatures above 70 °C are linked to a higher rate of anthocyanin degradation [33], the total anthocyanin content of C. ternatea flower extract was observed to increase with higher extraction temperatures (up to 90 °C) due to the elevated internal energy of pigment molecules, enhancing molecular diffusivity and solubility, as well as changes in plant cell membrane structure and permeability, which facilitated the extraction of anthocyanins [34]. The current study demonstrated that extracts obtained at 70 °C and 80 °C displayed similar visual appearances, but the visible light spectrum revealed differences that were attributed to the degradation of anthocyanins at a higher temperature (Figure 4).

Additionally, the aqueous extract of I. tinctoria leaves, obtained through ultrasonication for 30 min at 50 °C, exhibited a dark blue-green color with maximum absorbance observed at 290, 326, and 672 nm. The results corresponded well with a previous study, which reported that indigo blue in solution exhibits three characteristic absorption peaks at 286, 348, and 623 nm [35]. The dark blue-green color of the extract arises from indican, a colorless and water-soluble compound, hydrolyzed by beta-glucosidase, an enzyme in the leaves of I. tinctoria, into glucose and indoxyl, which is slightly yellow but undergoes dimerization and oxidation to form the strongly blue indigo dye [36]. However, at temperatures above 60 °C, the enzyme becomes denatured, significantly reducing indican hydrolysis and limiting indigo production [36]. The extracting temperature hence affected the color of the extracts, as shown in Figure 5. Additionally, higher temperatures, particularly in the presence of oxidants, led to the degradation of indigo blue, causing a gradual spectral shift as indigotin oxidizes to form dehydroindigo and then isatin [35]. This process shifts the colorimetric coordinates from bluish to greenish and ultimately to yellowish, as blue indigo is converted into yellow isatin [35]. Therefore, it is recommended that the extraction temperature be kept below 60 °C, and the extraction time be as short as possible [36]. Some previous studies have suggested extraction durations of 24–72 h at 25 °C, 4 h at 40 °C, and 30 min at 50 °C [36]. The I. tinctoria extract derived from the 30 min extraction at 50 °C was selected for further investigations due to its shortest extraction time within the degradation temperature threshold.

3.6. Herbal Mixtures Containing L. inermis Leaves, C. ternatea Flowers, and I. tinctoria Leaves

The natural dye from L. inermis, characterized by its orange-brown color, and the dye from I. tinctoria, characterized by its dark blue color, are traditionally combined to achieve the desired black hair color [10]. In the current study, a factorial design of experiments was used to assess the influence of each component in the herbal mixtures on dyeing performance, focusing on UV–visible light absorbance and lightness (L*), to develop the most effective combination of extracts from L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves. Although hair color is a crucial parameter, starting with the external appearance of the mixture (in terms of UV–visible light absorbance and L* values) would help ensure performance from the very beginning, leading to more accurate and effective results. Additionally, the appearance of the mixture served as an initial indicator, efficiently identifying and narrowing down the most promising combinations, potentially leading to more predictable outcomes when applied to the hair tresses. The coded values of the variables along with the responses are presented in

Table 2. Design matrix for the absorbance at λ_max and lightness of each natural dye mixture, along with their external appearances and wavelength of maximum absorbance (λ_max).

Mixture	L. inerms (%)	C. ternatea (%)	I. tinctoria (%)	Water (%)	Absorbance	Lightness	Appearance	λmax (nm)
1 (0, −1, +1)	17	0	33	50	0.80 ± 0.05 i	44.80 ± 0.02 v		291
2 (−1, 0, 0)	0	16.5	16.5	67	1.40 ± 0.01 f	37.20 ± 0.01 p		289
3 (−1, +1, 0)	0	33	16.5	50.5	1.51 ± 0.01 f	34.03 ± 0.02 m		290
4 (0, −1, −1)	17	0	0	83	0.38 ± 0.01 k	52.55 ± 0.02 z		289
5 (−1, 0, −1)	0	16.5	0	83.5	1.15 ± 0.01 g	41.85 ± 0.02 r		289
6 (+1, +1, +1)	34	33	33	0	2.41 ± 0.05 a	23.43 ± 0.02 b		290
7 (−1, +1, +1)	0	33	33	34	1.35 ± 0.04 f	35.15 ± 0.03 n		289
8 (+1, 0, −1)	34	16.5	0	49.5	1.70 ± 0.02 e	31.14 ± 0.02 j		290
9 (0, 0, −1)	17	16.5	0	66.5	1.84 ± 0.01 d	32.79 ± 0.02 l		291
10 (0, +1, 0)	17	33	16.5	33.5	2.20 ± 0.07 b	26.27 ± 0.02 e		289
11 (−1, −1, +1)	0	0	33	67	0.76 ± 0.10 i	46.44 ± 0.03 w		288
12 (−1, 0, +1)	0	16.5	33	50.5	1.32 ± 0.02 f	36.42 ± 0.02 o		289
13 (−1, −1, 0)	0	0	16.5	83.5	0.53 ± 0.01 j	47.20 ± 0.02 x		291
14 (+1, +1, −1)	34	33	0	33	1.70 ± 0.01 e	29.50 ± 0.01 i		291
15 (+1, +1, 0)	34	33	16.5	16.5	2.53 ± 0.03 a	20.04 ± 0.01 a		291
16 (+1, −1, −1)	34	0	0	66	0.51 ± 0.01 j	50.02 ± 0.02 y		290
17 (0, 0, +1)	17	16.5	33	33.5	2.01 ± 0.05 c	29.26 ± 0.02 h		291
18 (0, 0, 0)	17	16.5	16.5	50	2.01 ± 0.07 c	28.60 ± 0.01 g		291
19 (+1, −1, +1)	34	0	33	33	0.95 ± 0.03 h	43.03 ± 0.01 t		289
20 (+1, −1, 0)	34	0	16.5	49.5	1.04 ± 0.01 g,h	42.31 ± 0.02 s		291
21 (0, −1, 0)	17	0	16.5	66.5	1.03 ± 0.01 g,h	44.66 ± 0.02 u		289
22 (+1, 0, 0)	34	16.5	16.5	33	2.34 ± 0.04 a	24.30 ± 0.02 c		289
23 (0, +1, +1)	17	33	33	17	2.12 ± 0.02 b,c	28.23 ± 0.02 f		291
24 (−1, +1, −1)	0	33	0	67	1.37 ± 0.02 f	40.41 ± 0.02 q		291
25 (0, +1, −1)	17	33	0	50	1.70 ± 0.01 e	31.31 ± 0.02 k		289
26 (+1, 0, +1)	34	16.5	33	16.5	2.33 ± 0.05 a	25.08 ± 0.04 d		291

NOTE: The letter denotes significant differences among the absorbance at λ_max and lightness of each natural dye mixture (p < 0.05).

.

The wavelength of maximum absorbance (λ_max) for all the combinations was consistently detected between 288 and 291 nm but with varying intensities. The absorbance at their λ_max ranged from 0.38 ± 0.01 to 2.53 ± 0.03. Although some combinations exceed the linear range predicted by the Beer–Lambert law, a fundamental principle in spectroscopy that establishes a linear relationship between absorbance and concentration [37], the observed trends still provide meaningful insights despite the challenges posed by high concentrations. According to the full 3³ factorial design, the main factors and interaction effects on absorbance at λ_max are presented in Figure 7. The normal probability plot of the residuals, as shown in Figure 7a, indicates that the residuals are normally distributed since most of the data points fall closely along the reference line, with only slight deviations at the extreme ends. This pattern suggested that the assumption of normality for the residuals is reasonable and supported the validity of the statistical model used for the absorbance response. Higher absorbance represents a darker color because more light is absorbed by the mixture, indicating that the dye has a deeper or more intense color. Figure 7b shows the main effects of the concentration of the extract from L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves on the absorbance at λ_max. In the case where the largest vertical line moved from −1 to +1, it indicated the main effect of the factor on absorbance. This change reflects how the factor (e.g., concentration of an extract) influences the absorbance, with the magnitude of the change showing the strength of its impact. The presence of all extracts positively affected the absorbance of the herbal mixture, particularly the extract from C. ternatea flowers. However, increasing concentration from the mid-level (16.5 or 17% w/w) to the highest level (33 or 34% w/w) was found to have minimal effects. Although the impact was found to depend on concentration, the reduced effects at higher concentrations may be attributed to a saturation effect, wherein the natural dye extracts reached a plateau in absorbance, or the increased concentration of one extract could interfere with the absorption characteristics of the others, leading to a less pronounced increase in absorbance with further addition. In contrast to the other extracts, I. tinctoria leaf extract displayed an opposite trend, where increasing the concentration led to a decrease in absorbance. The likely explanation was that increasing the concentration of one component in the combination of three natural dye extracts necessarily reduced the relative concentrations of other ingredients to maintain the total at 100%. This dilution diminished the contributions of other components, especially when their functional effects are concentration dependent.

The interaction plots, as shown in Figure 7c, show parallel lines among the lowest (−1), middle (0), and highest (+1) level values, indicating that there is no significant interaction between the factors. This suggests that the effect of one factor on the response is independent of the levels of the other factor. However, it was observed that the mid and highest levels exhibited almost the same plot, while the lowest level showed a lower response. Therefore, the absence of any extract could significantly affect the absorbance of the mixture, emphasizing the importance of combining all three components. This highlights that each extract contributes to the overall dyeing performance, and omitting any one of them may lead to a noticeable reduction in absorbance.

The interaction plots in Figure 7c also demonstrate positive interactions between the combinations of each pair of extracts. A higher slope observed in their plot suggested that the combined effect of these two extracts on the response is greater than the individual effects of each extract, highlighting their synergistic interaction. In contrast to the others, the plot between C. ternatea flowers and I. tinctoria leaves shows a steep slope from the lower level to the mid-level but then plateaus from the mid-level to the higher level. The likely explanation is that once the concentration reaches a certain level (16.5% w/w), additional increases up to 33% w/w have little or no effect on the absorbance. This suggested that the response had reached a saturation point or maximum effect, where further increases in the concentration do not yield significant changes in the absorbance.

The absorbance at λ_max can be represented by the following model, which has an adjusted R² of 72.40%, indicating a good fit to the statistical model.

Absorbance = 0.424 + 0.00985 A + 0.0332 B + 0.0106 C + 0.000357 A × B +
0.000258 A × C − 0.000156 B × C,

(2)

where A, B, and C represent the effect of the extracts from L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves. The negative and positive signs of each coefficient reflect the effect of each independent variable on the absorbance, while the magnitude of the coefficient indicates the degree of significance of each independent variable. Therefore, absorbance increased with the concentration of each extract, as the main effects of all extracts are represented by positive coefficients. However, based on the varying absolute values of the coefficients, C. ternatea flower extract is suggested to have the greatest impact on absorbance due to the highest coefficient (0.0332), followed by I. tinctoria leaf extract (0.0106) and L. inermis leaf extract (0.00985), respectively. Regarding the interaction terms between pairs of extracts, the positive values suggest a synergistic effect that increases absorbance when L. inermis leaf extract was combined with either C. ternatea flower extract or I. tinctoria leaf extract. On the other hand, the negative value suggested that when C. ternatea flower extract and I. tinctoria leaf extract were combined, their effect on absorbance was less than the sum of their individual effects.

The order of significance based on magnitude was in good agreement with the resulting sequence from the Pareto chart, as shown in Figure 7d. The sequence of the significant terms and the main interaction effects with respect to the absorbance had the following order: C. ternatea flower extract > L. inermis leaf extract > I. tinctoria leaf extract > L. inermis leaf and C. ternatea flower extracts > L. inermis leaf and I. tinctoria leaf extracts > C. ternatea flower and I. tinctoria leaf extracts. The critical F-value was found to be 2.086, which represents the threshold above which a factor’s F-value is considered statistically significant. The C. ternatea flower, L. inermis leaf, and I. tinctoria leaf extracts, with F-values greater than 2.086, had a statistically significant impact on the absorbance of the mixture (p < 0.05). On the other hand, although the combination of L. inermis leaf extract with either C. ternatea flower extract or I. tinctoria leaf extract had a positive coefficient, the values were below the critical threshold, indicating that they were not statistically significant and had a smaller influence on the response.

The ANOVA results for absorbance at λmax, as shown in

Table 3. Analysis of variance (ANOVA) for absorbance at λ_max.

Source	DF	Sum of Squares	Mean Square	F-Value	p-Value
Regression	6	9.6346	1.60576	12.37	0.000 *
L. inermis	1	2.0808	2.08080	16.02	0.001 *
C. ternatea	1	6.5884	6.58845	50.74	0.000 *
I. tinctoria	1	0.7606	0.76056	5.86	0.025 *
L. inermis × C. ternatea	1	0.1200	0.12000	0.92	0.348
L. inermis × I. tinctoria	1	0.0631	0.06308	0.49	0.494
C. ternatea × I. tinctoria	1	0.0217	0.02167	0.17	0.687
Error	20	2.5971	0.12985
Total	26	12.2317

Note: DF: degree of freedom. An asterisk (*) denotes a significant effect on the absorbance at λ_max (p < 0.05).

, highlight that the concentration of individual natural dye extracts has a statistically significant effect on the color intensity of the herbal mixture, as measured by its absorbance. However, two-way interactions (i.e., combinations of two different dye extracts) were not found to be significant in the current study.

Apart from measuring absorbance at λ_max, the color intensity was also evaluated by lightness (L*). In contrast to absorbance, a lower L* value represents a darker color. The normal probability plot of the residuals for the lightness response, shown in Figure 8a, indicates a significant deviation from normality. The residuals do not align closely with the reference line, and there are clear gaps and clustering in the data, particularly around −5 and 5, as well as extreme outliers near −15 and 10. This pattern suggested that the residuals are not normally distributed, indicating a violation of a key assumption in regression analysis. Consequently, lightness may not be a reliable parameter for representing the color intensity of the natural dye combinations, unlike the absorbance.

Figure 8b shows the main effects of the concentration of the extract from L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves on lightness. Different patterns were observed in each herbal extract. In the case of L. inermis leaf extract, no effect was observed from the lowest (0% w/w) to the mid-level (17% w/w). However, increasing the concentration from the mid-level to the highest level (34% w/w) negatively affected the L* value, indicating a darker color. It can be summarized that a concentration of at least 17% w/w is required to observe this effect. On the other hand, the strong negative effects observed in C. ternatea flower extract, particularly from the lowest (0% w/w) to the mid-level (17% w/w), could be attributed to the fact that at low concentrations, the extract is not enough to noticeably darken the color. However, as the concentration increases, it causes a significant shift toward a darker hue. Once the concentration of C. ternatea flower extract reaches a certain level, the color intensity may approach a saturation threshold. In contrast to L. inermis leaves and C. ternatea flowers, I. tinctoria leaves exhibited a different trend, with almost no effect observed. A slight negative effect (darker color) was noted from the lowest concentration (0% w/w) to the mid-level (16.5% w/w). However, from the mid-level (16.5% w/w) to the highest concentration (33% w/w), the effect turned positive, resulting in a lighter color. The slight negative effect observed from 0% to 17% w/w is likely due to the gradual increase in pigmentation, which contributes to the desired darker color. However, beyond a certain point, adding more extracts may overpower the mixture, reducing the overall color intensity and leading to a paler color.

In contrast to L. inermis leaves and C. ternatea flowers, I. tinctoria leaves exhibited a different trend, with almost no effect observed. A slight negative effect was noted from the lowest (0% w/w) to the mid-level (16.5% w/w), but the effect turned positive from the mid-level (16.5% w/w) to the highest level (33% w/w). The slight negative effect (darker color) observed from 0% to 17% w/w is likely due to the gradual increase in pigmentation, which contributes to the desired darker color. However, adding more extract beyond a certain point may overpower the mixture, causing a reduction in overall color intensity and leading to a paler color.

The interaction plots, as shown in Figure 8c, indicate that the plots for the mid-level and highest level are parallel but cross the low-level plot, suggesting a non-linear interaction between the factors at different concentration levels. The effect of natural dye concentration is consistent across the mid and high levels, as both follow a similar trend. This indicates that increasing the concentration beyond the mid-level does not result in a significant change in the lightness. However, the crossing with the low-level plot suggested a significant difference between the low and higher levels. At the low level, the factor has a different effect compared to the higher levels, implying that the response at the low level is more sensitive or behaves differently. Therefore, the presence of all extracts is essential, as the low level (0% w/w) represents the absence of the extracts, which significantly impacts the color intensity in terms of lightness (L*). These findings were consistent with the absorbance results, highlighting that the absence of any extract can significantly affect the absorbance of the mixture, emphasizing the importance of combining all three components.

The lightness (L*) can be represented by the following model, which has an adjusted R² of 33.10%, indicating that lightness is a less reliable predictor of the natural dye extract compared to absorbance.

Lightness (L*) = 26.57 + 0.557 A + 0.284 B + 0.518 C − 0.02333 A × B − 0.01636 A × C − 0.01405 B × C,

(3)

where A, B, and C represent the effect of the extracts from L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves. It was noted that the main effects of the individual herbal extracts all positively influence lightness (L*), meaning that increasing the concentrations of any extract resulted in a lighter color. However, the interaction effects between pairs of extracts are all negative, indicating that when two extracts are combined, the overall lightness is somewhat reduced. Since the R² of the equation was only 33.10%, the order of significance based on magnitude did not align with the sequence shown in the Pareto chart (as shown in Figure 8d). Only C. ternatea flowers and their combination with L. inermis leaves were found to have a statistically significant impact on the lightness (L*) of the mixture (p < 0.05), with F-values greater than 2.086.

The ANOVA results for lightness, as shown in

Table 4. Analysis of variance (ANOVA) for lightness (L*).

Source	DF	Sum of Squares	Mean Square	F-Value	p-Value
Regression	6	1575.92	262.653	3.14	0.025 *
L. inermis	1	50.17	50.167	0.60	0.447
C. ternatea	1	583.00	582.997	6.98	0.016 *
I. tinctoria	1	0.34	0.339	0.00	0.950
L. inermis × C. ternatea	1	514.04	514.044	6.15	0.022 *
L. inermis × I. tinctoria	1	252.73	252.725	3.03	0.097
C. ternatea × I. tinctoria	1	175.64	175.644	2.10	0.163
Error	20	1670.89	83.544
Total	26	3246.80

Note: DF: degree of freedom. An asterisk (*) denotes a significant effect on the lightness (p < 0.05).

, highlight that only the concentration of C. ternatea has a statistically significant effect on the color intensity of the herbal mixture, as measured by the small value of lightness. These results were consistent with the Pareto chart.

The multiple response prediction that identified the optimal combination of extracts to achieve the desired absorbance and lightness is shown in Figure 9. The most suitable formulation consisted of 34% C. ternatea flower extract, 33% L. inermis leaf extract, and 16.5% I. tinctoria leaf extract, which corresponded to formulation 15. The experimental results confirmed this prediction, with formulation 15 showing the significantly highest absorbance of 2.53 ± 0.03 (p < 0.05), which closely matched the predicted value of 2.3330, falling within the 95% confidence interval (2.1530, 2.5130), as shown in

Table 5. Summary of the predicted and actual values of the absorbance and lightness of the natural dye mixture with standard error, confidence interval, and prediction interval.

Response	Fit	SE Fit	Actual Values	95% CI	95% PI
Absorbance	2.3330	0.0863	2.53 ± 0.03	(2.1530, 2.5130)	(1.9363, 2.7297)
Lightness	18.51	8.78	20.04 ± 0.01	(−1.73, 38.75)	(−12.98, 50.00)

Note: Fit: predicted values; SE Fit: standard error of the fit; 95% CI: 95% confidence interval; 95% PI: 95% prediction interval.

. A slight discrepancy could be due to minor experimental variability or other influencing factors. Similarly, regarding lightness (L*), the prediction suggested formulation 15 as the optimal combination for achieving the lowest lightness, with a value of 20.04 ± 0.01 (p < 0.05). The predicted lightness value of 18.51 also fell within the 95% confidence interval (−1.73, 38.75), confirming the model’s alignment with the experimental results. Overall, both the absorbance and lightness predictions were well supported by the experimental data, with only minor discrepancies, suggesting the model’s reliability.

Formulation 15 was identified through the design of experiment analysis as the optimal combination for achieving the most intense hair color with the highest absorbance and the lowest lightness (L*). The predicted values were in strong agreement with the experimental results, which showed formulation 15 with the highest absorbance of 2.53 ± 0.03 and the lowest L* value of 20.04 ± 0.01, making it the selected formulation for further evaluation.

3.7. Irritation Properties of Natual Extracts and Their Mixture (Formulation 15)

Safety is a critical concern in hair dyeing, leading to a growing interest in natural plant dyes over the past few decades as a safer alternative to synthetic dyes. Although natural dyes may not always be cost effective, they are considered environmentally safer due to their lower toxicity to humans and biodegradability [8]. In the current study, the irritation properties of extracts from L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves, along with their mixture, which exhibited the darkest color (formulation 15), were evaluated for their irritation properties using the hen’s egg test on chorioallantoic membrane (HET-CAM). The chemical hair dye at the same concentration as the herbal extracts was also evaluated in the HET-CAM test. While traditional animal testing has been a standard method to ensure safety, increasing concerns for animal welfare have led to legislative efforts to minimize animal pain and injury [38,39]. The HET-CAM test, which uses seven-day-old hen’s eggs in the early embryonic stage of development during the initial incubation phase, exempts the need for ethical committee approval and serves as an alternative approach that reduces reliance on animal testing while maintaining safety standards [40,41]. The irritation properties of each natural dye extract, their combination (formulation 15), and the chemical hair dye are shown in Figure 10. No signs of irritation were observed on the CAM when exposed to either the individual dyes or their combination, with results similar to those of the normal saline solution, which was used as a negative control. In contrast, the chemical hair dye induced severe irritation, similar to the aqueous solution of 1% w/v sodium lauryl sulfate (SLS), which was used as a positive control (

Table 6. Irritation score (IS) and irritation classification of L. inermis, C. ternatea, I. tinctoria, their mixture (formulation 15), and chemical hair dye.

Sample	IS	Classification
Positive control	13.42 ± 1.8 a	Severe irritation
Negative control	0.00 ± 0.0 b	No irritation
L. inermis extract	0.00 ± 0.0 b	No irritation
C. ternatea extract	0.00 ± 0.0 b	No irritation
I. tinctoria extract	0.00 ± 0.0 b	No irritation
Herbal mixture formulation 15	0.00 ± 0.0 b	No irritation
Chemical hair dye	15.63 ± 1.5 a	Severe irritation

Note: Positive control = 1% w/v of SLS aqueous solution, negative control = 0.9% w/v of NaCl solution. The IS was classified into four categories, including no irritation (0.0–0.9), mild irritation (1.0–4.9), moderate irritation (5.0–8.9), and severe irritation (9.0–21.0). The letter denotes significant differences among the irritation scores of each sample (p < 0.05).

). Although the irritation levels were similar, the signs of irritation were different. SLS, with an irritation score (IS) of 13.42 ± 1.8, induced vascular lysis, with bleeding from the blood vessels within 5 min of exposure. Coagulation was also observed within 5 min and became more severe with prolonged exposure. Hemorrhage was detected after 5 min, and some blood vessels disappeared. In contrast, the chemical hair dye, with severe irritation (IS of 15.63 ± 1.5), induced hemorrhage and coagulation within 5 min of exposure, which became more severe with prolonged exposure. Vascular lysis was also observed within 5 min of exposure, but most blood vessels remained intact. Therefore, the findings from this study conclusively demonstrate the superior safety profile of herbal extracts compared to chemical hair dyes, confirming the safety and feasibility of using natural dyes derived from L. inermis leaves, C. ternatea flowers, I. tinctoria leaves, and their combination as hair dyes. As the HET-CAM test is recognized as an alternative to the Draize eye test [17], the safety results from the HET-CAM assay support the potential use of these extracts on the scalp, where there is a possibility of contact with the eyes.

3.8. Color Staining Performance of Natual Extracts and Their Mixture

The bleached human hair tresses treated with the chemical hair dye and extracts from L. inermis leaves, C. ternatea flowers, and I. tinctoria leaves, along with their mixture (formulation 15), are shown in Figure 11, where distinct hair colors were observed. The chemical hair dye produced black hair, with no difference in the L* value representing similar darkness across different dyeing durations (Figure 11b). The extract from L. inermis leaves resulted in hair with an orange-brown color, which slightly darkened with longer treatment durations (Figure 11c). The results were consistent with the lightness of the hair tresses, as shown in Figure 11g, indicating that dyeing durations ranging from 5 to 30 min resulted in slightly lower L* values, reflecting greater lightness or lower intensity of darkness. On the other hand, C. ternatea extract was observed to have minimal impact on hair color, though the colors appeared more intense with a longer hair dying duration. The external appearance of the hair tress dyed with C. ternatea extract (Figure 11d) was consistent with its lightness in Figure 11g. Conversely, I. tinctoria was found to dye the hair tresses brown with a purple undertone at shorter treatment durations, which shifted to a blue undertone at longer durations (Figure 11e). Among these individual natural dye extracts, L. inermis leaf extract demonstrated the most potent dyeing properties, followed by I. tinctoria leaves and C. ternatea flowers, respectively. Interestingly, the combination of these three herbal extracts resulted in hair tresses with a darker, naturally dark brown color, as shown in Figure 11f. However, the treatment durations were also found to affect the darkness of the hair tresses. Although the color difference was barely noticeable visually, the colorimetric method revealed that a 30 min treatment with the herbal mixture (formulation 15) significantly darkened the color of the hair tresses, reducing the lightness, with an L* value of 8.3 ± 0.2 compared to L* values of 10.9 ± 0.1, 12.5 ± 0.2, and 17.5 ± 0.2 after treatment durations of 15, 10, and 5 min, respectively (p < 0.05). Therefore, a treatment duration of 30 min was recommended for further applications.

3.9. Color Stability After Washing of Hair Treated with Natual Extracts and Their Mixture

The color stability of hair tresses treated with the chemical hair dye, each individual natural dye extract, and their mixtures (formulation 15), as shown in Figure 12, was confirmed by the changes in hair lightness displayed in Figure 13. It was observed that the hair color remained durable with the application of the chemical hair dye, whereas the hair lightened progressively, indicating color fading, as the number of washes increased with the natural hair dyes. As presented in

Table 7. Color fading rate (L* per washing time) of hair tresses treated with chemical hair dye and 50% w/v natural extracts: L. inermis, C. ternatea, I. tinctoria, and their mixture (formulation 15) comprising L. inermis, C. ternatea, and I. tinctoria extracts with DI water in a 2:2:1:1 ratio before and after washing for up to 4 times.

Treatment Duration (min)	Color Fading Rate (L* Per Washing Time)
	Chemical Hair Dye	L. inermis Extract	C. ternatea Extract	I. tinctoria Extract	Herbal Mixture Formulation 15
5	0.03 ± 0.00 α (R2 = 0.964)	1.20 ± 0.03 a,γ (R2 = 0.914)	0.47 ± 0.02 a,ε (R2 = 0.880)	0.53 ± 0.01 c,δ (R2 = 0.986)	1.52 ± 0.01 a,β (R2 = 0.986)
10	0.02 ± 0.00 α (R2 = 0.956)	0.75 ± 0.04 b,δ (R2 = 0.911)	0.43 ± 0.09 a,ε (R2 = 0.968)	0.83 ± 0.00 a,γ (R2 = 0.979)	1.39 ± 0.01 b,β (R2 = 0.991)
15	0.01 ± 0.00 α (R2 = 0.983)	0.45 ± 0.04 c,ε (R2 = 0.985)	0.78 ± 0.01 b,γ (R2 = 0.950)	0.58 ± 0.01 b,δ (R2 = 0.949)	1.05 ± 0.02 c,β (R2 = 0.993)
30	0.01 ± 0.00 α (R2 = 0.818)	0.47 ± 0.03 c,δ (R2 = 0.817)	0.76 ± 0.01 b,β (R2 = 0.868)	0.57 ± 0.01 b,γ (R2 = 0.818)	0.50 ± 0.01 d,δ (R2 = 0.976)

Note: Different letters (a, b, c, and d) indicate statistically significant differences in the color fading rate among various treatment durations for each extract, as determined by one-way ANOVA followed by Tukey’s post hoc test (p < 0.05). Different symbols (α, β, γ, δ, and ε) denote statistically significant differences in the color fading rate among various extracts within the same treatment duration, as determined by one-way ANOVA followed by Tukey’s post hoc test (p < 0.05).

, the hair color fading rate was notably affected by the duration of treatment. However, different patterns were observed in each natural dye. For the individual L. inermis extract and the herbal mixture (formulation 15), the hair color demonstrated greater durability with longer treatment durations. The fading rate significantly decreased, indicating enhanced color retention over extended treatment times. In contrast, a lower fading rate was observed for treatments with C. ternatea extract at 5 and 10 min and I. tinctoria extract at 5 min, which could be attributed to the lower initial color intensity, reflected by higher L* values, resulting in smaller relative changes in color that were less perceptible in the fading rate.

3.10. Color Stability After Light Exposure of Hair Treated with Natual Extracts and Their Mixture

The color stability of the hair tresses after exposure to light or kept in the dark condition, as shown in Figure 14, was corroborated by the hair lightness (L*) shown in Figure 15. It was noticed that exposure to light led to the hair color fading over time, as shown in

Table 8. Color fading rate (L* per day) of hair tresses treated with chemical hair dye and 50% w/v natural extracts: L. inermis, C. ternatea, I. tinctoria, and their mixture (formulation 15) comprising L. inermis, C. ternatea, and I. tinctoria extracts with DI water in a 2:2:1:1 ratio before and after storage under light exposure and darkness for up to 7 days.

Samples	Color Fading Rate (L* Per Day)
	5 min		10 min		15 min		30 min
	Light	Dark	Light	Dark	Light	Dark	Light	Dark
Chemical hair dye	0.06 ± 0.01 a (R2 = 0.987)	0.04 ± 0.01 d,* (R2 = 0.991)	0.03 ± 0.01 b (R2 = 0.999)	0.02 ± 0.01 c,* (R2 = 0.966)	0.02 ± 0.01 c (R2 = 0.985)	0.02 ± 0.00 b (R2 = 0.912)	0.01 ± 0.00 d (R2 = 0.793)	0.01 ± 0.00 a (R2 = 0.992)
L. inermis extract	3.08 ± 0.01 a (R2 = 0.986)	−0.21 ± 0.01 d,* (R2 = 0.474)	2.39 ± 0.01 c (R2 = 0.965)	−0.14 ± 0.00 c,* (R2 = 0.281)	2.49 ± 0.01 b (R2 = 0.981)	0.19 ± 0.00 b,* (R2 = 0.909)	1.16 ± 0.86 d (R2 = 0.988)	0.40 ± 0.01 a,* (R2 = 0.875)
C. ternatea extract	1.00 ± 0.00 a (R2 = 0.929)	0.24 ± 0.00 a,* (R2 = 0.970)	1.01 ± 0.01 a (R2 = 0.949)	0.18 ± 0.01 c,* (R2 = 0.974)	0.81 ± 0.01 b (R2 = 0.847)	0.20 ± 0.01 b,* (R2 = 0.904)	0.30 ± 0.01 c (R2 = 1.000)	0.25 ± 0.01 a,* (R2 = 0.969)
I. tinctoria extract	0.98 ± 0.01 a (R2 = 0.964)	0.00 ± 0.00 c,* (R2 = 0.002)	0.68 ± 0.01 b (R2 = 0.999)	0.08 ± 0.01 b,* (R2 = 0.921)	0.44 ± 0.16 c (R2 = 0.971)	0.27 ± 0.01 a,* (R2 = 0.958)	0.27 ± 0.01 d (R2 = 0.939)	0.27 ± 0.01 a (R2 = 0.993)
Herbal mixture	0.53 ± 0.00 a (R2 = 0.964)	0.42 ± 0.01 a,* (R2 = 0.002)	0.47 ± 0.01 b (R2 = 0.999)	0.37 ± 0.00 b,* (R2 = 0.921)	0.36 ± 0.00 c (R2 = 0.971)	0.35 ± 0.01 c (R2 = 0.958)	0.28 ± 0.02 d (R2 = 0.939)	0.26 ± 0.02 d (R2 = 0.993)

Note: Different letters (a, b, c, and d) indicate statistically significant differences in the color fading rate among various treatment durations for each extract, as determined by one-way ANOVA followed by Tukey’s post hoc test (p < 0.05). An asterisk (*) denotes statistically significant differences in the color fading rate between light-protected conditions and exposure to light, as determined by a t-test (p < 0.05).

. The duration of both chemical and natural hair dye treatment significantly influenced color stability, especially when the hair tresses were exposed to light. Although the hair treated with the chemical hair dye was very dark, color fading was observed with short treatment durations of 10 min or less. The hair color from L. inermis was found to be the least stable when exposed to light, as significantly higher fading rates were observed over 7 days. While longer treatment durations slowed the fading rate, the 30 min treatment was still insufficient to prevent fading for L. inermis extract. Interestingly, storing the hair tresses dyed with the L. inermis extract in dark conditions effectively prevented color fading, with almost no fading detected. Therefore, it was evident that the color of L. inermis degraded when exposed to light. These findings supported previous studies, which have reported that the color of L. inermis undergoes significant degradation due to its inherent spectrum limitations and instability when exposed to UV light [42]. Similar to the L. inermis extract, the hair colors from individual C. ternatea or I. tinctoria extracts were unstable after being exposed to light, though with a lower rate of color fading. Interestingly, the combination of the three herbal extracts demonstrated enhanced stability when exposed to light, making it a more favorable formulation for achieving improved color retention and minimizing fading compared to the individual extracts.

3.11. Hair Morphology

The morphology of bleached hair fibers, along with those treated with natural hair dye extracts, their mixtures, and chemical hair dye, is shown in Figure 16. The hair fibers were found to be about 50–110 μm in diameter, which is consistent with previous reports [43]. The SEM micrographs depict the cuticle, which served as a barrier protecting the underlying cortex from external environmental damage. The shape and orientation of the cuticle cells are responsible for limiting friction between hair fibers, with smooth, tightly aligned, and flat cuticles representing healthy hair, as they protect the inner cortex from environmental damage and moisture loss [44]. The findings from this study obviously showed that natural hair dye extracts did not adversely affect hair quality, as the hair cuticles displayed a smooth appearance with tightly overlapping cuticle scales. In contrast, chemical hair dye damages the hair cuticles due to the alkaline and oxidant components present in the dye [45]. Although the damage caused by chemical hair dye is not visible to the naked eye, the hair often appears rough, dry, or crumbly [45]. SEM revealed the microstructural changes in the cuticle layer, showing the separation, breakage, and cracking of the cuticle surface scales, thereby emphasizing the extent of the damage caused by the dye’s chemical components. Additionally, the hair fibers treated with chemical hair dyes were found to have a larger diameter, with the cuticle appearing swollen, likely due to the highly alkaline nature of chemical hair dyes, which caused the cuticle to swell and allowed dye molecules to penetrate the cortex [46]. Therefore, natural hair dyes provide various benefits over chemical dyes, especially when it comes to maintaining hair health. In addition to being plant based and eco-friendly, which helps reduce environmental impact, natural dyes nourish and strengthen the hair, improving softness, moisture retention, and overall health [47]. While chemical hair dyes can lead to dryness, brittleness, and potential hair loss due to their harsh ingredients [46], natural hair dyes are a safer and more beneficial choice for maintaining healthy hair.

4. Conclusions

The present study demonstrated the potential of natural dyes derived from L. inermis, C. ternatea, and I. tinctoria as viable, sustainable alternatives to chemical hair dyes. All extracts were obtained using the green extraction method, with water as the solvent, which could offer a sustainable and eco-friendly alternative. Through the optimization of dye ratios using Minitab, the effective combinations yielding the most intense color, in terms of the highest absorbance and lowest lightness, were identified. The combination was found to be effective in hair dyeing, resulting in a long-lasting color that is more stable for washing and light exposure compared to each individual extract. Additionally, the extracts demonstrated favorable safety profiles, with no damage to the hair cuticle, positioning them as a promising alternative to synthetic hair dyes. The application of mixture formulation 15, comprising the extracts of L. inermis, C. ternatea, I. tinctoria, and water in a 2:2:1:1 ratio for 30 min, was recommended for optimal results. These findings highlighted the potential of herbal mixtures as a safer and more environmentally sustainable alternative for hair coloring, thereby contributing to the advancement of natural cosmetic products.