Production of Microbial Lipids by Saitozyma podzolica Zwy2-3 Using Corn Straw Hydrolysate, the Analysis of Lipid Composition, and the Prediction of Biodiesel Properties

Abstract

Although Saitozyma podzolica Zwy2-3 can use the enzymatic hydrolysate of corn stalks treated with an ammonium carbonate-steam explosion (EHCS-ACSE) as a substrate for lipid accumulation, the inefficient conversion of sugars from EHCS-ACSE into lipids necessitates the further optimization of fermentation parameters. Response surface design was used to optimize the primary fermentation parameters. Under the optimized conditions of the reducing sugar concentration of 89.44 g/L, yeast extract concentration of 3.88 g/L, rotational speed of 219 rpm, and incubation time of 122 h, the maximum lipid production achieved 11.45 g/L, which was 2.28 times higher than the results of the previous study. In addition, lipid profiling showed the presence of four fatty acid methyl esters, with the highest percentage being 61.84% oleic acid, followed by 21.53% palmitic acid, 13.05% stearic acid, and 3.58% linoleic acid. It is noteworthy that the composition and relative abundance of microbial lipids remained constant under different culture conditions. The characteristics of Zwy2-3 biodiesel, such as the iodine value (62.09), cetane number (59.29), density (0.87 g/cm³), and oxidation stability (35.53), meet the international standards (ASTM D6751-02 and EN 14214) for biodiesel. The present study further demonstrated that S. podzolica Zwy2-3 can efficiently utilize EHCS-ACSE for microbial lipid accumulation, and its lipids have favorable qualities that make them suitable for biodiesel production.

1. Introduction

The massive consumption of fossil fuels not only causes serious environmental pollution but also triggers the energy crisis. Thus, biodiesel, as a potential renewable alternative to fossil fuels, has gained increasing recognition [1]. Biodiesel has the potential to be carbon-neutral, as well as environmentally friendly with low emissions, and can be a viable alternative to conventional petrodiesel due to its similar fuel properties to petrodiesel [2]. More than 95% of biodiesel is produced from edible oils, which account for approximately 60–80% of the production costs [3]. Using recycled cooking oil can significantly reduce costs. However, cooking waste oil can be of poor quality [4]. In addition, the large-scale conversion of edible oils to biodiesel may bring about problems such as leading to an imbalance in global food supply and demand [5]. Therefore, alternative renewable sources should be considered for biodiesel production as a substitute for edible oils. Microbial oils, also known as single-cell oil (SCO), are being explored as biodiesel feedstocks due to their fatty acid composition, which is similar to that of vegetable oils [6]. In addition, their natural advantage is that the cultivation process is not affected by climate, seasonality, or location, and they require less land [7]. It is estimated that the total production cost of microbial lipids using glucose as a substrate is approximately $2270 per ton, which is 2.5 times higher than that of plant-based biodiesel [8]. Therefore, it is urgent to find abundant and affordable substrates for the large-scale production of microbial oils.

Corn stalk (CS), as an abundant and low-cost lignocellulosic biomass from agriculture, has been utilized in the production of chemicals and biofuels [9]. The production process of SCO from lignocellulosic biomass mainly includes pretreatment for breaking down the complex and resistant structure, the conversion of cellulose and hemicellulose into monosaccharides via chemical or biological hydrolysis, and fermentation by oleaginous microorganisms [10]. Among the different pretreatment processes, hydrothermal, steam explosion (SE), ammonia fiber explosion (AFEX), dilute acid (DA), and dilute alkali (AL) have been widely used by researchers [11]. However, one significant drawback associated with the conventional pretreatment methods is the formation of lignin and sugar degradation compounds, such as furan derivatives, weak acids, and phenolic compounds, which have been found to hinder the growth of microorganisms and the accumulation of desired products [11]. Through a study by Yu et al. [12], DA-CS and AL-CS were used for microbial lipid production by Trichosporon dermatis 32903, resulting in lipid production of 7.46 g/L and 6.81 g/L, respectively. However, the lipid production significantly increased to 11.43 g/L and 20.36 g/L, respectively, when the washing of DA-CS and AL-CS was employed. It was found that the lipid content of Trichosporon cutaneum CX1 increased to 23.5% when utilizing the biodetoxified CS hydrolysate, which was twice as high as the lipid content achieved using the non-detoxified hydrolysate [13]. It was also found that inhibitors derived from lignocellulosic derivatives not only had a negative effect on microbial cells but also on cellulase enzymes, which are essential for enzymatic digestion after pretreatment [14]. For these reasons, there is a need to find a better pretreatment method to reduce the effect of inhibitors on subsequent enzymatic and microbial fermentation. SE combined with weak base reagents is an environmentally friendly and efficient pretreatment method, which effectively removes lignin and produces minimal amounts of inhibitors [15]. In this study, the CS is pretreated via an ammonium carbonate-steam explosion (AC-SE) according to previous research in the laboratory [16].

In addition to the medium components, such as the type and concentration of the carbon and nitrogen sources, the concentration of dissolved oxygen (DOC) is one of the important factors that affects the growth and lipid accumulation of oil-producing microorganisms [17]. Changes in rotational speed, fermentation volume, aeration, and the oxygen mass transfer coefficient are important factors that can induce variations in DOC, which can significantly impact microbial lipid accumulation. For instance, increasing the aeration flow rate from 0 to 2.0 vvm resulted in a noticeable improvement in both biomass and lipid yields in Rhodotorula glutinis, and further increasing the aeration flow rate from 2.0 to 3.0 vvm did not yield any significant changes [18]. In the investigation of Y. lipolytica W29 and Y. lipolytica NCYC 2904 [17], it was observed that increasing the oxygen transfer rate (k_La) from 22 h⁻¹ to 87 h⁻¹ resulted in approximately a 7-fold and a 2.8-fold improvement in microbial lipid production, respectively. However, further increasing the k_La from 87 h⁻¹ to 125 h⁻¹ was found to be detrimental to cellular growth and lipid production. Additionally, the lipid accumulation of Y. lipolytica from pork lard [19] and waste cooking oil [20] was also influenced by the DOC.

The complete hydrolysis of lignocellulosic materials primarily produces glucose and xylose, along with smaller amounts of arabinose, mannose, and lactose [21]. To effectively produce lipids, oleaginous microorganisms must possess the ability to efficiently metabolize a range of sugars, particularly xylose. The yeast strain Saitozyma podozolica Zwy2-3, which was discovered and identified in early laboratory research, shows great potential for xylose metabolism and is a suitable candidate for SCO production [22]. Additionally, it can utilize mixed sugars obtained from EHCS-ACSE [16]. As an unconventional oleaginous yeast, most studies on the yeast strain of S. podzolica primarily utilize pure glucose or xylose as substrates [23,24]. Research on the utilization of low-cost substrates for lipid production by S. podzolica is severely limited. The existing literature only includes two studies: one on S. podzolica Zwy2-3 using sugar derived from corn stover as a substrate [16] and another on S. podzolica DSM 27192 using a combination of glucose and sodium acetate as substrates [25]. Although S. podzolica Zwy2-3 has been observed to use EHCS-ACSE for lipid accumulation, it takes 168 h for the strain to consume the majority of the reducing sugar and the biomass and lipid production achieved are relatively low, at 10.56 g/L and 5.03 g/L, respectively [16]. Additionally, research on S. podzolica DSM 27192 found that the higher DOC resulted in increased lipid accumulation [26]. Previous optimizations primarily targeted media components such as the carbon and nitrogen types and the concentrations, the concentration of MgSO₄·7H₂O, pH, etc., while neglecting factors such as rotational speed and the initial fermentation volume [22]. Therefore, it can be inferred that DOC may be the primary limiting factor for the Zwy2-3 strain to increase biomass and lipid synthesis. This indicates the need for further research and improvements to optimize the fermentation process. Conventional optimization techniques typically focus on examining individual factors while holding all other variables constant. Nevertheless, these methods are associated with several limitations, including being time-consuming, requiring significant effort, and neglecting the potential interactions among the variables being investigated. To address these shortcomings, a statistical experimental design has emerged as a valuable approach [27]. Over the past few decades, a response surface methodology (RSM) has gained widespread popularity in industrial and biotechnological processes. Examples include total lipid production [28] and the production of docosahexaenoic acid (DHA) involving DOC [29].

Microbial oils produced by oleaginous microorganisms possess a distinct fatty acid composition, and the lipid profile can vary depending on the microbial species, substrate type, and growth conditions [30]. Although, the majority of microbial fatty acids are similar to those found in vegetable and animal oils, consisting mainly of C16 and C18 fatty acids [31]. Certain types of oleaginous microorganisms can accumulate specific functional fatty acids, including Crypthecodinium cohnii, Schizochytrium sp, and Ulkenia sp for the synthesis of DHA (22:6, n⁻³); Mortierella alpina ST1358 and Y. lipolytica for the synthesis of eicosapentaenoic acid (EPA, 20:5, n⁻³); M. alpina for the synthesis of arachidonic acid (ARA, 20:4, n⁻⁶); M. alpina 1S-4 for dihomo-gamma-linoleic (DGLA, 20:3, n⁻⁶); and Mucor. circinelloides for the synthesis of γ-linoleic acid (GLA, C18:3, n⁻⁶) [32]. These fatty acids can be used as nutritional supplements for human health. When evaluating the suitability of microbial oils for biodiesel production, it is important to analyze various properties. These include the cetane number, oxidative stability, kinematic viscosity, and cold filter plugging point, all of which must comply with the international standards (such as ASTM D6751-02 [33] and EN 14214 [34]) for biodiesel.

In recent years, numerous researchers have developed mathematical models to predict the relationship between fatty acid composition and biodiesel properties [35]. The easiest method is to use online prediction software (Version 2.2) that is based on these mathematical models. This software (Version 2.2) allows for the direct prediction of biodiesel-related properties by inputting the fatty acid composition and proportions of microbial oils [36].

Although significant progress has been made in utilizing oil yeast as a raw material for biodiesel, the high production cost hampers the commercialization of microbial oil. In the present study, a high-level microbial lipid production from the corn stalk hydrolysate by S. podzolica Zwy2-3 was further conducted. In addition, previous research has provided limited comprehensive analyses regarding the impact of various fermentation conditions on the lipid composition and proportion of microbial strains. This study aims to address this gap by conducting a comprehensive analysis of the fatty acid composition and proportion in microbial strains. Specifically, the investigation examines the effects of various factors, including the dissolved oxygen conditions, types of substrates, and substrate concentrations, both individually and in combination.

2. Materials and Methods

2.1. Processing of Corn Stalks

CS was collected from the outskirts of Chongqing. The EHCS-ACSE was obtained based on previous research conducted in the laboratory [16]. Through a series of treatments, the maximum concentration of reducing sugar from EHCS-ACSE was 47.43 g/L, including glucose (33.02 g/L), xylose (11.36 g/L), and arabinose (3.05 g/L).

The concentration of the total reducing sugars in the enzymatic hydrolysate was determined using the 3,5-dinitrosalicylic acid (DNS) method [37], with glucose as the reference standard. The concentrations of glucose, xylose, and arabinose were determined using high-performance liquid chromatography (Waters e2695, Waters Corporation, Milford, MA, USA) equipment with an Aminex HPX-87P column at 65 °C. The mobile phase consisted of 5 mmol/L of H₂SO₄ at a flow rate of 0.6 mL/min.

2.2. Strain, Mediums, and Culture Conditions

S. podzolica Zwy2-3 was provided by the Sichuan Center of Microbial Resource Collection. The yeast extract peptone dextrose (YPD) medium (g/L) contained glucose 20.0, yeast extract 10.0, and peptone 10.0. The seed medium has the following composition (g/L): X 60.0, yeast extract 10.0, KH₂PO₄ 3.0, and MgSO₄·7H₂O 1.5. X stands for glucose, xylose, or a mixture of sugars (glucose:xylose, 3:1). The main types of sugars used in the seed medium should be the same as those required for the fermentation medium. The initial fermentation medium compositions (g/L) were as follows: reducing sugar 60.0, yeast extract 1.5, and MgSO₄·7H₂O 1.5. All culture media were adjusted to an initial pH of 6.0 and sterilized at 115 °C for 20 min.

The inoculum preparation requires two activations with incubation conditions of 28 °C and 180 rpm. (1) A 5 mL YPD broth was inoculated with a 250 μL glycerol stock of S. podzolica Zwy2-3 (acquired from a single colony, stored at −80 °C) for 72 h; (2) Then, a 2.5 mL cell suspension was added to a 100 mL seed medium in a 500 mL flask and the culture was incubated for 24 h, which was used as the inoculum. In the initial fermentation experiments, a 100 mL fermentation medium with a 5% inoculation volume was incubated at 28 °C, 180 rpm for 144 h.

2.3. Effect of Major Fermentation Parameters on Biomass and Lipid Production

In order to examine the impact of fermentation volume on lipid production, a series of fermentations were conducted at varying initial fermentation volumes (30, 50, 70, 100, and 120 mL) while keeping all other conditions constant. Similarly, to investigate the influence of shaker speed on lipid production, fermentations were carried out at different shaker speeds (100, 140, 180, 220, and 260 rpm) using the optimal fermentation volume. Additionally, in order to investigate the impact of the reducing sugar concentration on lipid production, fermentations were conducted at various levels of reducing sugar concentrations (40, 60, 80, 100, and 120 g/L) while maintaining the optimal fermentation volume and shaker speed. All experiments were conducted in 500 mL conical flasks, with an initial pH of 6.0, a shaker temperature of 28 °C, and a fermentation duration of 144 h.

2.4. Experimental Design for Optimization of Lipid Production Using RSM

Based on previous experimental results [22] and a single factor experiment in this study, the production of lipids by the strain Zwy2-3 was optimized using a 3-level 4-factor Box–Behnken design (BBD) experiment. The software Design-Expert V8.0.6.1 (Stat-Ease Inc., Minneapolis, MN, USA) was used for this purpose.

The BBD can evaluate potential interactions between pairs of factors and minimize the number of experiments required. Four independent variables, including the reducing sugar concentration (A), yeast extract concentration (B), rotation speed (C), and fermentation time (D), were studied at three different levels (−1, 0, and +1) to determine the optimal fermentation conditions and their interactions for lipid production (

Table 1. Levels of factors for the Box–Behnken design.

Factors	Independent Variables	Range and Levels
		−1	0	+1
A	Reducing sugar concentration (g/L)	60.00	80.00	100.00
B	Yeast extract concentration (g/L)	1.50	3.00	4.50
C	Rotation speed (rpm)	120	180	240
D	Fermentation time (h)	96	120	144

).

A total of 27 runs were performed using the BBD. Three of these runs were repeated at the center point to estimate the variance of pure experimental uncertainty.

Table 2. Experimental results based on 3-level 4-factor Box–Behnken response surface design.

	Variables				Response
Run	Reducing Sugar Concentration (g/L)	Yeast Extract Concentration (g/L)	Rotation Speed (rpm)	Fermentation Time (h)	Lipid Production (g/L)
	Factor A	Factor B	Factor C	Factor D	Actual	Predicted
1	60.00	1.50	180	120	7.18	7.02
2	100.00	1.50	180	120	7.04	6.95
3	60.00	4.50	180	120	6.65	6.69
4	100.00	4.50	180	120	9.75	9.86
5	80.00	3.00	120	96	4.85	5.19
6	80.00	3.00	240	96	10.15	9.95
7	80.00	3.00	120	144	6.48	6.63
8	80.00	3.00	240	144	9.66	9.27
9	60.00	3.00	180	96	7.41	7.28
10	100.00	3.00	180	96	9.04	8.52
11	60.00	3.00	180	144	6.94	7.34
12	100.00	3.00	180	144	9.19	9.20
13	80.00	1.50	120	120	5.76	5.46
14	80.00	4.50	120	120	5.28	5.45
15	80.00	1.50	240	120	8.14	7.85
16	80.00	4.50	240	120	10.28	10.46
17	60.00	3.00	120	120	4.97	4.62
18	100.00	3.00	120	120	6.11	6.08
19	60.00	3.00	240	120	8.03	8.23
20	100.00	3.00	240	120	9.35	9.87
21	80.00	1.50	180	96	7.16	7.76
22	80.00	4.50	180	96	8.32	8.24
23	80.00	1.50	180	144	7.07	7.32
24	80.00	4.50	180	144	9.86	9.43
25	80.00	3.00	180	120	10.37	10.27
26	80.00	3.00	180	120	10.01	10.27
27	80.00	3.00	180	120	10.44	10.27

lists the experiments designed via the software (V8.0.6.1) for optimization research, along with the variable values at various levels. According to the experimental results, a quadratic model was generated, and the significance of the model was tested using ANOVA analysis.

All response surface experiments were conducted in 500 mL shake flasks containing 50 mL of the fermentation broth. The experiments involved simultaneous changes in the reducing sugar concentration (60, 80, and 100 g/L), yeast extract concentration (1.5, 3.0, and 4.5 g/L), rotation speed (120, 180, and 240 rpm), and fermentation time (96, 120, and 144 h). In addition, the initial pH of 6.0, the concentration of MgSO₄·7H₂O at 1.5 g/L, and the fermentation temperature at 28 °C remain unchanged.

2.5. Determination of Cell Dry Weight and Lipid Extraction

Cell dry weight (CDW) analysis was conducted in triplicate for each sample using gravimetric analysis. The fermentation broth was collected in a pre-weighed 100 mL centrifuge tube and centrifuged at 8000 rpm for 5 min. The supernatant was utilized to determine the concentration of residual sugar, while the yeast cells were washed twice with 10 mL of physiological saline (0.9% w/v, NaCl) and dried at 80 °C until a constant weight was achieved.

Lipid extraction was performed using an improved version of the Folch method [38]. A 0.5 g sample of dried biomass was placed in a 50 mL centrifuge tube and mixed with 15 mL of a 4 mol/L HCl solution. The mixture was then agitated for 3 h at a temperature of 28 °C and a speed of 180 rpm. The pretreated biomass was subjected to boiling in a water bath for 10 min, followed by rapid cooling in an ice water bath for 10 min, with this process repeated once. Subsequently, a mixture of 7.5 mL of chloroform and methanol (2:1, v/v) was used as the extraction reagent. The mixture was shaken for 3 h and then centrifuged at 8000 rpm for 5 min. Finally, the chloroform layer was transferred to a pre-weighed container and placed in a constant temperature oven at 80 °C to facilitate the complete evaporation of the solvent. The weight of the extracted lipid was determined using the gravimetric method, and the lipid was stored in a sealed glass tube for further analysis. The biomass, lipid production, and lipid content were evaluated using the following equations:

Biomass (g/L) = Weight_{dried yeast cell}/Volume_{fermentation medium}

(1)

Lipid production (g/L) = Weight_{extracted lipid}/Volume_{fermentation medium}

(2)

Lipid content (wt%) = 100% × (Weight_{extracted lipid}/Weight_{dried yeas tcell})

(3)

where Volume_{fermentation medium} is the volume of the fermentation medium. Weight_{extracted lipid} and Weight_{dried yeast cell} are the weights of the extracted lipid and the dried yeast cell, respectively.

2.6. Lipid Composition Analysis by Fatty Acid Methyl Ester (FAME)

Fatty acids from the S. podzolica Zwy2-3 were converted into fatty acid methyl esters (FAME) via transesterification [39]. The specific operation procedure was as described later. Initially, 1 μL of lipid was added to a glass tube containing 2 mL of a 0.6 mol/L potassium hydroxide–methanol solution and 2 mL of n-hexane. After shaking for 2 min, the mixed solution was put into a 37 °C water bath to fully react for 15 min. Subsequently, 5 mL of primary water was added to the mixture, which allowed the layers to separate. Finally, the upper n-hexane layer containing FAME was extracted. The transesterification product was diluted with n-hexane to achieve the desired concentration and analyzed using a GC-2010 Pro (Shimadzu Corporation, Kyoto, Japan) equipped with an HP88 capillary column (100 m × 0.25 mm × 0.20 μm). The column temperature was maintained at 150 °C for 1 min and then increased to 250 °C at a rate of 10 °C/min. It was held at 250 °C for 2 min. Nitrogen was used as the carrier gas at a flow rate of 5 mL/min. The split ratio and the injection volume were 1:10 (v/v) and 1 μL, respectively. The percentage of the fatty acid methyl ester was calculated using the area normalization method [40]. In order to enhance the comprehension of the procedures involved in biomass determination, oil production determination, oil extraction, and fatty acid methyl ester determination, a schematic diagram illustrating the entire process has been created (Figure 1).

2.7. Prediction of Biodiesel Properties Based on the Lipid Profile

In this research, a range of physicochemical characteristics of biodiesel, including the cetane number (CN), oxidation stability (OS), density (ρ), iodine value (IV), high heating value (HHV), and saponification value (SV), were assessed using the online platform Biodiesel Analyzer (Version 2.2). The evaluation was performed for each individual fatty acid and can be accessed at the following URL: http://www.brteam.org/analysis/ (Accessed on 8 September 2022) [36].

2.8. Statistical Analysis

The data analysis of response surfaces was carried out using the internal statistical tool in Design Expert 8.0 software (Stat-Ease, Minneapolis, MN, USA). The “*” symbol indicates significance (p < 0.05), “**” indicates a high level of significance (p < 0.01), and “ns” is used for non-significant results. Multiple comparisons were performed using SPSS 19.0 statistical software (IBM, Chicago, IL, USA). One-way ANOVA and Tukey’s test (p < 0.05) were used for the analysis. Different letters indicate significant differences between groups. The letters or combinations of letters marked with the same letter indicate that the differences between the two sets of data are not significant. The fatty acid profiles of microbial oils were expressed as the average of the three experiments.

3. Results and Discussion

3.1. Optimization of Single-Factor Experiments

3.1.1. Effect of Fermentation Volume and Shaker Speed on Lipid Accumulation

As can be seen from Figure 2a, the growth and lipid accumulation of the strain Zwy2-3 were significantly influenced by different initial fermentation volumes. At the initial fermentation volume of 30 mL, the oil production reached a maximum of 8.48 g/L, which was not significantly different from the production of 8.25 g/L at 50 mL. Additionally, the biomass was approximately 15.0 g/L under both conditions. However, as the initial fermentation volume increased, both the biomass and oil production decreased significantly. The lowest biomass and oil yield was observed at an initial fermentation volume of 120 mL, measuring 11.77 g/L and 5.68 g/L, respectively. From the experimental results, it can be seen that reducing the initial fermentation volumes leads to a more favorable accumulation of biomass and oil. This phenomenon can be attributed to the increase in the dissolved oxygen levels as the volume decreases. Notably, when the volume is reduced to 50 mL and 30 mL, there is no significant difference in oil production. Taking into account the overall reliability of the experimental data, it has been determined that a volume of 50 mL is the optimal volume for the subsequent fermentation experiments. In the laboratory research phase, it is feasible to use a smaller fermentation volume. In the context of large-scale fermentations, where high volumes and densities are required, it is not feasible to regulate DOC by reducing the fermentation volume. Various methods can be employed to increase DOC, such as increasing the rotational speed, enhancing aeration, utilizing gas mixing technology (combining air and pure oxygen in specific proportions), and introducing oxygen carriers. However, an alternative approach that shows promise for future commercial production is the co-fermentation of yeast and algae [41]. This approach effectively meets the yeast’s requirement for dissolved oxygen without incurring any additional costs.

The impact of shaker speed on growth and lipid accumulation is illustrated in Figure 2b. In the range of shaker speeds from 100 to 180 rpm, both biomass and lipid production gradually increased with higher shaker speeds. When the speed reached 180 rpm, the lipid content peaked at 54.44%, and the biomass reached 14.96 g/L. When the rotational speed was increased to 220 rpm and 260 rpm, there was a slight increase in biomass and oil yields compared to 180 rpm, but the differences were not significant. The carbon source concentration of 80 g/L was selected for the following experiments.

The increase in rotational speed and the decrease in initial fermentation volume had a similar effect on the biomass and oil production of the strain Zwy2-3. The DOC is a crucial factor that impacts the microbial fermentation process [42]. During fermentation, changes in agitation speed and fermentation volume can indirectly affect the DOC by altering the rate of gas–liquid mass transfer in the fermentation solution and the surface area-to-volume ratio [43]. Changes in DOC can impact the biomass and lipid production of oil-producing microorganisms. This is because the energy metabolism and biosynthesis, particularly the lipid desaturation process, of these microorganisms depend heavily on molecular oxygen. Once a certain threshold is reached, other nutrient concentrations may limit the increase in biomass and oil production.

3.1.2. Effect of Reducing Sugar Concentration on Lipid Accumulation

Figure 2c illustrates the impact of the reducing sugar concentration on the growth and lipid accumulation of the strain Zwy2-3. The biomass and lipid production of the strain Zwy2-3 increased as the concentration of reducing sugar increased from 40 to 80 g/L. The maximum biomass concentration of 17.01 g/L and the lipid content of 54.08% were achieved at a glucose concentration of 80 g/L. After reaching a peak at 80 g/L, biomass and lipid production decreased as the reducing sugar concentration increased up to 100 g/L. The lowest production was observed at 120 g/L. This phenomenon in the study may be attributed to a fact that during the concentration process, the corn straw enzymatic hydrolysate experienced an increase in the concentration of trace inhibitors, which hindered the strain’s ability to accumulate lipids. Depending on the specific pretreatment method employed and the various process parameters, such as temperature and duration, lignin is degraded into phenolics, while fermentable sugars are primarily degraded into furfural, HMF, acetic acid, levulinic acid, ferulic acid, and glucuronic acid [11]. Some studies have shown that the enzyme hydrolysis of corn stover pretreated with SE did not exhibit the inhibition of furfural, HMF, and acetic acid. However, it was found that lignin degradation was the primary factor contributing to the inhibition of microbial fermentation [44]. In our previous experimental investigation, the application of AC-SE showed similar effects on cellulose and hemicellulose while demonstrating a superior effectiveness in removing lignin compared to SE [16]. Therefore, it can be inferred that the main inhibitory effect on yeast growth and lipid accumulation in this study is caused by phenolic compounds. Nevertheless, ferulic acid, 3-methyl-4-hydroxybenzoic acid, 4-hydroxyacetophenone, and vanillic acid are typical phenolic compounds [45]. Therefore, a thorough and extensive investigation is necessary to determine which of these substances has a significant inhibitory effect on fermentation in the future work.

Currently, various substrates are used in the research of microbial oils, including glucose and other sugars, glycerol, lignocellulose, acetic acid, and volatile fatty acids (VFAs) [46]. However, due to cost considerations, researchers are now focusing on utilizing affordable and easily accessible substrates. According to Wang et al., the annual global production of CS is estimated to be 1661.25 million tons [47], which represents approximately 27.2% of the total agricultural waste generated worldwide [48]. Hence, the use of CS not only significantly decreases manufacturing costs but also helps to reduce environmental pollution caused by incineration.

3.1.3. Effect of Yeast Extract Concentration on Lipid Accumulation

In this experiment, four yeast extract concentration gradients of 0.5 g/L, 1.5 g/L, 3.0 g/L, and 4.5 g/L were selected and incubated for 144 h under the following conditions: reducing sugar concentration of 80.0 g/L, fermentation temperature of 28 °C, fermentation volume of 50 mL, and shaking speed of 180 rpm. The effects of varying concentrations of yeast extract on the growth and oil accumulation of Zwy2-3 were investigated (Figure 2d). When the concentration of yeast extract was in the range of 0.5–3.0 g/L, both yeast biomass and oil production increased. When the concentration of yeast extract reached 3.0 g/L, the maximum oil production was observed at 10.04 g/L. However, when the concentration of yeast extract was further increased to 4.5 g/L, the biomass reached its maximum, but oil production decreased. The yeast extract concentration of 3.0 g/L was selected as the best.

The nitrogen source plays a crucial role in microbial growth and product synthesis. It serves as the primary source of nitrogen for the intracellular synthesis of nucleic acids, proteins, enzymes, and other essential compounds. Under nitrogen-limited conditions, rapid oil accumulation occurs when the carbon source is sufficient. In other words, a high carbon-to-nitrogen ratio (C/N) promotes oil accumulation, whereas a low C/N favors the growth of microbial organisms [46]. A research investigation conducted on R. glutinis, a type of yeast, cultivated using glucose as the primary carbon source, demonstrated a significant 32% increase in lipid content when the C/N ratio was elevated from 20 to 70, while the lipid concentration exhibited a substantial 4.5-fold increase [49]. Therefore, optimizing the nitrogen concentration is crucial for maximizing the biomass and lipid production of the strain Zwy2-3.

3.1.4. Effect of Fermentation Time on Lipid Accumulation

Other fermentation conditions were kept constant, and the impact of the fermentation period (24 h, 48 h, 72 h, 96 h, 120 h, and 144 h) on the biomass and oil yield of the strain Zwy2-3 was examined (Figure 2e). The impact of fermentation time on the oil accumulation of the strain Zwy2-3 reveals that the biomass, oil production, and oil content of the strain increase as the time extends from 0 to 120 h. The highest values were observed at 120 h, with biomass and oil production reaching 19.42 g/L and 10.3 g/L, respectively. However, a slight decrease in oil production was observed as the fermentation continued. This may be due to the fact that at 120 h, the carbon source was basically exhausted, and the cells began to utilize intracellular substances to maintain their basic physiological state during continued fermentation. Therefore, choosing 120 h as the fermentation period would be more appropriate.

3.2. Assessment of the Regression Model and Adequacy Check for Lipid Production

The design matrix of the variables with experimental results is shown in Table 2. It is evident that lipid production varies depending on the level of each factor. The experimental data from 27 runs was used to estimate second-order polynomial equation models for lipid production, and the predictive equation (Equation (4)) was established as shown below.

Lipid production (g/L) = 10.27 + 0.77 × A + 0.65 × B + 1.85 × C + 0.19 × D +
0.81 × A × B + 0.045 × A × C
+ 0.16 × A × D + 0.65 × B × C + 0.41 × B × D − 0.53 × C × D − 1.37 × A² − 1.27 × B² − 1.70 × C² −
0.81 × D²

(4)

where the lipid production is the response value; A, B, C, and D is four independent variables representing the reducing sugar concentration (g/L), yeast extract concentration (g/L), rotation speed (rpm), and fermentation time (h), respectively.

The analysis of variance (ANOVA) results for lipid production by the yeast strain Zwy2-3 are presented in

Table 3. Analysis of variance (ANOVA) results for the actual responses using BBD quadratic model for lipid production of the yeast strain Zwy2-3.

Source	Sum of Squares	Df	Mean Square	F-Value	p-Value
Model	80.78	14	5.77	31.18	<0.0001	significant
A	7.21	1	7.21	38.95	<0.0001	**
B	5.06	1	5.06	27.33	0.0002	**
C	40.92	1	40.92	221.14	<0.0001	**
D	0.43	1	0.43	2.32	0.1536
AB	2.62	1	2.62	14.18	0.0027	**
AC	8.100 × 10−3	1	8.100 × 10−3	0.044	0.8378
AD	0.096	1	0.096	0.52	0.4849
BC	1.72	1	1.72	9.27	0.0102	*
BD	0.66	1	0.66	3.59	0.0825
CD	1.12	1	1.12	6.07	0.0298	*
A2	10.07	1	10.07	54.39	<0.0001	**
B2	8.60	1	8.60	46.49	<0.0001	**
C2	15.39	1	15.39	83.17	<0.0001	**
D2	3.54	1	3.54	19.14	0.0009	**
Residual	2.22	12	0.19
Lack of Fit	2.11	10	0.21	3.97	0.2178	not significant
Pure Error	0.11	2	0.053
Cor Total	83.00	26
R2 = 0.9732			Pred R2 = 0.8504
Adj R2 = 0.9420			Adeq Precision = 18.187

Note: * indicates the significant level with p < 0.05; ** indicates the extremely significant level with p < 0.01.

. It has been found that the quadratic model is statistically significant, as indicated by a high F value (31.18) and a small p value (<0.0001). This indicates that the probability of the F value being caused by noise is only 0.01%. The lack of fit is not significant with a p value of 0.2178 (>0.05), indicating that accidental factors may not have a significant impact on the response value. The value of R² (0.9732) indicates that 97.32% of the total variation for lipid production can be explained via the independent variables, while the remaining 2.68% represents the portion that the model cannot account for. The adjusted R² value (0.9420) corrects the R² values by removing the influence of the number of independent variables. This adjustment enables a more precise assessment of the explanatory capability of the quadratic regression equation. In this case, there is a strong correlation between the adjusted R² value (0.9420) and the R² value (0.9732), indicating that the model’s predicted and actual data have a statistically significant relationship. This indicates reliability in the experimental design (Figure 3a). The residuals for the experimental runs are shown in Figure 3b, and it is evident that the values are randomly distributed within the range of ±3. The normal plot of the residuals (Figure 3c) shows that the errors are randomly distributed along a straight line, convincingly demonstrating that the model has a strong predictive ability for optimizing lipid production. Adequate precision measures the ratio of signal to noise, with a value greater than 4 being a necessary prerequisite for a good model fit [50,51]. In this study, the precision of 18.187 indicates that the model is sufficiently accurate for exploring the design space. In short, all the results discussed above indicate that the second-order polynomial model accurately describes the observed data of lipid production. The data in Table 2 also shows the comparison between the experimental value and the model-predicted value of lipid production for each experiment.

3.3. Response Surface Plots Analysis

The three-dimensional (3D) graph of the response surface and its corresponding contour map can conveniently determine the optimal process parameters and help us understand the interactions between variables [52].

It is evident from Table 3 that the linear terms, A, B, and C, as well as their quadratic counterparts, A², B², C², and D² and the interaction terms, AB, BC, and CD, are significant model terms. To gain a deeper understanding of the relationships between these important factors, we examined the impact of the interaction between two independent variables on lipid production. This was achieved by creating a 3D surface plot and the corresponding contour plots, while also keeping all other variables constant. The 3D response surface plots and the corresponding contour plots for the interactions between independent variables are shown in Figure 4. It has been found that the process variables have significant interaction effects.

The results in Figure 4a revealed the effects of the reducing sugar concentration and the yeast extract concentration on the response value. The results showed that lipid production increased with the increasing concentrations of reducing sugar and yeast extract. A maximum response value was obtained when the concentration of reducing sugar was 89.44 g/L and the concentration of yeast extract was 3.88 g/L. Subsequently, the lipid production of the strain Zwy2-3 declined as the concentration of reducing sugar and yeast extract increased. Figure 4b illustrates the impact of the yeast extract concentration and rotation speed on lipid production. The results indicate that lipid production rapidly increases initially and gradually decreases with an increase in both factors. Similarly, Figure 4c demonstrates the effect of rotation speed and fermentation time on the response variable, exhibiting a similar trend to that observed in Figure 4b.

According to the analysis findings, the optimal conditions for the concentration of reducing sugar and yeast extract, rotation speed, and fermentation time were found to be 89.44 g/L, 3.88 g/L, 219 rpm, and 122 h, respectively. The experimental lipid production achieved under these conditions was 11.45 g/L, which closely aligned with the predicted result of 11.26 g/L. These results suggest that the regression model is accurate and appropriate for predicting the lipid production of S. podzolica Zwy2-3.

3.4. Lipid Accumulation Ability of S. podzolica Zwy2-3 under Optimized Conditions

To achieve high production of microbial lipids, the strain Zwy2-3 was employed for the shake flask fermentation using EHCS-ACSE as the carbon source. Under the optimum conditions, the time courses of biomass, lipid production, lipid content, and sugar consumption by S. podzolica are shown in Figure 5. S. podzolica Zwy2-3 exhibited a favorable growth pattern in the CS hydrolysate medium.

The data shows that cell growth and lipid accumulation rapidly and steadily increase in the first 72 h, accompanied by a significant decrease in the glucose levels (Figure 5a). The lipid content rapidly increases within 24 h, then gradually stabilizes during the remaining fermentation time (Figure 5a). The initial high lipid content may be attributed to the high carbon-to-nitrogen ratio in the medium during the second activation incubation. This led to some extent of lipid accumulation in the yeast, in addition to the inherent characteristics of the strain. When the glucose sugar was completely utilized within 96 h, xylose and arabinose were subsequently utilized. In microbial fermentation, there is a phenomenon known as the glucose effect. This occurs when the strain begins to utilize other carbon sources after all the glucose has been consumed or when its concentration drops below a certain threshold. Despite the glucose effect, there was no significant delay in the utilization of other pentose sugars (Figure 5a). The research on the R. glutinis conversion of the switchgrass hydrolysate into lipids revealed that the utilization of xylose was not significantly observed until the concentration of glucose dropped below 10 g/L [53]. Similar results were also obtained when the researchers investigated the utilization of the corn stover hydrolysate as a carbon source for producing microbial oils by Trichosporon cutaneum [13], as well as the co-utilization of corn stover hydrolysates and glycerol by Cryptococcus curvatus for lipid production [54]. Considering the prevalence of glucose inhibition, except for a very few strains that can utilize glucose and xylose simultaneously [55], the ability to ferment without a significant delay period is of great importance in reducing the overall fermentation time. Compared to previous research results [16], this study observed that the two key parameters, biomass and lipid production, achieved high values in a relatively short time (Figure 5a). Additionally, the ability of microorganisms to assimilate sugars was significantly enhanced, primarily due to the increased DOC resulting from higher rotational speed and reduced fermentation volume. This heightened DOC plays a crucial role in facilitating the rapid growth of microorganisms. A similar phenomenon was observed in the study of oil accumulation in Y. lipolytica W29. The study found that increasing k_La from 22 h⁻¹ to 87 h⁻¹ resulted in a 95% increase in substrate assimilation and an approximate 7-fold increase in microbial lipid production [17]. After 122 h of fermentation, a DCW of 21.75 g/L was achieved, with a total lipid production of 11.45 g/L and a lipid content of 52.64%. (Figure 5a). After fermentation, the analysis revealed a concentration of 0.53 g/L of arabinose sugar. This finding indicates that the majority of the sugar present in the CS hydrolysate medium, as prepared in this study, was converted into biomass and SCO. Based on previous research [16] and the results of this experimental study, the mass balance of lipid production from corn stalks was evaluated and is presented in Figure 5b. The results indicate that a lipid coefficient of a 61.4 g/1000 g corn stalk was achieved in this investigation.

The yeast strain Rhodotorula taiwanensis AM2352 accumulated 60.3% intracellular oil and reached a biomass concentration of 18.7 g/L after 108 h of fermentation on the corncob hydrolysate [56]. Although the lipid production of 11.28 g/L aligns closely with the findings of this study, the lipid coefficient of 55.8 (g/1000 g corncob) is lower than that of 61.4 (g/1000 g CS) [56]. Under the optimum conditions, Trichosporon dermatis 32,903 demonstrated a lipid production of 20.36 g/L and a lipid content of 55.97% when cultivated in a medium containing the AL-CS hydrolysate [12]. In comparison, T. dermatis 32,903 produced 11.43 g/L of lipids when grown in a medium containing the DA-CS hydrolysate [12]. This finding is consistent with the lipid production observed in this study. In another report, the cultivation of R. toruloides Y4 in the CS hydrolysate pretreated with an ionic liquid-based organic electrolyte resulted in a maximum lipid production of 7.04 g/L [57]. Many similar results have also been reported, such as 7.43 g/L of lipid production by Cryptococcus curvatus [58] and 7.90 g/L of lipid production by Rhodotorula graminis [59]. In short, the yeast strain Zwy2-3 employed in this investigation exhibited a higher capacity for microbial lipid production compared to most other strains. This observation implies that the yeast strain Zwy2-3 has the potential to be a suitable candidate for SCO production using low-cost substrates.

3.5. Comparative Analysis of Fatty Acid Composition and Proportions under Various Culture Conditions

Many studies have shown that the main fatty acids in the majority of analyzed yeast strains are palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1), and linoleic acid (18:2). Minor fatty acids include myristic acid (14:0), palmitoleic acid (16:1), α-linolenic acid (18:3), peanut acid (20:0), behenic acid (22:0), and lignoceric acid (24:0) [60,61]. Related studies have also demonstrated that the fatty acid composition of microbial lipids has a significant impact on the physical properties of biodiesel and the most suitable fatty acids for biodiesel production are primarily C16 and C18 fatty acids [31].

The fatty acid profiles of the lipid samples analyzed via GC-MS are shown in Figure 6. Under optimal fermentation conditions, the major fatty acids were shown to be oleic acid (61.84%), palmitic acid (21.53%), stearic acid (13.05%), and linoleic acid (3.58%). When glucose, xylose, and a mixed sugar (G:X, 3:1) were used as substrates for lipid production under optimal fermentation conditions, the main fatty acids detected were oleic acid (60.90%, 62.03%, and 62.55%), palmitic acid (20.35%, 21.56%, and 19.72%), stearic acid (13.16%, 11.94%, and 12.92%), and linoleic acid (5.59%, 4.48%, and 4.81%). This composition was similar to what was observed when EHCS-ACSE was used as the substrate (Figure 6a). The change in carbon source does not significantly affect the proportion of the main fatty acid composition of the strain Zwy2-3, except for linoleic acid. Compared to the other three substrates, the strain Zwy2-3 obtained the smallest proportion of linoleic acid (3.58%, Figure 6a) in fatty acids when using EHCS-ACSE as a substrate. This may be due to the presence of specific inhibitors in the concentrated CS hydrolysate that disrupt the fatty acid desaturation pathway in yeast. This phenomenon has also been found in a previous study, in which the strain Zwy2-3 was fermented using EHCS-ACSE as the substrate at an initial concentration [16]. Additionally, research has demonstrated that the fatty acid composition of the oil produced by C. curvatus ATCC20509 remains relatively stable when fermenting glucose or xylose as substrates for a duration of 67 h [62]. However, it has been observed that the proportion of individual fatty acids varies depending on the substrates used in the investigation of Y. lipolytica strains [17] and Lipomyces starkeyi DSM 70296 [62].

Given the significant impact of dissolved oxygen on the growth and lipid accumulation of yeast strains, it is crucial to investigate the effects of higher shaker speeds with a 50 mL initial fermentation volume and lower shaker speeds with a 100 mL initial fermentation volume on the composition and proportion of fatty acids. Therefore, glucose, xylose, mixed sugar (G:X, 3:1), and EHCS-ACSE were used as carbon sources for lipid production under different rotation speeds and fermentation volumes. It is evident that an increase in dissolved oxygen, resulting from rotation speed and fermentation volume, has little impact on the composition and proportion of the main fatty acids during fermentation with the same carbon source (Figure 6b,c). Similar observations were reported by Qian et al., who cultivated S. podzolica DSM 27192 using glucose as a substrate and found that increasing the dissolved oxygen (from 600 rpm to pO₂ > 40%) did not significantly alter the composition and proportion of various fatty acids [26]. In addition, it was also found that after significantly increasing the rotation speed, the proportion of various fatty acids remained stable when using different carbon sources as substrates, except for linoleic acid (Figure 6b vs. Figure 6a). In conditions of reduced dissolved oxygen, the composition of lipids obtained from the strain Zwy2-3 remains unchanged when different carbon sources are used as substrates (Figure 5c). Additionally, the proportions of each component remain consistent with those observed under other conditions of high dissolved oxygen (Figure 6c vs. Figure 6a,b). Similar results were reported by Rakicka et al., who cultivated Y. lipolytica using molasses and crude glycerol as substrates under different oxygenation conditions and found that there was only a slight alteration in the composition and proportion of fatty acids [63]. Furthermore, in order to comprehensively evaluate the impact of modifications in culture conditions on lipid composition, we simultaneously manipulated the concentration of reducing sugar and yeast extract, shaker speed, and fermentation time. This allowed us to analyze the influence of these modifications on the composition and ratios of fatty acids. The results showed that, despite the slight change in values, the composition and proportion of fatty acids remained relatively stable and consistent with those of the other experimental groups (Figure 6d).

A recently reported study on the enhancement of lipid production by Sporobolomyces roseus CFGU-S005 yeast found that when fermentation was carried out at the lowest k_La (16.6 h⁻¹), the contents of monounsaturated fatty acids (MUFAs) and saturated fatty acids (SFAs) in the microbial oils were 57% and 40%, respectively, while, the contents of polyunsaturated fatty acids (PUFAs) in the microbial oil significantly increased at k_La values of 22.44 and 32.16 h⁻¹ [64]. In the investigation of Y. lipolytica, researchers observed that the relative percentage of fatty acids in Y. lipolytica W29 was significantly affected by the level of oxidation. They found that there was a 50% increase in unsaturated fatty acids as the k_La increased from 22 h⁻¹ to 87 h⁻¹ [17]. In the field of mold research, it was found that after 168 h of fermentation, the proportion of fatty acid composition changes with varying levels of dissolved oxygen. Specifically, two types of fatty acids, oleic acid and arachidonic acid, experienced the most significant changes [65]. In the field of microalgae research, it is widely recognized that culture conditions, particularly aeration and stirring, are vital for the growth of microalgae and the production of fatty acids. Optimal conditions for biomass formation require high DOC, while low DOC is more favorable for the production of DHA [66]. In the field of bacterial research, the necessary lipid composition can be achieved by regulating DOC. For example, Alcanivorax borkumensis SK2 was found to accumulate triacylglycerol and wax esters under conditions of limited oxygen supply, and the production of polyhydroxyalkanoates can be induced by alternating DOC [67].

In conclusion, for most oil-producing microorganisms, an increased DOC in the appropriate range has the potential to promote cell growth and lipid accumulation. However, it is important to note that the lipid composition, as well as the proportion of fatty acids, may be affected by this increase. This variation is usually detrimental to the oil-producing strains used in biodiesel production. However, it can be a regulatory optimization strategy for the production of functional PUFAs, such as DHA [29]. Microbial strains that demonstrate the ability to maintain a consistent fatty acid composition and proportion under different DOC and substrate conditions are more advantageous for the biodiesel production.

Although it has been reported that the distribution of fatty acids can vary depending on the strain type, strain genetic makeup, cultivation conditions, and medium composition [68]. It has been found that at least 94% of the fatty acids obtained from S. podzolica are C16 and C18 fatty acids. Furthermore, oleic acid (18:1) and palmitic acid (16:0) account for over 80% of the total fatty acids, with proportions of approximately 60% and 20%, respectively [16,26]. The consistent stability of the primary fatty acid composition and proportions under various cultivation conditions makes S. podzolica more suitable for biodiesel production. The use of alternative sources of inexpensive carbon, such as glycerol, acetate, and molasses, should be considered in future studies to assess their impact on the fatty acid composition and ratios of the strain S. podzolica Zwy2-3.

Table 4. Comparison of fatty acid composition between vegetable oil ^a and the lipid from S. podzolica Zwy2-3 ^b.

Lipid Source	Fatty Acid Relative Percentage (%)
	C14:0	C16:0	C18:0	C18:1	C18:2	C18:3
Corn	ND c	7–13	2.5–3	30.5–43	39–52	1
Soybean	ND c	2.3–11	2.4–6	22–30.8	49–53	ND c
Safflower (high oleic)	ND c	4–8	2.3–8	73.6–79	11–19	2–10.5
Palm	0.62.4	32–46.3	4–6.3	37–53	6–12	ND c
Sunflower	ND c	3.5–6.5	1.3–5.6	14–43	44–68.7	ND c
Olive	1.3	7–18.3	1.4–3.3	55.5–84.5	4–19	ND c
S. Podzolica Zwy2-3	ND c	21.53	13.05	61.84	3.58	ND c

^a The composition of different vegetable oils refers to the previous report [69]; ^b The fatty acid composition of lipid from S. podzolica Zwy2-3 under optimal fermentation conditions with hydrolysate of (NH₄)₂CO₃-SE corn stover; ^c ND means “not detected”.

provides a comparison of the fatty acid composition between the lipids extracted from S. podzolica Zwy2-3 and the commonly used vegetable oils in biodiesel production [69]. It can be observed that the fatty acid composition of the lipid extracted from the strain Zwy2-3 is similar to that of vegetable oils. This similarity makes it a potential substrate for biodiesel production.

3.6. Prediction of Biodiesel Properties Based on the Lipid Profile

The synthesis of biodiesel is influenced by several factors, including the type of catalyst, reaction temperature, reaction time, alcohol–oil molar ratio, and the water and free fatty acid content [70]. Among these factors, the choice of catalyst is crucial because different catalysts may result in the superior transesterification of specific types of oils and alcohols [71]. The synthesis of biodiesel and the determination of its fuel properties is a time-consuming and complex process. The fatty acid composition of the feedstock is correlated with biodiesel properties, and therefore, many studies have focused on modeling equations based on the fatty acid composition and ratios, fatty acid chain lengths, and molecular structures to predict the fuel properties [72].

To reduce the cost, efforts, and resources, in this study, the theoretical properties of biodiesel were evaluated based on the fatty acid composition using BiodieselAnalyser software (Version 2.2) [36]. By comparing biodiesel derived from S. podzolica Zwy2-3, biodiesel from S. podzolica DSM 27192 [26], the reported sunflower biodiesel derived from vegetable oil [73], and the international biodiesel standards developed by the European Union and the United States [33,34], it was found that the main characteristics of Zwy2-3 biodiesel not only meet the international standards but also outperform DSM 27192 biodiesel and sunflower biodiesel in some indicators (

Table 5. Estimated properties of obtained biodiesel in comparison with DSM 27192 biodiesel, sunflower biodiesel, and international standards.

Properties	Unit	Zwy2-3 Biodiesel	DSM 27192 Biodiesel	Sunflower Biodiesel	EN 14214	ASTM D6751-02
IV	-	62.09	76.53	120.4	≤120	≤120
CN	-	59.29	56.09	50.54	≥51	≥47
υ	mm2/s	4.10	4.02	4.68	3.5–5.0	1.9–6.0
SV	mg/g	202.46	202.11	191.1	-	-
OS	h	35.53	14.50	2.00	>8	>3
ρ	g/cm3	0.87	0.88	0.84	0.86–0.9	0.86–0.9
HHV	MJ/kg	39.56	39.58	43.9	-	-
PP	°C	0.054	−0.82	−3.74	-	−15 to 10
CP	°C	6.33	5.53	4.57	-	−3 to 12

IV: Iodine Value, CN: Cetane Number, υ: Kinematic Viscosity (mm²/s), SV: Saponification Value (mg/g), OS: Oxidation Stability (h), ρ: Density (g/cm³), HHV: Higher Heating Value (MJ/kg), PP: Pour Point (°C), CP: Cloud Point (°C).

).

Specifically, these values, including the iodine value (62.09), cetane number (59.29), kinematic viscosity (4.10 mm²/s), oxidation stability (35.53 h), density (0.87 g/cm³), pour point (0.054 °C), and cloud point (6.33 °C), all fall within the range specified by the EU (EN 14214) and the US (ASTM D6751-02) standards [33,34]. The saponification value and higher heating value of Zwy2-3 biodiesel are 202.46 mg/g and 39.56 MJ/kg, respectively. These values are very similar to those of DSM 27192 biodiesel and sunflower biodiesel. Among all the important parameters of biodiesel, the cetane number indicates the combustion quality and autoignition potential, while the iodine value controls the cold filter clogging point and the oxidation stability [74]. The biodiesel with a higher cetane number not only has a good cold start performance and ignition performance, but also reduces the formation of white smoke [72]. In this regard, the cetane number value of Zwy2-3 biodiesel is higher than the minimum value required by the international standards, as well as higher than that of DSM 27192 biodiesel and sunflower biodiesel. This gives Zwy2-3 biodiesel a significant advantage. In addition to the cetane number value, oxidation stability is also one of the most important fuel properties that can affect the long-term storage and usage of biodiesel fuel [75]. The predicted oxidation stability value of Zwy2-3 biodiesel is 35.53, which is approximately 2.5 and 18 times higher than that of DSM 27192 (OS: 14.50) and sunflower biodiesel (OS: 2.00), respectively. The reason for this significant difference is that there are very few PUFAs in the fatty acid composition of S. podzolica Zwy2-3 under the optimal fermentation conditions. A higher oxidation stability value has also been observed in the research of Rhodococcus sp. YHY01 biodiesel, in which the oxidation stability is shown as infinity and contains only MUFAs (61.68% palmitoleic acid and 0.05% oleic acid) and no PUFAs [76]. In other words, oxidation primarily occurs due to the presence of double bonds in the chain. Raw materials rich in PUFAs are more prone to oxidation compared to raw materials rich in saturated or MUFAs [77]. According to Demirbas [78], the higher heating value (HHV) range for biodiesel falls between 39 and 41 MJ/kg, as there are no specified limits for HHV in either the ASTM D6751-02 or EN 14214 standards. The biodiesel derived from Zwy2-3 and DSM 27192 exhibited HHV ranging from 39.56 to 39.58 MJ/kg, which falls within the typical range of 39–41 MJ/kg for biodiesel. This indicates that the HHV of these biodiesel samples was superior to that of sunflower biodiesel (>41 MJ/kg).

The pour point (PP) and cloud point (CP) are two important parameters for the low-temperature application of fuel. The PP is the lowest temperature at which the cooled oil sample can flow, and the CP is the first visible temperature at which wax appears when the fuel is cooled [79]. In this study, the PP and CP of Zwy2-3 biodiesel were 0.054 and 6.33, respectively, which meet the ASTM D6751-02 standard. According to the above analysis, it is evident that the lipid extracted from S. podzolica Zwy2-3 is suitable for replacing or mixing with conventional oil in biodiesel production.

4. Conclusions

In the present study, the EHCS-ACSE was effectively utilized for the growth and lipid accumulation of S. podzolica Zwy2-3. In a single-factor experiment, it was found that the DOC had a significant impact on the growth and oil accumulation of the yeast. By one-factor optimization, the yeast was able to achieve an oil yield of 10.3 g/L. Based on this, a 4-factor 3-level RSM experiment was conducted to fully optimize the fermentation parameters of the Zwy2-3 strain. Under the optimal fermentation conditions (reducing sugar concentration: 89.44 g/L, yeast extract concentration: 3.88 g/L, shaker speed: 219 rpm, and fermentation time: 122 h), a final oil production of 11.45 g/L was achieved. This provides a significant advantage over other oil-producing yeasts that utilize the corn stover hydrolysate as a substrate for fermentation. The lipid profile of the strain Zwy2-3 was determined via GC-MS under the optimal culture conditions. It was found that the main fatty acid compositions were C16 and C18 fatty acids. The fatty acid compositions and proportions (%) in the fermentation of EHCS-ACSE, glucose, xylose, and mixed sugar (glucose: xylose, 3:1) as the substrate were as follows: oleic acid (61.84, 62.03, 60.90, and 62.55), palmitic acid (21.53, 21.56, 20.35, and 19.72), stearic acid (13.05, 11.94, 13.16, and 12.92), and linoleic acid (3.58, 4.48, 5.59, and 4.81). Furthermore, it was observed that the fatty acid composition and ratios of the strain Zwy2-3 remained consistent despite variations in the dissolved oxygen levels, simultaneous changes in the substrate and dissolved oxygen, and simultaneous changes in the substrate concentration and dissolved oxygen. This indicates that the fatty acid of the strain Zwy2-3 is highly suitable for biodiesel production. BiodieselAnalyzer (Version 2.2) was finally used to predict the properties of biodiesel based on the fatty acids obtained from the fermentation of EHCS-ACSE as a substrate. The following values were obtained: iodine value (62.09), cetane number (59.29), kinematic viscosity (4.10 mm²/s), saponification value (202.46), oxidative stability (35.53), density (0.87 g/cm³), high calorific value (39.56 MJ/kg), pour point (0.054 °C), and cloud point (6.33 °C). The predicted values of the listed biodiesel properties comply with the international standards for biodiesel. The main characteristics, such as the cetane number, oxidative stability, and iodine value, were better than those of Saitozyma podzolica DSM 27192 and sunflower biodiesel. This further demonstrates that the oil from the strain Zwy2-3 is highly suitable as a feedstock for biodiesel production. All the results suggest that S. podzolica Zwy2-3 has the potential to be developed as an industrial strain for the production of microbial oil, which can be used to generate biodiesel.