1. Introduction
Considering the high biological potential of the essential oils, it is advisable to look for new efficient methods for their extraction. In this study, the raw plant material has been treated with ultrasound action towards increasing the efficiency of the hydrodistillation process, which means an increase in the amount of obtained essential oil per unit of weight of the raw material.
The mechanism of extraction of the essential oils by the ultrasound process is based on two main physical phenomena such as the diffusion across the cell walls and membranes and mechanical destruction of plant cell walls by pressure waves and cavitation, and next, rinsing the contents of cells [
1]. From an economic standpoint, the application of ultrasound-assisted hydrodistillation of essential oil is beneficial not only on the grounds of more effective extraction, but also saving energy, solvent, and time consumption [
2]. Due to the shortening of the exposure time of the raw material to high temperatures, this process avoids the disadvantages of essential oil conventional extraction, such as the formation of the by-product by decomposition of thermolabile and heat-sensitive compounds [
3].
The objective of this study was to improve the yield of the essential oil extraction using distillation assisted by sonication of raw material. There is no information in the literature about the effect of sonication of celery seeds (
Apium graveolens L.,
Apiaceae Lindl.) on the hydrodistillation of essential oil as well as the obtained product quality. The only research on ultrasound-assisted extraction was conducted by Zor et al. (2017), which was proved as a fast and efficient method for deriving alcoholic extracts from celery seeds, a rich source of antioxidants [
4]. The celery and its essential oil are known for their therapeutic, medical, and industrial attributes [
5]. Due to antioxidant activities and antimicrobial effects against bacteria, yeast, and moulds, the celery seed oil may be used as alternative natural food preservatives, functional foods, and nutraceutical ingredient [
6]. The presented research places special emphasis on environmental action, so the waste celery seeds which did not meet quality criteria and without any sowing value were used.
To maximize the process efficiency, the experiment was planned according to a design of experiment (DOE) approach with the use of the Taguchi method to optimize parameters of the seeds sonication process for savings in time and energy consumption. The qualitative and quantitative composition, physicochemical parameters, biological activity, and aroma profile of the essential oil obtained from raw material sonicated in optimal parameters were analyzed.
The hydrodistillation process, to which sonication of raw material has been applied, is characterized by the high increases in the amount of essential oil per unit of weight of the raw material relative to hydrodistillation. The yield of celery seed essential oil has increased by 48.3%.
3. Materials and Methods
3.1. Plant Material
The experimental part of this research was performed with the use of the dried celery seeds (Apium graveolens L.) that showed no germination power and were industrial waste with no utility and economic value. Raw material was donated by the manufacturer as part of a cooperation with the Polish seeds company Przedsiębiorstwo Hodowlano-Nasienne W. Legutko Sp. z o.o (Jutrosin). A voucher specimen is stocked in the manufacturer’s premises (Jutrosin, Poland).
The seeds were grounded in laboratory grinders (Basic A11D, IKA, Staufen, Germany) for 20 s.
3.2. Taguchi Experimental Design Approach
The experiment has been planned according to a DOE approach with the use of the Taguchi method. The selected control factors and their applicable working levels were: time of sonication 5, 20, 50 min; pulse range 0.1, 0.5, 1.0; power control 20, 60,100 %, as well as the content of water 350, 700, 950 mL. The content of water shall mean the water in the sample treated with ultrasounds. The pulse range was: setting 1 meaning continuously switched on, setting 0.6 meaning power discharge 0.6 s, and pause 0.4 s. In this research, the L9 orthogonal array was used, which has nine rows corresponding to the number of tests, with four columns at three levels. Each level of each parameter has been tested three times, which means that the number of required experiments for this module was 27. In the case of the DOE approach, randomness is desired and should be maintained when possible. In relation to this rule, all the trials and repetitions were unbiased and performed in a completely randomized order. Next, the ANOVA was performed (using the Statistica 13.1 software). The Taguchi’s loss function the larger the better was adopted as the best possible loss function for maximizing product yield. In the case of this function, the best quality standard is infinity, and the higher the actual value (the yield of the essential oil), the better. Based on the analysis of Eta values, the best sets of input parameters have been determined and the optimal parameters of the optimized process have been chosen. Last, run confirmation test with optimum conditions have been done in triple repetition. The essential oil obtained from seeds sonicated in optimal conditions was used for further research as the study sample.
The theoretical yield of essential oil was calculated from the formula for Eta in using loss function: where n is the number of iterations and yi is the value of the output variable (the essential oil yield).
3.3. Application of Ultrasounds
The sonication of 100 g of grounded celery seeds (Seeds Company W. Legutko, Poland) was conducted in sonicator UP400S (400W, 24kHz, Hielschier, Germany) with the sound protection box SB1-16 (Hielschier Ultrasonics, Teltow, Germany), electronic timer for controlling the acoustic irradiation duration (Hielschier, Germany), and the titanium sonotrode Tip H3 type (Hielschier Ultrasonics, Teltow, Germany) for transmitting the ultrasound into the liquid. The application of ultrasound was performed according to the L9 orthogonal array containing applicable working levels for each control parameter according to
Table 6. The tests were carried out with care to make sure that the sonotrode was always drowned in the same depth.
3.4. Hydrodistillation Process
The extraction of essential oils was performed by the hydrodistillation process with the use of a modified Deryng apparatus [
29]. The 1000 mL of distilled water per 100 g of celery seeds was used. The control sample constituted the essential oil from untreated celery seeds (without ultrasound processes). The yield of essential oil was calculated as an average of three hydrodistillation processes. The percentage increase of yield of essential oil was calculated according to the formula:
, where:
YUAH means the yield of essential oil obtained by ultrasound-assisted hydrodistillation, and
YHD means the yield of essential oil obtained by hydrodistillation.
3.5. Physicochemical Parameters
Physicochemical characteristics have been described with an optical rotation α (a polarimeter Autopol IV, Rudolph Research Analytical, Hackettstown, NJ, USA), a refractive index nD20 (an automatic refractometer JI57 Donserv, Rudolph Research Analytical, Hackettstown, NJ, USA), and a density (an automatic densitometer DDM2910, Rudolph Research Analytical, Hackettstown, NJ, USA). Each measurement was repeated three times.
3.6. Gas Chromatography-Mass Spectrometry with Flame Ionization Detection (GC-FID)
Gas chromatography-mass spectrometry analyses were carried out using a Trace GC Ultra gas chromatograph and a DSQ II mass spectrometer (Thermo Electron Corporation, Beverly, MA, USA) with an Rtx-1 column (length 60 m, internal diameter 0.25 mm, film thickness 0.25 mm, Restek Corporation, Bellefonte, PA, USA). A flow divider (an MS-Column Flow Splitter, SGE Analytical Science, Melrose Park, Australia) collected of signals concurrently from two detectors (FID, MSD, Thermo Fisher Scientific, Waltham, MA, USA). The split ratio was 1:20. The temperature program was from 50 °C (3 min) to 300 °C (30 min), at a gradient of 48 °C (21 min). The temperature of the injector (SSL) and detector (FID) was 280 °C and 300 °C, respectively. The helium flowing at a constant pressure of 200 kPa was used as the carrier gas. The mass spectrometer ionization energy was 70 eV, and the ion source temperature was 200 °C. A full scan was conducted in the mass range from 33 to 420. The analysis was repeated three times. The NIST Library, Wiley 8th edition, and the Adams 4th edition were used.
3.7. Similarity Analysis
The essential oil from ultrasound-processed in optimal conditions seeds has been compared in term of chemical composition with the essential oil from celery seeds obtained by hydrodistillation without ultrasound-process. Analysis was performed by the Mann–Whitney test with a significant level of 0.05 and by NIR and MIR spectroscopy.
3.8. Near-Infrared (NIR) and Mid-Infrared (MIR) Spectroscopy
Oil samples were scanned in the infrared spectrometer Nicolet iS50 FT-IR (Thermo Fisher Scientific, Waltham, MA, USA) that allows to generate spectra in two ranges: MIR and NIR. MIR analyses were carried out with a DTGS KBr detector, IR light source, and KBr beamsplitter. The range of scans was from 4000 to 400 cm−1 with a resolution of 4.00 cm−1. The samples were placed in IR Sample Cards (Real Crystal, US) with KBr glass (9.5 mm aperture) and all spectra were accumulated from 32 scans. During NIR measurements there were used: InGaAs detector, white light, and CaF2 beamsplitter. The range of scans was from 12,000–4400 cm−1 with a resolution of 8.00 cm−1. All spectra were accumulated from 32 scans. The samples were placed in the middle of the borosilicate glass tubes (6 × 50 mm) made by Kimble Glass, US. All calculations were made using a commercial analysis software: OMNIC 9.3.30 and TQ Analyst, v. 9.4.45 (Thermo Fisher Scientific, Waltham, MA, USA).
3.9. Antimicrobial Activity Assay
The microbes which were originated from the American Type Culture Collection (ATCC) and the Center of Industrial Microorganisms Collection of the Institute of Fermentation Technology and Microbiology, Lodz University of Technology, Poland, WDCM 105 (LOCK) were examined. Strains tested were both Gram-positive (Bacillus subtilis ATCC 6633, Staphylococcus aureus ATCC 6538) and Gram-negative bacteria (Escherichia coli ATCC 8739, Pseudomonas aeruginosa ATCC 15442), as well as fungi (the yeast Candida vini LOCK 0008, the moulds Penicillium expansum LOCK 0535, Aspergillus niger LOCK 16404). Microbial strains were sub-cultured on the medium plate count agar (PCA, Merck, Darmstadt, Germany) for bacteria or potato dextrose agar medium (PDA, BTL, Warsaw, Poland) for fungi, and next, activated through double passaging in Tripticase Soy Broth (TSB, Biocorp, Warsaw, Poland) for bacteria or Sabouraud Dextrose Broth (SDB, BTL, Warsaw, Poland) for fungi. Conditions of incubation were 30 °C (E. coli, B. subtilis), 37 °C (S. aureus, P. aeruginosa) for 24h, and 25 °C for 72 h for fungi. Inoculum of 24-h cultures of each strain was prepared in standard 0.85% sodium chloride solution and adjusted to the final concentration of approximately 107 CFU/mL.
Antimicrobial activity assay was carried out using an impedimetric method (Bactometer M64, bioMerieux, Craponne, France). The celery seed essential oil was diluted in ethyl alcohol (pure P.A., Avantor Performance Materials Polan, Gliwice, Poland) in a ratio of 1:1. Each sample included 0.1 mL of the standardized microorganism inoculum and the essential oil at the examined concentration within the range from 50 to 500 μg/mL. Next, each Bactometer well was filled until a final volume of 1 mL was obtained with the use of the growth medium: general purpose medium (GPM, bioMerieux, Craponne, France) for B. subtilis, S. aureus, and P. aeruginosa, coliform medium (CM, bioMerieux, Craponne, France) for E. coli, and yeast and mould medium (YMM, bioMerieux, Craponne, France) for A. niger, P. expansum, and C. vini. In the negative controls, instead of the essential oil, 0.5 μg/mL of novobiocin for bacteria or 0.2 μg/mL of cycloheximide for yeast and moulds were added. Positive controls did not contain essential oil, but only 0.1 mL of cell standardized suspension in 0.9 mL of the appropriate medium. The microorganisms were incubated for 72 h at their optimal growth temperatures (30 °C for E. coli, B. subtilis, 37 °C for S. aureus, P. aeruginosa, and 25 °C for fungi). Next, the tested strains were checked for their viability by streaking on the PCA medium. Plates were incubated for 72 h for bacteria and 120 h for fungi at the optimal growth temperatures. The results which were the mean value of three measurements were presented as MIC and MBC.
3.10. Antioxidant Activity
Radical scavenging activity was examined using the DPPH assay. First, the 1-μM DPPH (Sigma-Aldrich, Hamburg, Germany) solution in methanol (pure p.a., Chempur, Poland) was prepared. Next, the 200 μL of DPPH was added to 100 μL of the solution of essential oil in methanol at concentrations of 2.5, 5, 10, 20, 50, and 100 g/L, respectively. The analysis was performed using a 96-well polystyrene plate (Nest Biotechnology, Wuxi, China). The samples were incubated in the dark at room temperature for 30 min. As the reference, a methanol solution of Trolox ((±)-6-hydroxy-2,5,7,8-tetramethylchromane-2-carboxylic acid, Sigma-Aldrich, Saint Louis, MO, USA) was used. The absorbance of the DPPH radical was spectrophotometrically analyzed at a wavelength of 517 nm (Modular Multimode Microplate Reader TriStar2 S, Berthold Technologies, Oak Ridge, TN, USA). The results were shown as the percentage of inhibition of the DPPH radicals calculated according to the formula: , where, A0 is the absorbance of the control sample (DPPH), and A1 is the absorbance of the sample with the essential oil. The results were expressed as the arithmetic mean value of the three consecutive measurements with a standard deviation value.
3.11. Aroma Profile
The aroma profile of essential oil was performed by untrained analytical teams of five males and five females. A hedonic ten-point scale test was applied. After inhalation, the panellists marked the intensity scale of aroma from 1 to 10 points, where 0 is none or not perceptible intensiveness, and 10 is strong intensiveness. The distinguishing features as fatty, mossy, green, camphoraceous, herbal, citrus, fruity, earthy, flowery, fresh, medicinal, mushroom, sweet, spicy, and woody were evaluated. Sensory hallmarks were according to El-Zaeddi et al. (2016) with modifications [
30]. The results were shown as the average of all measurements.