Esential oils extraction: a 24-hour steam distillation systematic methodology

Abstract Steam distillation is known to be the most prevalent method of essential oil extraction. Despite many studies on extraction methods, there is no report about the impact of distillation process duration on the yield and oil quality. A new 24-h steam distillation process for extraction of plant essential oils is presented. For improving the total yield, prolonged and continued isolation was used. A selection of plant species from Lamiaceae and Apiaceae families was subjected to direct steam distillation and essential oils were collected at different times (1, 2, 3, 6, 12 and 24 h). The analysis included either annual or perennial species monitored in terms of different harvesting time. From these studies, it is conclusively that there is no rule about appropriate extraction time, and different plants need different periods for the essential oils to achieve the desired quality or quantity of extract. Thus, extraction duration is directly dependent on what the study is conducted for.


Introduction
Essential oils (EOs) are naturally occurring volatile compound mixtures generally obtained by steam distillation, having the characteristic aroma of the plant part they come from. Generally, EO is present at low concentrations and it requires high-performance extraction techniques in order for high yields to be achieved (Tongnuanchan & Benjakul 2014). EOs are produced by different methods, including solvent extraction, supercritical fluid extraction, hydro distillation and steam distillation. Recently, ultrasound-and microwave-assisted processes have become attractive (Gómez & Witte 2001;Mohammad & Karamatollah 2008;Malekydozzadeh et al. 2012;Ranitha et al. 2014).
Plant EOs are usually the complex mixture of natural compounds, both polar and nonpolar (Masango 2005). They are composed principally of terpenoids, including monoterpenes and sesquiterpenes (diterpenes may also be present), and their related oxygenated derivatives. A variety of other molecules may also occur, such as aliphatic hydrocarbons, acids, alcohols, aldehydes, acyclic esters or lactones, and exceptionally nitrogen-and sulphur-containing compounds, coumarins and phenylpropanoid homologues. However, these molecules are extremely sensitive to heat and due to this, they are often subject to chemical changes, and some losses of volatile compounds may occur depending on the extraction method used (Alves De Barros et al. 2013).
EOs have an unexpectedly large range of applications: they have been widely used as food flavours (Burt 2004), and possessing antioxidant and antimicrobial activities, it serves as natural additives in foods and food products (Tongnuanchan & Benjakul 2014). They are known for their antiseptic (i.e. bactericidal, virucidal and fungicidal), medicinal properties and their fragrance. EOs are used in embalmment, preservation of foods and as antimicrobial, analgesic, sedative, anti-inflammatory, spasmolytic and local anaesthetic remedies (Bakkali et al. 2008).

Steam distillation as the method of EO extraction
The EO extraction method is important, in that its chemical composition is somewhat dependent on the applied practice (Xavier et al. 2011). However, the choice of each technique depends on the objective to be achieved by research. Distillation-based recovery processes such as steam or hydro distillation are preferred for the extraction of EOs from plant materials. These processes are flexible, versatile, do not generally lead to EOs decomposition and provide the possibility of operating from small (Amenaghawon et al. 2014) to huge volumes. On the other hand, these methods suffer from some disadvantages including losses of volatile compounds, long extraction times and high levels of energy consumption (Gavahian et al. 2015). However, these methods are the simplest and the equipment they require is often more available.
Steam distillation is a separation process for temperature sensitive materials like oils, resins, hydrocarbons, etc. which are insoluble in water and may decompose at their boiling point. This process, as the most widely used method for plant EOs extraction, has been actively pursued since the beginning of the 1980s (Reverchon & Senatore 1992). The fundamental nature of the process is that it enables a compound or mixture of compounds to be distilled at a temperature significantly below the corresponding individual constituent(s) boiling point(s). EOs contain substances with boiling points up to 200 °C or higher, but in the presence of steam or boiling water, these substances are volatilised at the water boiling point (100 °C) at atmospheric pressure (Rao & Pandey 2007).
Fresh or dried plant material is placed in the plant chamber of a steel apparatus and the generated steam passes through plant material thus softening the cells and letting the EO escape in vapourised form. The heating of the system must be mantained to increment the EO vapour pressure, yet not so high to destroy the plant or burn the plant and hence the EO (Babu & Kaul 2005). As soon as it is released, tiny EO droplets forms and mix with the steam (the carrier) and converge into a cooling system. The mixture condenses to form a bilayer liquid. In the majority of cases, the oil is less dense (lighter) than water, forming the top layer of the distillate and can be separated easily using proper method and instruments (Rao & Pandey 2007). It has been reported that the proportion of EO extracted by steam distillation is around 93% and the remaining 7% can be further extracted by other methods (Masango 2005). EOs isolated by steam distillation are different in composition from those naturally occurring in plants, since the steam distillation conditions cause chemical reactions to occur which result in the formation of certain artificial chemicals, called artefacts. Some of these are considered beneficial -formation of chamazulene during the chamomile oil extraction (Povh et al. 2001), while others may not be -the hydrolysis of esters like linalyl acetate to the corresponding alcohols that was observed in clary sage oils (Schmaus & Kubeczka 1984). The effects of different distillation methods on oil content and composition of aromatic plants have been previously reported (Fathi & Sefidkon 2012;Aliboudhar & Tigrine-Kordjani 2014;Wong et al. 2014;Liu et al. 2015). Other factors such as cultivation, soil and climatic conditions, and harvesting time, can also determine the composition and quality of the EO (Pereura et al. 2008). An improved method has been developed for maximum yield of EO from Celery seeds which included hydro distillation using electric, microwave as well as steam heating (Jain et al. 2003). A number of factors determine the final quality of a steam distilled EO, but the most important are time, temperature and pressure. The effects of these parameters on the yield of Eucalyptus EO were investigated, showing their significant impact, and suggesting the optimisation of these important process parameters (Kabuba & Huberts 2009). Regarding SP, the oil yield rapidly decreased after the first hour of the extraction process, increasing in the last 21 h, with the same last fraction yield as of the first one ( Figure S1). This is a very unusual profile, as most EOs are normally extracted in the first 3 h. This plant was not monitored in terms of different harvesting time. The flowering material was collected in the second half of June, furnishing a 0.038% total yield over 24 h ( Figure S7). The overall EO amount increased from 0.1 to almost 0.4 grams per kg of plant subjected to steam distillation indicating a low EO content in this species. Interestingly, the higher amount of EOs was obtained in the first and sixth fractions, accounting for about 28% and 29% of extracted oil (Tables 1 and 2). MA has been monitored for 3 months, from July to September, covering the main stages (periods before, during and after flowering). July material gave the least amount of EO, but the yield curve was specific, showing the increase after the third fraction -in the last 12 h it reached almost half of the amount from the first fraction (Tables 1 and 2). Flowering material from August and fruiting material from September, compared to each other, show no significant differences in yield curves. However, within the first hour of extraction, the amounts of EO doubled in comparison with July ( Figure S2). Cumulative yields were very similar for all 3 months (0.030-0.039), indicating that harvesting period has no impact ( Figure S7, Table  2). Analysis of the EO amount shows similar values to that of SP, but differently, the greatest percentages were obtained within 1 or 2 h of extraction indicating that in case of industrial oil production the extraction should be stopped within 2 h, in particular if harvested in August or September.

Results and discussion
MS has also been monitored for the same period (3 months). In August and September, the 24-h yields were 2.5 times larger than those of July ( Figure S7, Table 2). This showed great impact of the plant growth stage, highlighting the blooming period as the most fruitful in terms of EO yielding. In general, the highest amounts (from each month) were obtained during the first three extraction phases (1-3 h) and the last 12 h of extraction, giving curves similar to that of SP ( Figure S3). Regarding the cumulative EOs quantities, MS gave a greater yield in the first two fractions, especially in August and September accounting for more than 50% of the total amount extracted in 24 h. Interestingly, the last two fractions (12 and 24 h) are those giving the greater amount of EOs if compared to the others EOs extractions ( Figure S7). Fresh CG plant material was extracted and monitored for 4 months, from July to October, thus covering periods before, during and after flowering. The results showed different amounts for each fraction, but usually the largest parts of EOs were extracted in the first 3 or 6 h ( Table 1). The EO from July material was extracted in the first 2 h, with no important yield after. The higher percent of EO was observed in August, which can be explained by quite extreme temperatures and other ecological conditions during that period. Even the total yield after 24 h was the highest in August (0.43%), which could as well be explained by extreme environmental conditions. Yield curves for these 2 months are quite similar to those of MA ( Figure S4). Blooming material from September was characterised by even higher EO yield, not only in the first three fractions, but also between the third and sixth hour of extraction. The last one, October material, was mainly in the fruiting phase, continuing to give EO during the first 6 h of extraction, with no important yield after ( Figure S7). Total yields were more or less the same, highest being in July (Table 2). Regarding the amount of EO production, CG is among the most productive plants furnishing after 24 h from 3.4 to 4.3 grams of EOs per kilogram of harvested plant. Interestingly the amount is low sensitive to the harvesting period and most of percentage is obtained in the first fractions reaching more than 90% of extracted EO in 2 or 3 h.
RS is an annual plant and all the material was collected in the end of July, with flowering and fruiting stages mixed together. Dried RS material was extracted in the same way. As the result, we had a very unusual yield curve, with the first maximum after the first hour of extraction, and the second one between the third and sixth hour of the extraction process ( Figure S5, Table 1). Total yield over the entire extraction time is shown in Table 2. The amount of extracted EO profile is similar to those of SP and MA with the only difference that at 24 h the yield is lower than in the previous fractions. In terms of production, after 24 h more than 7 grams of EO can be gathered, being almost uniformly spread among six fractions. FV was monitored from August to October, including vegetative and reproductive stages, showing almost no differences in the first 2 months (with almost the same 24-h yields). On the contrary, a great increase, of up to five times more, in EO content was noticed in October (Table 2) than when the plant was fruiting ( Figure S6). This was quite expected since Apiaceae species fruits are characterised by the highest EO amounts. What can also be noticed is the similarity of this month's yield curve to the RS one (both included fruiting periods). From the plot in Figure S7, FV is indicated as the plant species producing the highest EO amount, particularly the material harvested in October, which after 24 h furnished almost 13 grams. Of these, 65% were obtained after 3 h extraction clearly indicating that for industrial FV oil manufacture, more than 8 grams per kilogram of plant can be easily produced.
It was also observed that the fractions changed the colour and density during the process, becoming darker and more dense, which was most noticeable in the case of CG and RS. That came as the result of accumulation of more artefacts, directly affecting the physical characteristics of the EOs. Regarding physical and chemical behaviour of the curves reported in Figures S1-S6, very likely long exposition of the plant material to boiling water is the main factor for the observed profile. In any case, since this is the very first study for such long steam distillations, deeper studies are under evaluation to explain the different chemical constitution of the EOs fractions.

Aims and objectives
According to literature, EOs are mainly obtained by hydro distillation or steam distillation apparatus for only 3-4 h (optionally 5). On the other hand, most of the data are related to some specific period of harvesting, e.g. flowering or fruiting periods. During the research on MS, it was observed that, either shortening or extending the extraction time, a different chemical composition of the oil could be obtained with some variation in the related biological activity. As a consequence, to obtain the optimal chemical composition of the oil, two main questions arise: (1) When should the plant be harvested? and (2) How long should the extraction time be?
In line with the previous studies on MS (Garzoli et al. 2015), the systematic extraction procedure was expanded and applied to additional five plant species from Lamiaceae and Apiaceae families, either annual or perennial. Deeper detail on CG 24H extraction were also recently reported (Božović et al. 2017) The aim of this study was to develop a systematic EO extraction system using steam distillation technique, in terms of different harvesting and extraction times. Having in mind that EO is made up of many distinct compounds which come together to form its aroma and therapeutic properties, it should be emphasised that some of these components are chemical susceptible structures that can be altered or destroyed by adverse environmental conditions. Longer distillation process may give more complex EO in terms of chemical composition as further artifacts are accumulated. This may have a curious effect on the physical characteristics of EO (odour, density and colour) and its associated biological activities. Therefore, we limited our study at 24 h distillation for the following reasons: (1) plant material is subjected to degeneration, (2) EOs produced after 6 h are very likely due to hydrolysis phenomena and EOs composition at higher time of extraction could be a result of some artefact, similarly to that observable for Matricaria chamomilla L. where the blue compound chamazulene is obtained from matricine degradation (Safayhi et al. 1994).

Plant material
A series of aromatic plants from Lamiaceae and Apiaceae families were selected and deemed adequate considering the research conducted worldwide so far and the results obtained in this area. Depending on the abundance of their populations, the following species were considered for investigation and validate the 24-h extraction model: MS, CG, MA and SP from Lamiaceae family, and FV and RS from Apiaceae family. Aerial parts (herba) were collected in summer and early autumn periods of year 2015. All the material was collected in a wild area about 10 km from Tarquinia city (Viterbo, Italy) except SP which is an endemic of the Western part of Balkan peninsula and was collected in the urban zone of Podgorica city (Montenegro). MS and CG oils were extracted from the fresh plant material.
Air-drying of the other collected plants was performed in a shady place for 20 days. SP and RS are summer annuals with very short vegetative and reproductive periods -they complete their life cycles within June/July. They have been monitored just in terms of different extraction time, not period of harvesting. Other chosen species are perennials -they have been monitored for three (MS, MA and FV) or 4 months (CG).

Species determination
Taxonomic identification of the chosen species was conducted according to the official European flora (Tutin et al. 1968) and the national Italian one (Pignatti 1982).

EO extraction
EOs have been isolated by direct steam distillation using a 62 L steel distillator apparatus (Albrigi Luigi E0131, Verona, Italy). With this type of steam distillation apparatus, the remaining water (a by-product), called floral water or hydrosol, is recycled after its condensation, thereby not limiting the process duration. Thus, by modifying the conventional distillation apparatus, so that water flows from the separation funnel back to the retort ( Figure S8), the loss of oil to wastewater can be reduced (Gonzalez et al. 2012).
Fresh MS and CG material (about 2.5 kg) and dried material of other investigated species (about 1.7 kg) were subjected to steam distillation, and the EOs were separated at interval time of 1, 2, 3, 6, 12 and 24 h. At each interval, the accumulated oil/water double phase was extracted three times with diethyl ether (20 mL). The organic layers were then dried on anhydrous Na 2 SO 4 , filtered and deprived of the solvent in vacuo to furnish oils.

Conclusions
Steam distillation is the most common technique which serves to isolate EO from plant material. Classical distillation procedure is usually reported to be over in 2-4 h. Herein, for the first time, is reported a systematic 24-h EO extraction procedure applied to different plant species and monitored for different harvesting periods. A selection of Lamiaceae and Apiaceae species has been analysed leading to a conclusion that no rule can be given about the appropriate duration of steam distillation process. In fact, different plant species have different EO yields, and the dynamic of oil extraction from plant material could be considered as species specific. On the other hand, harvesting period is also very important, directly affecting these parameters. It could be more related to the family than to the species. Flowering period could be considered as the best for Lamiaceae species. On the contrary, since Apiaceae fruits are usually very rich in EO, the fruiting stage could be defined as the optimal period of harvesting.
The study has also included chemical analyses of obtained EOs. The extraction method applied gave fractions that differ greatly in their chemical compositions (Section S1). Although the main characterising compounds are usually present in every fraction, variations in their amount are particularly evident between the first three fractions (up to 3 h of extraction process) and the last ones (after 12 or 24 h). Furthermore, some compounds appear only with the development of the extraction process, and gradually increase in amount, being significantly present only in the last few fractions. Concerning the period of harvest, the chemical profile of an EO has been found to be heavily influenced by this factor. In some cases, the same plant gave EOs that were chemically completely different from each other, depending on the month in which it was harvested.
In order to monitor the biological variability, EOs were assayed by means of antifungal activity. Having in mind the processes of synergism and antagonism between EO compounds, overall potential of the isolated EO fractions was evaluated. Longer distillation may give higher yields, but on the other hand, it may lead to the accumulation of more artefacts. All of that may have an effect on the physical characteristics of EO as well as on its associated biological activities. The latter concept has been clearly proved by our results for MS, CG and FV (Section S1).