Comparison and optimization of conventional and ultrasound‐assisted solvent extraction for synthetization of lemongrass (Cymbopogon)‐infused cooking oil

Abstract The lemongrass plant, which is widely cultivated in Asia, Australia, and Africa, has been reported to have many significant health benefits such as antimicrobial, insecticide, anticancer, fight fever, and disinfection. Therefore, it is an added benefit to have lemongrass compounds in cooking oil. This study was aimed to compare the conventional (CSE), and ultrasound‐assisted solvent extraction (UASE) for citral compounds from lemongrass (Cymbopogon) leaves and to optimize the best extraction method using the response surface methodology (RSM) and ANOVA. RSM design of experiments using three types of cooking oils; palm oil, sunflower oil, and corn oil. The effect of three independent variables, which are temperature (48.2–81.8°C), extraction time (4.8–55.2 min), and solvent to leaves ratio (5.3–18.7), was investigated. The characterization of lemongrass‐infused cooking oil was evaluated by Fourier transform infrared spectroscopy (FT‐IR), Gas Chromatography‐Mass Spectrometry (GC‐MS) and Scanning Electron Microscopy (SEM) analysis for confirmation of the citral compound extraction. This extraction process is optimized using Response Surface Methodology (RSM) for producing the lemongrass‐infused cooking oil. After optimization, the UASE process gives 1.009 × 106 maximum citral area for palm oil and 1.767 × 106 maximum citral area for sunflower oil. CSE process only can give 2.025 × 105 and 2.179 × 105 citral area in the GC‐MS spectrum for palm oil and sunflower oil respectively. For both the UASE and the CSE, the optimum operating conditions are 81.8°C of extraction temperature and 55.2 min of extraction time except for lemongrass‐infused palm oil in the CSE process with 45 min extraction time. The optimum solvent to leaves ratio varies from 5.3:1 to 12.9:1. This study found that corn oil cannot be used as a solvent to extract lemongrass‐infused cooking oil due to the insignificant changes and no citral peak. The lemongrass (Cymbopogon)‐infused palm oil and sunflower oil extracted using the UASE have a higher maximum citral area than the CSE process.


| INTRODUC TI ON
Lemongrass is a native and widely cultivated in tropical and subtropical climates of Asia, Australia, and Africa and is usually used as an ingredient in cooking (Singh et al., 2014). An extraction process can produce lemongrass oil from its leaves. An ancient India has used it to fight fever and infection (Al-shaer, 2006). Lemongrass oil is characterized by having a significant component of citral, which contains a mixture of tran-citral and cis-citral (Weisheimer, 2008). It was reported to have many advantages such as antimicrobial activities, insecticide activity, and anticancer (Avoseh et al., 2015). Hence, various extraction process of lemongrass oil may give many advantages to pharmaceutical industries.
Many methods can extract volatile oil from different parts of a plant such as leaves, fruit, flower, and stem (Falcão et al., 2012).
Extraction methods such as solid-liquid extraction, also known as Soxhlet, distillation, enfleurage, and maceration, were traditionally used (Kostova et al., 2010). However, there new extraction methods.
such as Accelerated Solvent Extraction (ASE), Ultrasound-Assisted Solvent Extraction (UASE), Microwave-Assisted Extraction (MAE) and Supercritical Fluid Extraction (SCFE), had been introduced (Carlson et al., 2001;Falcão et al., 2012;Kaur & Dutt, 2013;Lim et al., 2017;Parniakov et al., 2015;Petigny et al., 2013;. The new extraction technologies can shorten the extraction time, reduce solvent consumption, increase pollution prevention, and higher consent for thermolabile constituents (Shams et al., 2015). The SCFE method is an extraction technique using fluids in elevated conditions above their critical point of temperature . Carlson et al. (2001) extracted the lemongrass (Cymbopogon citratus) essential oil using dense carbon dioxide at 23-50°C and 85-120 bar. The requirement for high pressures in the SCFE increases the cost compared to conventional liquid extraction. The usage of carbon dioxide as a solvent in the SCFE is non-polar and has limited dissolving power, which cannot always be used as a solvent on its own, particularly for polar solutes.
The UASE method is cheaper than other advanced extraction techniques, and its operation is much easier (Wang & Weller, 2006). This method induced cavitation bubbles and accelerates the release of organic compounds contained within the plant body. Thus, the yield of oil obtained can be increased (Parniakov et al., 2015). The design of green and sustainable extraction methods of natural products is currently a hot research topic in the multidisciplinary area of applied chemistry, biology, and technology (Chemat et al., 2012).
Hence, a green solvent can be used as alternative solvents or innovatory plant resources and eliminate petroleum-based solvents to ensure high quality extracted products (Li et al., 2014). In conjunction, many researchers used extracted products as value-added or additives in food industries. Thus, various types of cooking oil were used so that the oil base used in cooking is full of nutrients. This study was aimed to compare the conventional (CSE) and ultrasound-assisted solvent extraction (UASE) for citral compounds from lemongrass (Cymbopogon) leaves and to optimize the best extraction method using the response surface methodology (RSM) and ANOVA using three types of cooking oils; palm oil, sunflower oil, and corn oil. This paper focuses on examining the optimum conditions for synthetization of lemongrass (Cymbopogon)-infused cooking oil and the information derived is particularly important for scaling up the operation.

| Materials
All chemicals which were utilized in this study were of analytical grade. Distilled water was consumed in the preparation of solutions. The lemongrass (Cymbopogon) leaves were collected from a residential area of Kota Samarahan, Sarawak, Malaysia. Cooking oils such as sunflower oil, corn oil, and olive oil were purchased from the consumer market.
To extract the lemongrass oil, an ultrasonic cleaning bath and heating bathtub were used. Moreover, GC-MS and FT-IR were used for chemicals analysis purposes. Other than that, to analyse the oil's physical properties, Portable Density Meter DMA 35 and Atago 3850 Refractometer were used.

| Experimental design
Response surface methodology (RSM) based on the circumscribed central composite design (CCD) was employed to study the effect of the three independent variables; time, temperature, and solvent to lemongrass leaves ratio ( Lim et al., 2017). Each variable was scrutinized at five following levels: −1, −α, 0, +α, and +1, as tabulated in Table 1. These three independent variables were studied in a multivariate study with 20 experimental runs. As described in Equation 1, the empirical quadratic model explains the behavior of the extraction process.

Selected independent variables
Symbol Unit The response surface fittings were generated using MATLAB version R2019b to enable a surface fitting analysis of the response over an entire range of varying factors with 20 number of experimental runs. Data were analysed using variance (ANOVA) analysis, and the effective treatment means were separated by the least significant difference (LSD) at a 95% confidence level. The relationships between the citral area in the GC-MS chromatogram and all parameter were determined using regression analysis. All statistical analysis was performed using Minitab Statistical Software.

| Extraction procedures
The extraction procedures were carried out using both CSE and UASE methods to optimize the extraction operating conditions.
The lemongrass leaves have undergone two processes: ultrasoundassisted solvent extraction (UASE) and conventional solvent extraction (CSE). The extracts were aliquoted and analysed using instrumental analyses.

| Conventional solvent extraction
The experiments were conducted for the conventional solvent extraction using the lemongrass leaves, which were cut and weighed according to the fixed parameter in a 250 ml beaker. One hundred milliliter of solvent was measured and mixed with the weighed lemongrass leaves. The sample was then blended and subjected to treatment in an indirect heating bath with the time and temperature established in the experimental design. The extract and leaves mixture was filtered on a sieve to separate the leaves and extracted product. The extracted product was then stored in the dark and under room temperature (24°C) until used.

| Ultrasound-assisted solvent extraction
The extraction of lemongrass oil was performed in an ultrasound bath (Ultronique, Q 5.9/40A, Eco-Sonics), with the power of 176 W and ultrasound frequency of 100 kHz. In each experimental run, fresh lemongrass leaves were cut and weighed according to the 250 ml beaker's fixed parameter. Next, 100 ml of solvent was meas-

| RSM model development
Three types of solvents (palm oil, sunflower oil, and corn oil) were used to examine the effects of extraction solvent on citral compound using both the CSE and UASE. The multivariate study was (1) When the extraction time is higher, the time for the citral infused into the solvent is expanding. Hence, the amount of extracted lemongrass oil is increased.  Moreover, Figure 2c shows that the S/L ratio of 8 and below gives a higher mean of citral area. It is also demonstrated that the mean of citral area is rising as the duration time of extraction getting big, which insinuates that the extraction duration time is proportional to the time for the citral compounds to infuse into the solvent.
The optimum condition of the UASE process for palm oil is unfolded in Table S1. The estimated optimum operating condition is at the temperature of 81.8°C for 55.2 min with a solvent to leaves ratio of 5.3:1, which produced a maximum citral area of 1.009 × 10 6 mAu.
min citral area in the GC-MS spectrum.

Figures 3 and 4 show the response surface and contour plots
for the effects of sunflower oil as solvent using CSE and UASE respectively. The temperature versus time graph in Figures 3a and 4a shows that the citral area's mean increases as the temperature rises.
However, the 15 min duration of UASE extraction decreases over F I G U R E 1 Effects of palm oil as solvent using CSE. The response surface and contour plots of: (a) citral area versus temperature and time; (b) citral area versus temperature and solvent to lemongrass leaves ratio; and (c) citral area versus time and solvent to lemongrass leaves ratio the increasing temperature (Figure 4a). The temperature versus solvent to leaves ratio graphs (Figures 3b and 4b) show that the lower the solvent to leaves ratio gives a higher mean of citral area. This is due to the concentration of the citral is decreasing as the solvent to leaves proportion increasing.
Moreover, the time versus solvent to leaves graph Figure 4c displays that the solvent to leaves ratio of 8 gives a higher mean of citral area. Besides that, it is delineated that the mean of the citral area increases as the duration of extraction increases. When the extraction duration is higher, the time for the citral infused into the cooking oil is rising. Hence, the amount of extracted lemongrass oil is increased.
It is estimated that at 81.8°C, 55.2 min and 5.3:1 of solvent to leaves ratio, the optimum condition of the CSE process using the sunflower oil give a maximum citral area which is 2.179 × 10 5 GC-MS spectrum citral area. The optimum condition of the UASE process for sunflower oil as laid out in Table S1 is Table 4. The significance and suitability of the optimized conditions were then studied using analysis of variance (ANOVA). Statistical significance of each effect, including interaction terms, linear and quadratic impact, was validated by comparing the mean squared against estimated experimental error. Depending upon the degree of freedom (df) involved, F-ratio can be calculated (the mean squared error to the pure error). With a confidence level of 95%, F-ratio significance is evaluated using the p-value column. In this column, when the value is lower than 0.05, the effect is significant. Tables 3 and 4 show that all three parameters significantly affect the extraction process. Other than that, R 2 represents the proportion of the total variability explained by the regression model. It is a measure for the amount of response variation defined by the variables and will always increase when a new term is added to the model. The R 2 adj is an adjusted form of R 2 for several terms in the model. The R 2 and R 2 adj values are relatively high and are at an acceptable range.

| Optimized conditions
The optimized parameters for UASE and CSE processes are tabulated in Table 5. This study found that the UASE process gives five times higher maximum citral area than the CSE process. The optimum temperature is at 81.8°C, and the optimum extraction time is 55.2 min except for lemongrass-infused palm oil after the CSE process (45 min). The optimum solvent to leaves ratio is varied from 5.3:1 to 12.9:1.

| Characterization of the infused oils
To study the oil content, three types of analysis were done which are Scanning Electron Microscopy (SEM), Fourier Transform Infrared (FT-IR) Spectroscopy and Gas Chromatography-Mass Spectrometry (GC-MS). The lemongrass's morphological characteristics, before and after CSE and UASE processes, were illustrated in Figure 5. On the leaf's upper surface in Figure 5a, the body consists of prickle hair known as trichomes (Kaur & Dutt, 2013) and fixed shape Parenchyma cell, which is both in the circle. Due to CSE extraction's high temperature, the trichomes were ruptured, and the Parenchyma cells were shrunken compared to the fresh lemongrass leaf (Figure 5b).
The trichomes that contain the metabolite liquid are the source of lemongrass oil. Whenever the trichomes break, the lemongrass oil is produced, and this figure demonstrates that the extraction process occurred. The lemongrass leaf's cell and trichomes that underwent the UASE process ( Figure 5c) were ruptured and abraded, which might be due to the cavitation bubble induced by the ultrasonicator collapsing on the surface of the lemongrass leaf. This finding is consistent with the work as reported by Petigny et al. (2013) and Parniakov et al. (2015). Figure 6 shows the FT-IR spectrum of extracted oil samples attained in the wavenumber region between 4,000 and 600 cm −1 .
The result from FT-IR is tabulated in  (Wahab et al., 2015). The peak at 1,541.12 cm −1 indicates stretching of C=C of the alkenes group. The peak at 1,456.26 cm −1 is observed due to C-C stretching, which is in the aromatics group. FT-IR analysed each sample for all multivariate study. However, there is no significant change in the functional group after CSE and UASE were carried out for all cooking oil types. This is because all of the samples still contain the same functional group, organic functional groups. Other than that, it also shows that the component in the cooking oils did not change significantly.
Studies from GC-MS revealed that both CSE and UASE processes produced infused oil that has similar components. Figure S1 shows the GC-MS spectrum before the extraction process (virgin oil) and after the extraction process using palm oil, sunflower oil and corn oil.
From the GC-MS spectrum comparison between the virgin oil and

| CON CLUS IONS
This study found that the UASE process gives five times higher maximum citral area than the CSE process. The optimum temperature is at 81.8°C, and the optimum extraction time is 55.2 min except for lemongrass-infused palm oil after the CSE process (45 min). The optimum solvent to leaves ratio is varied from 5.3:1 to 12.9:1. The SEM analysis shows that the lemongrass leaf consists of hair like that containing metabolism of the plant, which is also known as an essential oil. After going through the extraction process, the trichome was ruptured, thus produced the lemongrass oil. This study concludes that ultrasound-assisted could assist the extraction process by induced cavitation bubbles and accelerates the release of organic compounds within the plant body. Thus, oil production is increasing, and the yield of oil production will be high. The FT-IR Spectroscopy shows that the saturated aliphatic and aromatics group is present in the alkenes' infused oil. These entire groups give a significant peak in the FT-IR spectrum. The

GC-MS analysis reveals the additional component in the infused
oil. This study discovers that neral and geranial are the significant components added from the lemongrass.