Maximization of the Production of a Low-Cost Biosurfactant for Application in the Treatment of Soils Contaminated with Hydrocarbons

Abstract

Oil spills occur during different operations in the energy sector, such as crude oil transport, tank filling and cleaning, and fueling. Such spills are one of the major causes of the accumulation of oil derivatives in the environment, requiring the remediation of soil and marine environments. The production of a biosurfactant by Starmerella bombicola ATCC 222214 was maximized by investigating the effect of different carbon/nitrogen sources and culture conditions. The mineral medium selected for its production was supplemented with 2.0% potato peel flour, 5.0% waste canola frying oil, and 0.20% urea. The culture conditions were a 200 rpm shaking speed, a fermentation time of 180 h, and a 4.0% inoculum size. The yield of isolated biosurfactant was 7.72 g/L. The emulsification rates of heavy oil and motor oil were 65.55 and 95.00%, respectively, indicating an affinity for complex hydrocarbons. In stability tests performed at different pH values, temperatures, and NaCl concentrations, the surface tension ranged from 27.14 to 31.08 mN/m. The critical micelle concentration was 2.0 g/L, at which the surface tension was 33.26 mN/m. The biosurfactant was composed of 6,6-dimethoxy-octanoic acid and azelaic acid, and it exhibited low toxicity to Brassica oleracea and Solanum lycopersicum. In the kinetic test, the biosurfactant allowed for the removal of 82.30%, 96.65%, and 98.25% of exhaust motor oil from sand, silty soil, and clay soil, while in the static test in packed columns, the removal yields were 66.62%, 63.03%, and 58.45%, respectively. The biosurfactant produced in this study is promising for environmental remediation applications in the energy sector.

1. Introduction

The oil industry is responsible for the release of oily pollutants into the environment during the extraction of crude oil from reserves or during its transport. The industry often uses chemical surfactants in cleanup processes, which are also toxic to the environment [1]. In soil, which is a complex system, the presence of surfactants increases the hydrophilicity of microbial cells and the adsorption of surface-active agents, increasing the hydrophobicity of total petroleum hydrocarbons. The reduction in the surface and interfacial tension promoted by surfactants increases the solubility, mobility, and bioavailability of petroleum hydrocarbons. This facilitates biodegradation by microorganisms, which can break down large oil slicks and make the contaminants more easily accessible for natural processes. As a result, microorganisms can metabolize and reduce the contaminants to safe levels or eliminate them. In water, they can reduce the formation of dense oil plates, properties often associated with oil spills, and the droplet size of oil-in-water emulsions through the acceleration of natural biodegradation processes [2,3,4,5,6,7]. The toxicity of chemical surfactants, however, has motivated the development of solutions to replace them with natural correlates, which are known as green surfactants [8].

Green surfactants produced by microorganisms—denominated biosurfactants—have attracted attention in recent decades as attractive, environmentally sustainable alternatives. Biosurfactants are amphipathic molecules with hydrophilic and hydrophobic portions that have detergency, emulsification, de-emulsification, foaming capacity, solubilization, and phase dispersion properties [9].

Biosurfactants are less toxic than their chemical counterparts, exhibit stability in a wide range of pH, temperature, and salinity values, and are biodegradable. This enables the application of these natural surface-active agents in microbial-enhanced oil recovery, the remediation of contaminated soil and water, and the removal of heavy metals. Biosurfactants can also be used in refineries to separate oil from oily sludge [1,10].

Examples of low-molar-mass biosurfactants include glycolipids such as rhamnolipids, sophorolipids, trehalose lipids, and mannosylerythritol lipids, lipoproteins such as surfactin, and fatty acids. High-molar-mass bioemulsifiers include polysaccharides, alasan, and emulsan [11].

The yeast Starmerella bombicola stands out among biosurfactant producers due to its potential to produce high yields of glycolipids, especially sophorolipids [12]. These compounds are biosurfactants belonging to the class of glycolipids, consisting of a disaccharide known as sophorose (2′-O-β-D-glucopyranosyl-1-β-D-glucopyranose) linked to a nonpolar fatty acid tail of 16 or 18 carbons linked by a β-glycosidic bond. Sophorolipids are secondary metabolites synthesized mainly by the non-pathogenic yeast S. bombicola [13]. These compounds are produced from sugars and lipids within the cell and are subsequently secreted into the extracellular environment as a mixture of compounds with related chemical structures [14]. These properties make sophorolipids an attractive choice for various industrial sectors, including the cosmetic and pharmaceutical ones [15,16].

The main technological challenge in the commercialization of biosurfactants is the high production cost, which is directly associated with the price of the cultivation media and purification processes. To address these limitations, several innovative strategies are being adopted. The use of low-cost industrial byproducts as alternative sources of nutrients in growing media is a promising approach to reducing costs. Furthermore, the application of statistical tools and optimization methods, such as the design of experiments, has been shown to be effective in maximizing production yields. The development of large-scale bioreactors is also crucial to increase the efficiency and economic forecasts of biosurfactant production. These strategies are contributing to making biosurfactant production more accessible and sustainable at an industrial level [17,18,19,20,21,22].

The aim of the present study was to produce a low-cost biosurfactant from the S. bombicola ATCC 222214 strain, characterize it, and assess its performance in the remediation of a terrestrial environment contaminated by petroleum derivatives. The results of this study are expected to be of great importance in improving the sustainability of traditional operations in the energy sector.

2. Materials and Methods

2.1. Substrates Used to Produce the Biosurfactant

The insoluble substrates assessed for biosurfactant production were waste soybean frying oil and waste canola frying oil, collected from local establishments in the city of Recife, Brazil, and commercial cottonseed oil obtained from a local supermarket in the same city. The soluble substrates were sugarcane bagasse ground in an industrial blender at 18,000 rpm for approximately 2 min to achieve its pulverization [23], a polymeric resin, sugarcane juice, and sugarcane molasses, which were donated by a sugar plant located in the city of Vitória de Santo Antão, Brazil.

Flours were made from peels of Irish potato (Solanum tuberosum L.) and sweet potato (Ipomoea batatas (L.) Lam.) obtained from the local market in the city of Recife, Brazil. The peels were dried in an oven at 70 °C for 24 h and ground in an industrial blender at 18,000 rpm for approximately 2 min to achieve a granulometry similar to powder, as described by Kartini et al. [24] for roselle anthocyanins. The flours were stored in closed recipients until their use in the production medium.

Corn steep liquor, a byproduct of corn production rich in vitamins, amino acids, and mineral salt, was donated by Corn Products S.A. do Brasil located in the city of Cabo de Santo Agostinho, Brazil. Analytical-grade glucose and table sugar (sucrose) were also obtained in the city of Recife and used as substrates.

2.2. Microorganism and Inoculum Preparation

The yeast Starmerella bombicola ATCC 222214 was obtained from the American Type Culture Collection (ATCC^®) and was used as the biosurfactant producer. The microorganism was maintained in test tubes containing yeast mold agar (YMA) formulated with 0.3% yeast extract, 1.0% D-glucose, 0.5% peptone, and 2.0% agar, at pH 7.0. To obtain the inoculum, the yeast was cultivated in yeast mold broth (YMB) for 24 h at 28 °C, 150 rpm, and pH 7.0. The inoculum was standardized by cell count in a Neubauer chamber until obtaining a cell density of 10⁶ cells/mL and was added to the medium at a concentration of 5.0%.

2.3. Influence of Substrates and Culture Conditions on Biosurfactant Production

The yeast Starmerella bombicola ATCC 222214 was grown in a mineral medium with the following composition: 0.5% KH₂PO₄; 0.3% NaNO₃; 0.005% CaCl₂·H₂O; 0.005% FeCl₃; 0.04% MgSO₄·7H₂O; and 0.005% yeast extract. Combinations of soluble and insoluble substrates were added to this medium.

The insoluble substrates (waste soybean frying oil, waste soybean canola oil, and commercial cottonseed oil) were added separately to the mineral production medium at a concentration of 5.0% (w/v). The following soluble substrates were added to prepare different culture media: glucose, sucrose, sugarcane molasses, sugarcane juice, sugarcane bagasse, Irish potato peel flour, and sweet potato peel flour.

After the preparation of the media, their pH values were adjusted to 6.0 ± 0.2 with the aid of a 5.0 M NaOH solution, and they were autoclaved at 121 °C for 20 min. Fermentations for biosurfactant production were performed in 500 mL Erlenmeyer flasks containing 250 mL of each medium incubated under orbital stirring at 200 rpm and 28 °C for 180 h with an inoculum of 5.0%. At the end of culturing, samples were removed to determine surface tension and yield of the isolated biosurfactant.

After selecting the best carbon sources, the influence of their concentration (1.5%, 2.0%, 3.0%, 4.0%, and 5.0%) on biosurfactant production was assessed. The flasks were maintained under the same conditions with orbital stirring at 200 rpm and 28 °C for 180 h.

Other substrates were then tested to select the best nitrogen source for biosurfactant production: 0.1% (w/v) yeast extract, 0.1% (w/v) urea, and 3.0% (w/v) corn steep liquor. In this step, the yeast extract used in the composition of the mineral medium was replaced with these nitrogen sources. When corn steep liquor was tested, the mineral medium was also replaced with distilled water. The flasks were kept under orbital stirring at 200 rpm and 28 °C for 180 h.

After establishing the best nitrogen source for biosurfactant production (urea), the influence of its concentration was tested at 0.10%, 0.12%, 0.20%, 0.40%, and 0.60% (w/v). The flasks were kept under orbital stirring at 200 rpm and 28 °C for 180 h. After the definition of the best culture medium for biosurfactant production, the optimal culture conditions were assessed using different values of the fermentation time (96, 120, and 150 h), stirring speed (150 and 180 rpm), and inoculum size (1.0%, 1.5%, 2.0%, 2.5%, and 3.0%). The steps followed to choose the best medium and culture conditions are shown in

Table 1. Sequence of experiments with substrates and cultivation conditions tested to produce the biosurfactant by Starmerella bombicola ATCC 222214.

(a) Selection of insoluble and soluble carbon sources at 5% concentration in the mineral medium (180 h at 200 rpm, 28 °C, and 5% inoculum)
Insoluble Soybean frying oil Canola frying oil Cotton oil	Soluble Glucose Sucrose Sugarcane juice Sugarcane bagasse Molasses Irish potato peel flour Sweet potato peel flour
(b) Influence of the concentration of soluble carbon source selected (1.5%, 2.0%, 3.0%, 4.0%, and 5.0%) in the mineral medium (180 h at 200 rpm, 28 °C, and 5% inoculum)
Irish potato peel flour
(c) Influence of the concentration of the insoluble carbon source selected (1.5%, 2.0%, 3.0%, 4.0%, and 5.0%) in the mineral medium (180 h at 200 rpm, 28 °C, and 5% inoculum)
Waste canola frying oil
(d) Selection of nitrogen source in the mineral medium (180 h at 200 rpm, 28 °C, and 5% inoculum)
Yeast extract (0.1%) Urea (0.1%) Corn steep liquor (3.0%) Corn steep liquor (3.0%) in deionized water
(e) Influence of the concentration of nitrogen source selected (0.1%, 0.12%, 0.2%, 0.4%, and 0.6%) in the mineral medium (180 h at 200 rpm, 28 °C, and 5% inoculum)
Urea
(f) Influence of the shaking speed for the culture medium selected (180 h at 28 °C and 5% inoculum)
150, 180, 200, and 250 rpm
(g) Influence of the shaking time for the culture medium selected (200 rpm at 28 °C and 5% inoculum)
96, 120, 150, and 180 h
(h) Influence of the inoculum size for the culture medium selected (180 h, 200 rpm at 28 °C)
1.0%, 2.0%, 3.0%, 4.0%, and 5.0%

.

2.4. Isolation of Biosurfactant

The biosurfactant was isolated following the method described by Farias et al. [25]. The metabolic broth (200 mL) was centrifuged at 4400 rpm and 4 °C for 15 min. The cell-free broth was then transferred to a decantation funnel, to which the same volume of ethyl acetate had been added (1:1, v/v), and the resulting solution was shaken vigorously for 15 min. After phase separation, the organic phase was recovered, and the extraction was repeated a second time with the same volume of solvent. After evaporation of the solvent of the organic phases at 40 °C, the residue containing the biosurfactant was washed twice with hexane to remove traces of hydrophobic substances and then evaporated. The yield of the isolated biosurfactant, which was determined by gravimetry taking into consideration the volume of the fermented medium sample used, allowed for an accurate assessment of the efficiency of the process.

2.5. Determination of Biosurfactant Surface-Active Properties and Stability

The surface tension of the biosurfactant was determined in the cell-free broth by the immersion of the du Noüy ring (KSV Instruments, Helsinki, Finland) and the measurement of the force required to pull it through the liquid into the air with the aid of a Sigma 70 tensiometer (KSV Instruments, Helsinki, Finland). The critical micelle concentration (CMC) was assessed by weighing 0.1 g of the isolated biosurfactant, with successive dilutions and respective surface tension measurements using the du Noüy ring.

The emulsification indexes for the hydrophobic substrates, namely, exhaust motor oil (SAE 20W-50, Chevron Brasil Lubricantes, Duque de Caxias, Brazil), paraffin-based lubricating oil (mixture of hydrocarbons and additives, 15 cSt, Petrobras, Brasilia, Brazil), and OCB1 heavy oil (complex mixture of hydrocarbons, 620 cSt, Petrobras), were determined following the method described by Cooper and Goldenberg [26]. A volume of 2 mL of substrate solution was added to 2 mL of the cell-free broth in test tubes, which were capped and agitated in a vortex (K45-2820, KASVI, São José dos Pinhais, Brazil) at 50 Hz for 2 min. The emulsions were then left to rest for 24 h at 27 °C, and the emulsification index (E₂₄) was calculated as a percentage using the following formula:

$$E_{24}(\%)={\displaystyle \frac{h_e}{h_t}}\times 100$$

(1)

where h_e is the emulsification height and h_t is the total height of the mixture, both expressed in cm.

The effects of different environmental conditions on the activity of the biosurfactant were investigated at different concentrations of NaCl (2.0%, 4.0%, 6.0%, 8.0%, and 10%), pH values (2.0, 4.0, 6.0, 8.0, 10.0, and 12.0), and temperatures (5, 70, 100, and 120 °C) for 60 min, as well as different heating times at 90 °C (10, 20, 30, 40, 50, 60, and 120 min). The biosurfactant stability was assessed by determining the surface tension [27].

2.6. Phytotoxicity Test

The phytotoxicity test was performed on seeds of cabbage (Brassica oleracea) and tomato (Solanum lycopersicum), as described by Tiquia et al. [28]. Petri dishes with Whatman N° 1 filter paper discs were sterilized, followed by the addition of 5 mL of the isolated biosurfactant at different concentrations, namely, half the critical micelle concentration (CMC), the CMC, and twice the CMC, or 5 mL of distilled water (control). The test was carried out in triplicate, and for each repetition, ten seeds were sterilized with NaClO and placed symmetrically on a filter paper. The dishes were then incubated in the dark for five days.

Germination of the seeds was determined, and the lengths of the roots (≥5 mm) were measured per biosurfactant solution.

The relative germination of the seed (G_r) was calculated as a percentage using the following equation:

$$G_r(\%)={\displaystyle \frac{N_b}{N_c}}\times 100$$

(2)

where N_b is the number of seeds germinated in the presence of the biosurfactant, and N_c is the number of seeds germinated in the control.

The relative root length (L_r) was calculated as a percentage according to the following equation:

$$L_r(\%)={\displaystyle \frac{L_b}{L_c}}\times 100$$

(3)

where L_b is the average root length in the presence of the biosurfactant and L_c is the average root length in the control.

Lastly, the germination index (I_G) was calculated as a percentage using the following formula:

$$I_G(\%)={\displaystyle \frac{N_b\times L_b}{N_c\times L_c}}\times 100$$

(4)

2.7. Chemical Composition of Biosurfactant

The crude biosurfactant (10.0 g) was redissolved in approximately 300 mL of hexane and extracted three times with 100 mL of an 80% ethanol solution. The alcoholic extracts were combined, and the hexane phase was discarded. The alcoholic solution containing the biosurfactant was concentrated in an oven at 50 °C. When the volume reached approximately 50 mL, the material was redissolved in 150 mL of ethyl acetate, the resulting mixture was transferred to a separating funnel, and the organic phase was washed twice with distilled water. After drying ethyl acetate with about 20.0 g of anhydrous sodium sulphate, filtration with Whatman No. 1 filter paper and evaporation at 50 °C allowed for the obtainment of the purified biosurfactant.

The purified biosurfactant was dissolved in deuterated chloroform (CDCl₃) and analyzed in a 500 MHz spectrometer (Agilent, Santa Clara, CA, USA) operating at 300.13 MHz to determine chemical shifts (δ) in the ¹H NMR and ¹³C NMR spectra on the ppm scale relative to tetramethylsilane.

The purification described above was also necessary to dissolve the biosurfactant in methanol to enable analysis in an FT-IR spectrometer (Bruker IFS66, Karlsruhe, Germany) in the spectral region range from 4000 to 400 cm⁻¹. Precision was maintained in the wavenumber range from −0.1 to +0.1 cm⁻¹ [29].

The hydrophobic portion of the biosurfactant (fatty acid) was submitted to analysis in a gas chromatograph and mass spectrometer system (Thermo Scientific Trace 1300—ISQ Single Quadrupole, Waltham, MA, USA) with a TGMS-5 column (30 m × 0.25 mm; film thickness: 0.25 μm). The initial column temperature was 60 °C for 3 min, raised at 10 °C × min⁻¹ until reaching 300 °C, and maintained for 15 min. A volume of 10 µL of sample was injected using helium as the carrier gas. Injector and detector temperatures were 300 and 280 °C, respectively.

2.8. Kinetic Tests of Biosurfactant Removal of Exhaust Motor Oil Adsorbed to Sand and Soils

Samples of standard sand (NBR 7214, 2015) as well as clay and silty soils collected from the city of Jaboatão, Brazil, were contaminated with 10% exhaust motor oil, to which 20 mL of the isolated biosurfactant solutions in distilled water were added at concentrations equal to ½ × CMC (1 g/L), 1 × CMC (2 g/L), and 2 × CMC (4 g/L). The flasks were maintained in an orbital shaker (Tecnal, Piracicaba, Brazil) at 150 rpm and 28 °C for 24 h. After treatment with hexane and the evaporation of the solvent, the mass of the remaining oil was quantified [30].

2.9. Static Test of Biosurfactant Removal of Exhaust Motor Oil Adsorbed to Sand and Soils in Packed Columns

The removal of motor oil by the biosurfactant produced by S. bombicola ATCC 222214 was assessed using three different types of soil (standard sand, clay soil, and silty soil), based on Robert et al. [31]. Glass columns (57 cm in height and 4.0 cm in diameter) were filled with a mixture of 200 g of soil containing 10% (w/w) exhaust motor oil (15 cSt). The surface of each column was inundated with 200 mL of the solutions for the flushing of the hydrophobic contaminant (motor oil) adsorbed to the soils or sand. The tested solutions were deionized water (control), the crude biosurfactant, and solutions of the isolated biosurfactant at ½ × CMC (1 g/L), 1 × CMC (2 g/L), and 2 × CMC (4 g/L). The synthetic surfactant Tween 80 was used for comparison purposes at concentrations of the CMC (15 mg/L) and twice the CMC. After percolation of the solution in the soils or sand contained in columns (monitoring for 24 h), the samples were washed with 200 mL of hexane for the removal of the residual oil. The solvent was evaporated at 50 °C, and the quantity of oil extracted was determined by gravimetry [32].

2.10. Statistical Analysis

The data were submitted to statistical analysis using Statistica^® (version 7.0), followed by analysis of variance (ANOVA). All experiments were performed in triplicates, and the results (n = 3) were expressed as means ± standard deviations (SDs), with a confidence interval of 95% corresponding to a significance level of 5% (p < 0.05).

3. Results and Discussion

3.1. Influence of Nutrients and Culture Conditions on Biosurfactant Production

The analysis of soluble and insoluble carbon sources for biosurfactant production by Starmerella bombicola ATCC 222214 revealed similar surface tensions. When mixing the waste soybean and canola frying oils with any soluble carbon source, the surface tension was near 32.50 mN/m, while the cottonseed oil did not give satisfactory results when mixed with sugarcane bagasse and molasses, as the surface tension was higher than 39.90 mN/m (

Table 2. Surface tension (mN/m) of biosurfactant produced by Starmerella bombicola ATCC 222214 grown (a) on different soluble (5.0%) and (a) insoluble carbon sources (5.0%); (b) on Irish potato peel flour at different concentrations; (c) on waste canola frying oil at different concentrations and potato peel flour (2.0%); (d) on different nitrogen sources (0.1%) and corn steep liquor (3.0%); (e) on urea as a nitrogen source at different concentrations; (f) at different shaking speeds; (g) for different shaking times; and (h) using different inoculum sizes. Conditions (a–e) were assessed for 180 h at 200 rpm using a 5.0% inoculum, whereas only one of these variables was altered in conditions (f–h). Results are expressed as means ± SDs (n = 3), where means are significant at p < 0.05.

(a) Insoluble carbon source		(a) Soluble carbon source
		Glucose	Sucrose	Sugarcane juice	Sugarcane bagasse	Molasses	Irish potato peel flour	Sweet potato peel flour
Soybean frying oil		33.30 ± 0.03	32.90 ± 0.01	32.06 ± 0.04	32.06 ± 0.01	32.59 ± 0.03	32.02 ± 0.01	32.15 ± 0.01
Canola frying oil		32.87 ± 0.01	33.05 ± 0.02	32.05 ± 0.01	32.60 ± 0.02	32.77 ± 0.07	29.60 ± 0.01	31.19 ± 0.01
Cotton oil		32.87 ± 0.01	37.23 ± 0.03	37.81 ± 0.01	44.81 ± 0.03	39.91 ± 0.02	31.77 ± 0.06	30.05 ± 0.01
(b) Irish potato peel flour concentration (%)					(e) Urea concentration (%)
1.5	2.0	3.0	4.0	5.0	0.10	0.12	0.20	0.40	0.60
31.84 ± 0.05	31.57 ± 0.06	31.60 ± 0.07	31.56 ± 0.08	31.50 ± 0.09	31.22 ± 0.05	32.28 ± 0.06	30.61 ± 0.07	38.60 ± 0.09	39.55 ± 0.08
(c) Canola frying oil concentration (%)					(f) Shaking speed (rpm)
1.5	2.0	3.0	4.0	5.0	150	180	200	250
33.45 ± 0.06	33.47 ± 0.07	33.68 ± 0.08	33.59 ± 0.09	31.57 ± 0.05	41.04 ± 0.09	40.41 ± 0.07	30.61 ± 0.06	41.50 ± 0.08
(d) Nitrogen source					(g) Shaking time (h)
Yeast extract	Urea		Corn steep liquor in mineral medium	Corn steep liquor in deionized water	96	120	150	180
32.56 ± 0.09	31.54 ± 0.08		32.66 ± 0.07	33.17 ± 0.08	33.23 ± 0.06	35.46 ± 0.08	35.27 ± 0.09	30.61 ± 0.07
(h) Inoculum size (%)
1.0	2.0		3.0		4.0		5.0
37.64 ± 0.07	34.12 ± 0.05		33.25 ± 0.06		31.16 ± 0.08		33.00 ± 0.09

).

Irish potato peel flour and waste canola frying oil were selected as the most promising soluble and insoluble carbon sources, respectively, as their mixture led to a surface tension as low as 29.60 mN/m, and because both sources are easily accessible and economically viable. The best soluble carbon source was selected taking into account the most economically viable concentration without losing the biosurfactant production efficiency (Table 2 (a)).

As shown in Table 2, all concentrations yielded good results in terms of the surface tension, which ranged between 31 and 32 mN/m for the different mixtures of potato peel flour with waste canola frying oil (Table 2 (b)). Although the 5.0% concentration ensured a slightly lower surface tension, the 2.0% concentration (Table 2 (b)) was selected as the most promising due to the lower quantity. Moreover, as potato peel flour is very rich in starch, a higher percentage would have considerably increased the medium viscosity, thereby hampering the mixture of substrates and the production of the biosurfactant and increasing the costs.

In the experiments performed at concentrations of waste canola frying oil ranging from 1.5 to 5.0% (Table 2 (c)), the lowest surface tension of the biosurfactant (31.57 mN/m) was obtained at the highest concentration (5.0%). This result demonstrates the high affinity of S. bombicola ATCC 222214 for this oil, which is important for the production of the hydrophobic portion of the biosurfactant.

Although the use of different nitrogen sources led to similar surface tensions, the lowest value was achieved using urea (31.54 mN/m) (Table 2 (d)). As is well known, this nitrogen source is widely available on the market, is inexpensive, and has a high degree of purity, which reduces the amount of suspended particles in the medium. In contrast, corn steep liquor contains a lot of solid substances in variable proportions, which causes complications from the industrial standpoint when considering the scaleup and purification process.

Considering the importance of meeting all the microbial nutritional needs in the most economical manner, different concentrations of urea were assessed to establish the minimum concentration of the nitrogen source without compromising the biosurfactant production. As shown in Table 2 (e), the 0.20% concentration, ensuring a surface tension of 30.61 mN/m, was the most promising considering both the production efficiency and cost.

With regard to the physical conditions, the best shaking speed for the production of the S. bombicola ATCC 222214 biosurfactant was 200 rpm, resulting in a surface tension of 30.61 mN/m, which is even 10 mM/m lower than that under the other shaking conditions (Table 2 (f)). This can be explained by the fact that gentler shaking results in less aeration and a smaller contact area offered for the yeast to grow, consume nutrients, and produce the biosurfactant (Table 2). For some species, however, too-high agitation rates have been shown to affect the interactivity of the microorganisms by offering very short contact with nutrients, thus hindering the biosurfactant production due to the high shear stress and consequent limited heat and mass transfer [33].

The biosurfactant surface tension progressively increased by increasing the shaking time from 96 to 120 h, and it suddenly dropped to the minimum value of 30.61 mN/m after 180 h (Table 2 (g)). This result indicates that most of the yeast cells encountered stressful conditions after this time, likely due to either the depletion of some nutrient or accumulated shear stress, stimulating the production of the biosurfactant as a secondary metabolite.

The results obtained using different inoculum sizes revealed that the biosurfactant production was enhanced by increasing the initial cell concentration up to 4.0%, reaching a minimum surface tension of 31.16 mN/m (Table 2 (h)), while the increased value of this parameter at the highest concentration (5.0%) suggests the possible excess viscosity of the culture.

Based on all these results taken as a whole, the biosurfactant produced in a mineral medium containing 5.0% waste canola frying oil, 2.0% potato peel flour, and 0.20% urea for 180 h at 200 rpm with a 4.0% inoculum was selected for the characterization step and environmental applications.

Alyousif et al. [34], after investigating different carbon and nitrogen sources and their concentrations, inoculum sizes, and incubation periods to optimize the production of a Staphylococcus epidermidis biosurfactant in a minimal salt medium, assessed its performance, determining the emulsification activity, emulsification index, and oil dispersion. The optimal biosurfactant production was achieved after 5 days in a medium containing 2% olive oil as the carbon source and 0.2% glutamic acid as the nitrogen source, using a 3% inoculum size.

In the present study, S. bombicola ATCC 222214, which is a yeast and, therefore, a more complex organism, needed not only an insoluble carbon source (waste canola frying oil) but also a soluble carbon source (potato peel flour) and a longer period (approximately 8 days) to effectively produce the biosurfactant. In any way, the process is still more economical and sustainable due to the use of low-cost, renewable waste materials that would have otherwise been discarded.

Shah et al. [35] investigated three hydrophobic substrates, Tapis oil, Melita oil, and Ratawi oil, for the production of a biosurfactant by S. bombicola ATCC 22214, which achieved surface tensions of 36.38, 37.84, and 38.92 mN/m, respectively. These values were comparable with the biosurfactant produced using palm oil (35.38 mN/m).

Maddikeri et al. [36] reported the production of sophorolipids by S. bombicola in a medium containing waste cooking oil and studied the effects of the stirring speed, the presence of ultrasound, and the mode of operation (batch or fed-batch). The surface tension (32.60 mN/m) obtained in the best conditions (i.e., fed-batch mode assisted by ultrasound) was only slightly higher than the value found in the present study (30.61 mN/m). Biodiesel has also been described as an interesting hydrophobic feedstock for the microbial synthesis of sophorolipids from S. bombicola, which reduced the surface tension to 34.2 mN/m [37].

3.2. Determination of Critical Micelle Concentration

The yield of isolated biosurfactant produced by S. bombicola ATCC 222214 was 7.72 g/L. The behavior of the surface tension as a function of its concentration (Figure 1) demonstrates that the critical micelle concentration (CMC) was about 2.0 g/L with a surface tension of 33.26 mN/m. After reaching this threshold value, the surface tension remained practically unaltered as the formation of micelles occurred.

3.3. Emulsification Index

To assess the emulsification activity, an analysis was performed on the emulsions formed by the biosurfactant from S. bombicola ATCC 222214. An excellent emulsification index was found for exhaust motor oil (95.43%) with the biosurfactant. This value is higher than that reported by Rocha Junior et al. [38] for the biosurfactant of the yeast Candida tropicalis, which emulsified 92% of exhaust motor oil. The emulsification index of the biosurfactant and OCB1 heavy oil mixture (65.55%) also appears rather promising considering the high viscosity of this oil, which hampers the dissolution of substances with different densities. Alfian et al. [39] described the production of sophorolipids from S. riodocensis GT-SL1R using the co-carbon substrates glucose and palm oil in shake flasks and produced biosurfactants with an emulsification activity of 54.59% against kerosene compared to the S. bombicola BCC5426 strain with an activity of 60.22%.

3.4. Stability of Biosurfactant under Different Environmental Conditions

The biosurfactant produced by S. bombicola ATCC 222214 was tested under different environmental conditions to simulate applications with extreme conditions.

The salt (NaCl) concentration did not significantly affect the biosurfactant, which proved to be highly stable under high-salinity conditions. The surface tension remained between 28.83 and 28.97 mN/m (Figure 2a), indicating that the biosurfactant could be used in the marine environment. These surface tension values are lower than those reported by Ashish and Debnath (Das) [40] for a biosurfactant produced by C. tropicalis, which ranged from 42 to 49 mN/m.

As for the effect of the pH, Figure 2b shows a slight increase in the surface tension under acidic (30.94 mN/m) and extremely alkaline (32.28 mN/m) conditions, with the best result (26.41 mN/m) obtained at pH 8. Souza et al. [41], who assessed the stability of a natural surfactant produced by the yeast Wickerhamomyces anomalus under the same pH values, reported a higher surface tension range (30.5–34.7 mN/m). These results demonstrate that the S. bombicola ATCC 222214 biosurfactant was effective within the pH range tested, with the exception of pH 12, in the interaction with water and possibly with oil.

A progressive reduction in the surface tension occurred after exposing the samples for 60 min to increasing temperatures (Figure 2c), which could be explained by the fact that the molecules become agitated and interacted more in the medium as the temperature increased. Particularly, the surface tension ranged from 27.14 to 30.29 mN/m, with the lowest value found at 120 °C. Somosa-Coutiño et al. [42], who investigated the production of a biosurfactant by Pseudomonas aeruginosa B0406, also found a reduction in the surface tension to 30 mN/m as the temperature was progressively increased from −20 to 120 °C, similar to what occurred in the present study.

The analysis of the biosurfactant stability to exposition to 90 °C for different times revealed a slight reduction in the surface tension over time, with values ranging from 28.79 to 29.94 mN/m (Figure 2d).

The values of the surface tension of the biosurfactant from S. bombicola ATCC 222214 in the presence of different concentrations of NaCl, pH values, temperatures, and times of heating at 90 °C can be found in Table S1 in the Supplementary Material.

3.5. Phytotoxicity Tests

In the phytotoxicity test on Brassica oleracea (cabbage), the germination and root growth increased with the increase in the concentration of the biosurfactant up to its CMC, suggesting that the plant used the biosurfactant as a nutritional source for growth. Overall, the biosurfactant proved nontoxic to cabbage, with a germination index (I_G) above 91% at the CMC and twice the CMC (

Table 3. Germination indexes (%) of Brassica oleracea and Solanum lycopersicum in the presence of different concentrations of the biosurfactant produced by Starmerella bombicola ATCC 222214. Results are expressed as means ± SDs (n = 3), where means are significant at p < 0.05.

Concentration of Biosurfactant	Brassica oleracea (Cabbage)	Solanum lycopersicum (Tomato)
½ × CMC	56.9 ± 0.7	79.3 ± 0.9
1 × CMC	91.1 ± 0.7	34.3 ± 0.6
2 × CMC	91.3 ± 0.8	29.5 ± 0.6

). Surprisingly, when the biosurfactant was used at ½ × CMC, the I_G was lower (56.9 ± 0.7%), possibly because such a biosurfactant concentration is insufficient to stimulate germination.

The phytotoxicity test on Solanum lycopersicum (tomato) demonstrated that the root growth remained above 50% even with the increasing biosurfactant concentration. However, both the germination and root growth decreased with increasing biosurfactant concentration, suggesting that tomato is more sensitive than cabbage to the biosurfactant. Nonetheless, the I_G was 79.3% (Table 3) and the relative root length was 99.99% when the biosurfactant was tested at half the CMC, demonstrating the low toxicity of the biosurfactant to tomato seeds. Pinto et al. [43], who investigated the toxicity of a biosurfactant produced by C. bombicola in a medium containing molasses, corn steep liquor, and waste soybean frying oil also using tomato as the bioindicator, reported comparable I_G values at half the CMC (64%) and the CMC (45%), but very strong toxicity (0%) at twice the CMC.

3.6. Nuclear Magnetic Resonance (NMR) Spectroscopy

The ¹H NMR spectrum suggested the presence of methyl and aliphatic groups between δ 0 and 3 ppm in the purified biosurfactant. The signal of a hydroxyl group bonded to carbon was found in the region between 3 and 5 ppm, and the presence of a double bond was detected between 5 and 6 ppm. The signal between 6 and 7.5 ppm is related to the residue of the chloroform used as a solvent (Figure 3A). The ¹³C NMR spectrum showed typical signals of aliphatic carbon between δ 0 and 45 ppm and of a carbonyl group between 50 and 74 ppm, while signals in the region between 75 and 85 ppm are ascribable to chloroform. Moreover, the peaks between 120 and 150 ppm point to the presence of a carbon double bond, and the ones between 165 and 190 ppm point to that of a carboxylic group (Figure 3B). These results taken as a whole suggest that the biosurfactant from S. bombicola ATCC 222214 contains an unsaturated fatty acid with a hydroxyl group. Finally, since the ¹H NMR spectrum also showed hydrogens bonded to aliphatic carbons in the 0–2 ppm range, it is reasonable to suppose a structure similar to that of azelaic acid [44].

3.7. Fourier Transform Infrared (FTIR) Spectroscopy

FTIR spectroscopy revealed the presence of a hydroxyl group in the broadening near 3500 cm⁻¹, peaks between 3000 and 2800 cm⁻¹ due to aliphatic carbon, and a peak at 1737.85 cm⁻¹ corresponding to carbonyl elongation (Figure 4). Likewise, the glycolipid produced by the Pseudomonas sp. KZ1 strain investigated by Zdarta et al. [45] showed the typical absorption bands of aliphatic carbon as well as hydroxyl and carbonyl groups in the same regions, corresponding to the fatty acid portion.

3.8. Gas Chromatography–Mass Spectrometry (GC-MS)

The GC-MS chromatogram of the purified biosurfactant exhibited five main peaks, two corresponding to acids commonly produced by microorganisms and then probably also by S. bombicola ATCC 222214 (i.e., 6,6-dimethoxy-octanoic acid [46] and nonanedioic acid (azelaic acid) [47]), and the other three corresponding to plant compounds still present in the sample after purification (i.e., palmitic, linolelaidic, and α-linolenic acids) (according to the data library).

The relative abundance of 6,6-dimethoxy-octanoic acid (retention time [RT]: 14.775 min) was 7,174,873,000 with a 7.144% peak area, that of azelaic acid (RT: 15.202 min) was 9,400,243,000 with a 9.309% peak area, that of palmitic acid (RT: 15.841 min) was 14,671,072,000 with a 45.108% peak area, and those of linolelaidic acid (RT: 18.923 min) and α-linolenic acid (RT: 19.615 min) were 17,449,328,000 and 5,195,392,000 with 32.240% and 6.199% peak areas, respectively (Figure 5). Azelaic acid produced by S. bombicola ATCC 222214 from waste canola frying oil was also identified by Anwar and Wahyuningsih [48], but after chemical synthesis via the oxidation of ricin (castor) oil. This biosurfactant was also detected in a pure cosmetic assessed by Yang et al. [49] using H₂SO₄ and methanol for the esterification, at an RT of 12.158 min.

3.9. Removal of Exhaust Motor Oil Adsorbed to Sand and Soils

The removal of exhaust motor oil from standard sand in the kinetic test increased up to a yield of 82.30 ± 0.07% by adding the isolated biosurfactant until the CMC, while no further increase was found when using the biosurfactant at twice the CMC (Figure 6). When silty soil was tested as an adsorbent, the use of the biosurfactant at only half the CMC allowed up to 94.20 ± 0.07% of this polluting oil to be removed, while an increase in the concentration up to twice the CMC did not have any significant effect, ensuring a removal yield of 96.65 ± 0.09%. Similar behavior was found when clay soil was used, but the removal yields were even higher at both ½ × CMC (98.25 ± 0.05%) and the CMC (98.25 ± 0.08%).

The motor oil removal yield observed in the kinetic test with standard sand using the S. bombicola ATCC 222214 biosurfactant at half the CMC (79.05 ± 0.09%) was higher than that reported for a biosurfactant from Bacillus cereus UCP 1615 (about 50%) [50], while those found at the CMC (82.30 ± 0.07%) and twice the CMC (84.0 ± 2.2%) were almost coincident. However, it should be noticed that an in-depth comparison should take in to account the lower CMC observed by these authors (0.50 g/L).

Using the same yeast, the biosurfactant produced from glucose had lower removal yields from silty soil (about 60%) and clay soil (about 80%) than those found in the present study [51].

The yield of motor oil removal from standard sand in the static test in packed columns using only deionized water (control) was lower than that achieved using the metabolic broth containing the crude biosurfactant (59.71 ± 0.08%) (Figure 7). The removal yield using the isolated biosurfactant increased with increasing the concentration from half the CMC to the CMC, reaching 66.62 ± 0.09%. Although the biosurfactant at twice the CMC was less effective in removing the contaminant, it was anyhow more effective than the synthetic chemical surfactant Tween 80 at half the CMC and the CMC.

The yield of motor oil removal from silty soil in the static test was higher compared to the control when using the isolated biosurfactant at half the CMC, reaching 63.03 ± 0.08%. Reductions in the yield were found when using the biosurfactant at the CMC and twice the CMC as well as the chemical surfactant Tween 80. The highest motor oil removal yield from this soil was achieved with the crude biosurfactant (63.60 ± 0.07%), which suggests the presence of some constituent present in the crude biosurfactant that facilitated the removal of the contaminant.

The yield of motor oil removal from clay soil was similar when using the isolated biosurfactant at half the CMC and the crude biosurfactant, with higher removal yields when using the isolated biosurfactant at the CMC and twice the CMC. In comparison to the biosurfactant, Tween 80 did not allow a significantly higher yield of removal of motor oil from this soil.

According to Sales da Silva et al. [51], oil removal yields from clay soil are hampered by the characteristics of this type of soil, such as its considerable water retention, low permeability, and high plasticity. Moreover, although the ionic charge of clay soils is negative in most cases, it can also become positive depending on the exchange of cations [52]. Contaminants have greater interaction with clay soil, which is why this type of soil is used as a contaminant contention barrier [53].

The biosurfactant from S. bombicola ATCC 222214 at the CMC was capable of removing motor oil from standard sand in packed columns with a yield close to that reported by Santos et al. [54] for the biosurfactant from the yeast Candida sphaerica UCP 0995 at the CMC (67%). With regard to silty and clay soils, the biosurfactant from S. bombicola ATCC 222214 grown in the medium containing potato peel flour and waste canola frying oil achieved higher oil removal yields than when it was produced in a glucose-containing medium [51] (approximately 20% for both types of soil).

4. Conclusions

The traditional use of petroleum-based surfactants is due to their low production costs, long shelf life, wide availability, and ability to perform at low temperatures. However, the non-renewable nature of petroleum-based surfactants and the environmental damage they cause have led to the need for a safe and renewable alternative. This has resulted in increased interest in the development of environmentally friendly products. One promising development in this area is the synthesis of green surfactants through biotechnological advances, with microbial surfactants being the most well known. Despite their potential, the high production cost of these surfactants is a significant barrier to increasing their production scale. Challenges in biosurfactant manufacturing include high substrate costs, low productivity, extensive downstream processing, and a lack of understanding of bioreactor systems. However, advances in petroleum biotechnology have revealed the growing recognition and value of various biosurfactants in the petroleum industry due to their diversity and efficiency. These compounds not only have a supporting role but are also beginning to play a crucial role, making them essential for both the petroleum industry and the environment.

In this work, a new yeast, Starmerella bombicola ATCC 222214, was able to produce a biosurfactant in a medium containing waste products as low-cost substrates, achieving a satisfactory reduction in the surface tension. The biosurfactant was considered nontoxic to the living organisms assessed and exhibited stability under different environmental conditions, as well as a promising emulsifying capacity. The purified extract analyzed contained two microbial biosurfactant molecules characterized as saturated fatty acids, namely, 6,6-dimethoxy-octanoic acid and azelaic acid. Thanks to its characteristics, this biosurfactant may find interesting applications in the removal of hydrophobic pollutants from different types of soil under static and dynamic conditions. Therefore, the natural surfactant from the yeast S. bombicola ATCC 222214 can be considered promising in several environmental applications related to the oil sector in particular and to the energy sector in general.