Coriandrum sativum L.—Effect of Multiple Drying Techniques on Volatile and Sensory Profile

Coriandrum sativum L. is a medicinal and aromatic plant spread around the world, with beneficial properties that are well recognized. Both coriander seeds and leaves are used for pharmaceutical and flavoring purposes. Even though coriander seeds tend to be more popular, the leaves are receiving a consistently growing interest, especially because of popularization of Mexican, Northern African, and Indian cuisines. This increased attention brings about the necessity for providing the product with guaranteed quality, which will retain its valuable characteristics, even after post-harvest treatment. For this reason, it is highly necessary to determine reliable protocols for cilantro preservation, which usually include drying procedures; in order to identify the optimal drying treatments, a spectrum of drying techniques—convective, vacuum-microwave, and a combination of convective and vacuum-microwave—were evaluated. Cilantro-based dried products were examined from the perspectives of volatile organic constituent composition and sensory quality. After headspace solid-phase microextraction-GC/MS analysis and sensory tests, the results demonstrate that convective drying at 70 °C for 120 min followed by vacuum-microwave drying at 360 W and convective drying at 70 °C were the optimal drying methods for preserving cilantro aroma quality, while convective drying at 70 °C for 120 min followed by convective finishing drying at 50 °C decreased cilantro aroma quality.

Undoubtedly, fresh, unprocessed material is the most valuable source of the bioactive and aromatic compounds characteristic of cilantro. Nevertheless, in light of the complicated production and transportation chains, it is impossible to always use it as a fresh ingredient; therefore, it is extremely important to assure the guaranteed quality and safety of cilantro-based products. The most essential quality determinants for MAPs are cultivation conditions (including agronomic practices and weather conditions), harvest time, postharvest treatments (including material preservation), and plant chemotype [15,16]. Among indicated determinants, post-harvest treatment in the form of drying is the most common processing used for preservation of MAPs quality in a long term [17,18]. The latest studies focused on developing cost-effective, thermo-based drying procedures [19,20], which need to be evaluated on various plant materials [21][22][23][24][25][26][27][28]. The greatest interest is focused on convective drying (CD), microwave drying, vacuum-microwave drying (VMD), freezedrying, infrared drying, or combined methods such as convective pre-drying followed by vacuum-microwave finish-drying (CPD-VMFD).
In addition, coriander seeds and cilantro were the objects of numerous studies which considered drying proceedings and their influence on the quality of obtained dried products. Some of them were mainly focused on mathematical modelling within cost and energy efficiency evaluation of the drying process [29][30][31], while other studies included evaluation of drying influence on quality in light of volatile organic constituents (VOCs) and non-volatile constituent composition, or even the sensory properties of dried products [32][33][34]. Unfortunately, none of these studies considered a comprehensive experiment design, the intention of which would be to show a complete picture of various drying methods' influence on VOC composition and their impact on aroma sensory quality of coriander leaves. The most extensive study was designed by Pirbalouti et al., (2017) [32], since it included five drying methods and their influence on cilantro EO composition; however, sensory evaluation was not performed. On the other hand, Fathima et al. (2001) [34] provided a detailed sensory evaluation of cilantro, without testing a wide spectrum of drying methods and VOCs analysis. Therefore, the aim of this study was to investigate the influence of various drying methods on the sensory quality of cilantro and its VOC composition. For this purpose, CD, VMD, and CPD-VMFD with various parameters were applied to obtain dried cilantro-based products, which were evaluated with headspace solid-phase microextraction-GC/MS (HS-SPME-GC/MS) analysis and multiple sensory analyses.

Plant Material
Cilantro (Coriandrum sativum L.) plants were delivered by Swedeponic Polska Sp. z o.o. company (Grodzisk Mazowiecki, Poland). The used plants were of a flat-leaf, parsley-like variety, purchased as potted plants (green parts weight 30 ± 2 g). Coriander leaves, immediately after delivery, were chopped, thoroughly mixed, and split for specific purposes-(i) drying procedures, (ii) chemical analyses of fresh material, and (iii) sensory analyses of fresh material. Plant material was vacuum packed and stored at −18 • C awaiting further proceedings.

Drying Methods
Two drying methods (convective drying, vacuum-microwave drying) and their combinations were utilized [35]. In each case, 80 g of fresh cilantro samples were used. Moisture content (Mc) of the samples was determined using a vacuum dryer pre-set at 70 • C and 100 Pa (SPT-200, ZEAMIL Horyzont, Krakow, Poland). Initial moisture content of the material was Mc = 92.62%. The maximum temperature of the sample was determined after each measurement using an infrared camera (Flir Systems AB, Stockholm, Sweden), right after the removal from the dryer. All drying variants were carried out in two technological repetitions and the process lasted until the moisture content was below 10% wet basis (wb).

Convective Drying (CD)
Convective drying was performed at 50, 60 and 70 • C using the equipment engineered at the Institute of Agricultural Engineering (Wroclaw University of Environmental and Life Sciences, Wroclaw, Poland). In addition, two-stage convective drying with temperature adjustment after 120 min was performed ( Table 1). The air velocity in each case was 0.5 m/s.

Modelling of Drying Kinetics
The drying kinetics were fitted on the basis of mass losses recorded during drying by different methods. Moisture ratio (MR) was calculated according to the simplified equation: where M (t) is the moisture content at a given time and M 0 is the initial moisture content. The drying model was fitted using Table Curve 2D software (Systat Software, San Jose, California, USA) according to the experimental points obtained in the study. The good fit of the applied model was determined on the basis of the highest values of R 2 and the lowest values of root mean square error (RMSE). Preliminary studies found that the best fit was obtained when Page's model was applied: where A, k, and n are drying constants and τ is the time taken for drying.

VOC Profiling
The VOC profile of cilantro was investigated according to Łyczko  For qualitative analysis, Varian Workstation software was used, whereas for quantitative analysis, ACD/Spectrus Processor (Advanced Chemistry Development, Inc., Toronto, ON, Canada) was utilized. VOCs were identified via determination of linear retention indices (LRIs) and mass spectra comparison (Adams, 2012 [37]; NIST 17 Mass Spectral and Retention Index Library). The LRIs filter was narrowed down to ±15 points and only mass spectra matches with similarity score ≥90% were qualified as hits. The molecular mass of VOCs was confirmed by chemical ionization mode performed with isobutane as the reagent gas. Quantification was done by peak area normalization against IS peak area.

Sensory Evaluation
The sensory quality evaluation of dried cilantro was carried out, as previously reported by Łyczko et al. (2020) [36], with a trained panel of 20 panellists (age 23-52) from the Food Quality and Safety Group (Escuela Politécnica Superior de Orihuela) of the Universidad Miguel Hernández de Elche (Orihuela, Alicante, Spain). The selection of panelists was performed according to ISO standard 8586-1 [38,39]. Before the actual evaluation, panelists were involved in 3 training meetings in order to adjust the panelist's senses for required aroma attributes. Sensory analyses were performed in a tasting room with individual booths, controlled light (70-90 fc), and controlled temperature (22 ± 1 • C). Coded cilantro samples were provided for sensory analysis, during which 3 tests were performed-a ranking test, napping test, and descriptive sensory analysis. In the first place, ranking and napping tests were carried out. The ranking test was used to evaluate the intensity of dried products aroma in comparison to fresh cilantro aroma and to reduce the sample number for further steps. The reduction in samples for descriptive sensory analysis was required, since it was impossible to fully study all samples initially available in this study [40,41]. For this, 20 panelists were asked to rank the coriander samples according to perceived intensity of the attribute "fresh coriander". The panelists used the ascending rank order, in which 1 meant the least and 13 meant the most intense sample, respectively. For descriptive sensory analysis, one sample was selected from each drying treatment/type, having high intensity in the ranking test and having no measurable intensity of off-flavors. For the descriptive sensory analysis, a sensory sheet was developed containing 21 sensory attributes previously selected from the according to the scientific literature, napping test, and orientation sessions. The samples were served in 30 mL covered cups and a random block design was used for sample presentation in order to avoid biases. Each panelist was asked to smell and score the samples using an 11-point numerical scale, where 0 represented none or no intensity and 10 extremely strong, with a 0.5 increment. Additionally, the napping technique was used to group dried samples and to support the reduction in sample number for descriptive sensory analysis. Napping consisted of the position of the products on a A3 sheet of paper according to their similarities by the panelists, who were also asked to add at least 3 descriptors for each group of samples [42]. As the last step, descriptive sensory analysis was carried out by 7 judges who evaluated 4 chosen samples. In this step, the panelists evaluated the basic tastes, aroma, and flavor of dry cilantro according to the lexicon first given by Łyczko et al. (2020) [36] with slight modifications (the full lexicon is available in the Supplementary Materials).

Statistical Analysis
Statistical analysis was performed with Statistica 13.3 software (StatSoft, Kraków, Poland). The one-way analysis of variance (ANOVA) of HS-SPME-GC/MS data and descriptive sensory analysis were carried out, followed by Tukey's test (p < 0.05), and for a ranking test, Friedman's rank-sum analysis (α = 0.05), supported by Tukey's honest significance difference (p < 0.05), was used. The statistical analysis and graphic interpretation (biplot) of the napping results were carried out with XLSTAT Premium 2016 software (Addinsoft, París, France). The procedure of data handling, collected during napping, consisted of determining the position of each sample on X and Y axis and the sensory description given by each panelist. Check-all-that-apply (CATA) and multifactor data analysis (MFA) analyses were performed to generate the biplot. In all relevant cases, standard deviation (SD) was given. Figure 1 shows the drying kinetics of the cilantro samples treated by different drying methods and Table 2 shows the drying kinetics constants (A, k, n), drying times, maximum temperature of sample heating (T max ), and final moisture content (Mc wb ) of cilantro dried using different methods. It can be seen that during CD with an increase in temperature from 50 to 70 • C, the duration of drying decreased over two times (from 570 min in the case of CD50 to 270 min in case of CD70). On the other hand, application of two-term CD led to an only slightly shorter time of the process than in case of CD50, and even longer times than in case of CD60 or CD70. Furthermore, CD with temperature adjustment after 120 min resulted in higher moisture content than in the case of CD60 or CD70. A smooth transition between temperatures in two-term CD can be noticed, except for CD50/70, where a sudden drop in MR value after 120 min can be seen (Figure 1b). still higher than when only VMD was applied. The use of this method resulted in moderate final Mc of the sample as well as moderate maximum temperature of the material recorded during drying. Page's model was successfully applied in the study to describe drying kinetics on the basis of high values of determination coefficient (R 2 > 0.99 in every case) and low values of RMSE (RMSE < 0.026). This section may be divided by subheadings. It should provide a concise and precise description of the experimental results, their interpretation, as well as the experimental conclusions that can be drawn.    CD-convective drying, CPD-convective pre-drying, VMD-vacuum-microwave drying, VMFD-vacuum-microwave finish-drying.

Drying
The highest Mc wb was obtained when CD70/60 was used and the highest water reduction, and therefore the lowest Mc, was reached when CD60 was applied. As for VMD, an increase in the power of magnetrons from 240 to 480 W significantly reduced the duration of the process (from 68 min in the case of VMD 240 to 42 min when VMD 480 was used). Overall, VMD resulted in much shorter processing time comparing to convective drying. Furthermore, the maximum temperature of the sample treated by VMD increased when higher power was applied.
Combined drying resulted in shorter time of the process than in the case of CD, but still higher than when only VMD was applied. The use of this method resulted in moderate final Mc of the sample as well as moderate maximum temperature of the material recorded during drying. Page's model was successfully applied in the study to describe drying kinetics on the basis of high values of determination coefficient (R 2 > 0.99 in every case) and low values of RMSE (RMSE < 0.026). This section may be divided by subheadings. It should provide a concise and precise description of the experimental results, their interpretation, as well as the experimental conclusions that can be drawn.

Sensory Analysis
Multiple sensory analyses showed a significant divergence, regarding dried cilantro odor quality on various levels. The very first trial-the ranking test, the results of which are presented in Table 5-evaluated the intensity of the cilantro-like aroma of dried samples. Samples CPD70-VMFD (rank sum 198 points) and CD70 (sum 184 points) were recognized as those with the most intense cilantro-like aroma, while CD50/70 (sum: 100 points), VMD 480 (sum: 98 points) and CD70/50 (sum: 95 points) were pointed out as those with the weakest aroma intensity. The napping test, the results of which are illustrated in Figure 2, showed that dried cilantro samples may be divided into five groups. Samples were grouped in the napping test according to the groups established in a clustering dendrogram prepared using Pearson's correlation based on the unweighted average. The two largest groups, A and C, consisted of three and four samples, respectively, which all included the CD method. Group C, with higher CD initial temperatures, was characterized as "dry grass", "hay-like", "parsley-like", "sweet", and "chamomile", while group A, where initial CD temperatures were lower, was characterized as "green", "lacteal", "fermented grass", "wet", "fresh", "balsamic", and with "high intensity". The next groups, D and E, both contained two samples. Group D, consisting of CPD50-VMFD and CD60/70, was characterized only as "dry leaves", while group E, consisting of VMD 240 and VMD 480, was characterized as "fermented" and "musty". One sample-VMD 360-was categorized as group B, characterized as "dry", "woody", "infusion", and "mushrooms", while the last sample, CD70/50, overall was not characterized with any descriptions. Foods 2021, 10, x FOR PEER REVIEW 11 of 17 Figure 2. Napping test results for cilantro treated with various drying methods presented as biplot generated with MFA, which explains 44.25% of the variance. Samples were grouped in the napping test according to the groups established in a clustering dendrogram prepared using Pearson's correlation based on the unweighted average.
As the last sensory test, descriptive sensory analysis was performed. For this trial, due to the high number of samples, only one representative for each drying method was selected after ranking and napping techniques-CD60, CD50/70, VMD 240, and CPD60-VMFD. In the case of basic taste descriptors, samples did not show any significant differences, while in case of flavor, pungency and spiciness were the differentiating features. More significant differences were observed among aroma descriptors. As visualized in Figure 3, chamomile, hay-like, herbaceous, vegetable, and woody were the descriptors, which diversified dried cilantro in the case of aroma. CD60 treatment was the one with the strongest intensity of the above-mentioned attributes. As the last sensory test, descriptive sensory analysis was performed. For this trial, due to the high number of samples, only one representative for each drying method was selected after ranking and napping techniques-CD60, CD50/70, VMD 240, and CPD60-VMFD. In the case of basic taste descriptors, samples did not show any significant differences, while in case of flavor, pungency and spiciness were the differentiating features. More significant differences were observed among aroma descriptors. As visualized in Figure 3, chamomile, hay-like, herbaceous, vegetable, and woody were the descriptors, which diversified dried cilantro in the case of aroma. CD60 treatment was the one with the strongest intensity of the above-mentioned attributes. Foods 2021, 10, x FOR PEER REVIEW 12 of 17

Drying Kinetics
The effect of the hot air temperature during CD as well as power of magnetrons during VMD was evaluated in the study. It can be seen that with an increase in temperature during CD, the time taken for drying decreases, maximum recorded temperature increases, and the drying rate of the process accelerates, which is consistent with other studies [26,29,44].
A similar effect was seen in case of VMD, where with an increase in the power of magnetrons, the duration of the process decreases. The same phenomena were also noticed in thyme [45]. Furthermore, increased wattage resulted in an increase in the maximum temperature recorded during VMD, which is consistent with other studies [46].
During CD, the hot air flows around the sample and efficiently removes surface moisture, which is directly correlated with the high drying rate at the initial stage of the CD. Yet, as can be seen in the Figure 1a, over time, the drying rate diminished as a result of slow diffusion of the moisture from the inside of the sample. On the other hand, faster evaporation during VMD is due to the heating in the whole volume of the sample as a result of microwave application, which mitigates the limitations of CD.
The beneficial effect of combined CPD-VMD was due to the phenomena occurring during both CD and VMD, where at the initial stage the surface water is efficiently removed by CD and, then, the VMD is applied to remove the remaining water which is strongly bounded to the plant cellular structure [47]. As a result, a shorter drying time can be obtained.

Drying Kinetics
The effect of the hot air temperature during CD as well as power of magnetrons during VMD was evaluated in the study. It can be seen that with an increase in temperature during CD, the time taken for drying decreases, maximum recorded temperature increases, and the drying rate of the process accelerates, which is consistent with other studies [26,29,44].
A similar effect was seen in case of VMD, where with an increase in the power of magnetrons, the duration of the process decreases. The same phenomena were also noticed in thyme [45]. Furthermore, increased wattage resulted in an increase in the maximum temperature recorded during VMD, which is consistent with other studies [46].
During CD, the hot air flows around the sample and efficiently removes surface moisture, which is directly correlated with the high drying rate at the initial stage of the CD. Yet, as can be seen in the Figure 1a, over time, the drying rate diminished as a result of slow diffusion of the moisture from the inside of the sample. On the other hand, faster evaporation during VMD is due to the heating in the whole volume of the sample as a result of microwave application, which mitigates the limitations of CD.
The beneficial effect of combined CPD-VMD was due to the phenomena occurring during both CD and VMD, where at the initial stage the surface water is efficiently removed by CD and, then, the VMD is applied to remove the remaining water which is strongly bounded to the plant cellular structure [47]. As a result, a shorter drying time can be obtained.
The study on leaf drying revealed that the quality of leaves is highly dependent on the duration of the process. However, higher temperatures might result in the degradation of bioactive compounds [29]. Two-term CD might mitigate the negative effect of high temperature on the quality of the sample since application of higher temperature of hot air during the first stage of drying does not increase the temperature of the sample. This is due to the intensive evaporation that cools down the sample and the dried material never reaches the temperature of hot air. After 120 min, when the temperature is reduced to 60 or even 50 • C, evaporation slows down and the temperature of the sample stays relatively low. In addition, in the case of CD50/70, the temperature adjustment changed the shape of the curve, which is marked by a sudden drop in the MR after 120 min, and significantly shortened the drying process.
Page's model applied in this study was previously successfully used to describe the drying kinetics of Origanum majorana [48] and Cassia alata [47].

Volatile Organic Constituents of Cilantro and Quality of Dried Products
The majority of the most abundant VOCs identified in fresh cilantro were qualified as aliphatic aldehydes and alcohols, what corresponds with de Melo et al. (2019) [49], Nurzyńska-Wierdak (2013) [50], and Padalia et al.'s (2011) [51] findings. Nevertheless, some exceptions were found, such as high contributions of (Z)-hex-3-en-1-ol, (Z)hex-3-en-1 acetate, and limonene-17.38 ± 0.38%, 34.54 ± 2.08%, and 2.43 ± 0.22%, respectively. In addition, in comparison to previous studies on cilantro, the high percentage of linalool was surprising, since it is much more characteristic for coriander seeds; however, some studies, such as the one performed by Shahwar et al., (2012) [51], supported this result. Such results may be reasoned in various plant chemotypes, which is well-documented in the scientific literature for coriander and other MAPs such as chamomile or sweet basil [52][53][54].
In terms of dried cilantro, some earlier studies were conducted; however, none of them included such a wide spectrum of drying methods, evaluated in the light of their impact on the sensory quality of products, as in the present study. Clearly, as shown in Table 5, the highest scores in the ranking test were granted to CPD70-VMFD and CD70 products. Furthermore, those two samples were qualified in the same group (C-"dry grass", "herbal", "sweet", "hay-like", "parsley-like" and "chamomile") during the napping test. Regarding the cilantro OACs, according to Cadwallader et al. (1999) [12], (Z)-hex-3-enal and (E)-alk-2-enals are essential for the characteristic cilantro aroma. Moreover, such results were also confirmed by Eyres et al. (2005) [43]. Therefore, focusing on indicated VOCs, the CPD70-VMFD and CD70 products showed some similarities in VOC composition. Both samples showed not statistically significant differences regarding (Z)-hex-3-enal, its derivative (Z)-hex-3-en-1-ol, (E)-undec-3enal, and (E)-tetradec-3-enal contribution, which are responsible for green, cut grass, fruity, solvent-like, chemical, pungent, spicy, aldehydic, and floral aroma notes. On the other hand, the CPD70-VMFD and CD70 products differed in the case of (E)-dec-2-enal, (E)-dodec-2-enal, (E)-tridec-2-enal, and (Z)-tetradec-2-enal. A slight difference in (E)-dodec-2-enal contribution (in favor of CD70 product over CPD70-VMFD-13.95% to 11.80%) and higher difference in (E)-dec-2-enal (in favor of CPD70-VMFD product over CD70 product-9.89% to 4.17%) may be responsible for the marginally higher score of the CPD70-VMFD sample. Such an implication is supported by Tamura et al.'s (2013) [55] findings, which report that among cilantro OACs, (E)-dodec-2-enal, due to the low odor threshold, has the most significant impact on cilantro aroma quality. Admittedly, in this case, the highest scored dried product (CPD70-VMFD) had a lower contribution of (E)-dodec-2-enal than the second in the rank (CD70); however, CPD70-VMFD had a higher content of (E)-dec-2-enal, which has a strongly similar odor description to that of (E)-dodec-2-enal (Eyres et al., 2005) [43]. One of the most surprising experimental findings was that during the descriptive sensory analysis, particular aroma attributes of CD60 scored the highest, while during the ranking test, the CD60 sample was placed lower than the CPD70-VMFD and CD70 samples. Such a result supports the hypothesis that the strongest does not always mean the finest, as was proven in our earlier studies on Thai basil and lavender flowers [23,36]. Furthermore, in the napping test, CD60 was qualified into the group with some off-flavor descriptions ("lacteal", "wet", "fermented grass"), which also may be the reason for locating this sample lower in the ranking test than CPD70-VMFD and CD70.
The lowest scores in the ranking test were received by CD50/70, CD70/50, and VMD 480 products. Nevertheless, samples were qualified to groups (A and E) or closely placed next to groups (CD70/50) characterized by MAPs with off-flavor descriptions such as "fermented grass", "wet", "lacteal", "mushrooms", "fermented" or "musty". Confronting these results with the cilantro VOC profile composition, it was unambiguous that for "fermented", "musty", and "mushroom" descriptions of drying treatments consisting only of VMD, high contributions of β-pinene and terpinolene were responsible, which, according to Breheret et al. (1997) [56], have a significant role in a wild mushroom aroma. In addition, it is interesting to mention that the compound responsible for the woody notes ( Figure 3) of dried cilantro was identified as (Z)-tetradec-2-enal. There was a clear relationship between the woody notes' intensity and the contents of this compound in the VOC profile analysis.
In earlier studies by Fathima et al., (2001) [43] and Pirbalouti et al., (2017) [32], it was pointed out that the application of microwave and high temperature (up to 60 • C) during cilantro drying decreased the quality of the dried products, which seems to support the part of obtained results in the present study, in terms of the results obtained in the case of using VMD individually. Overall, the lowest scores were granted to drying methods consisting only of VMD. In addition, VMD received the most negative odor descriptions (mushroom) during the napping test. On the other hand, the previously cited studies pointed out that applying high temperatures may also be destructive for cilantro aroma quality. Nevertheless, the current results show that utilizing only CD gave satisfying results, which corresponds to Kamel et al.'s (2013) [33] findings. The differences in quality of dried cilantro by CD may be caused by many factors, such as the humidity of the environment in which the process was performed, drying air velocity, and more. The overall satisfying results of combined drying, consisting of CPD followed by VMFD, applied for herbs was earlier confirmed in Chua et al., (2019bChua et al., ( , 2019a [47,57] [32], in their study, based the evaluation of the quality of dried cilantro on EO composition, which, as was proven in our earlier studies on lavender [23,35] and Thai basil [36], is not correct for the determination of the quality for whole plants.

Conclusions
The growing demand for medicinal and aromatic plants, especially in terms of flavoring purposes, results in an urgent need to determine optimal protocols in the case of preservation. Regarding the cost and time efficiency, drying technologies are leading postharvest treatments. Earlier studies showed that the universal drying protocol for all kind of medicinal and aromatic plants is questionable, since plants' morphological differences and specific volatile organic constituent compositions result in different process requirements. In this study, the evaluation of different drying methods for the preservation of Coriandrum sativum L. leaves (cilantro) was performed. The experimental results of the cilantro aroma profile (HS-SPME-GC/MS), multiple sensory analyses of dried products (ranking test, napping test, sensory descriptive analysis), and technological aspects of drying processes show that convective drying at 70 • C for 120 min followed by vacuum-microwave drying at 360 W and convective drying at 70 • C were the optimal drying treatments guaranteeing cilantro-based products the highest quality. The aroma profile of these samples was characterized by low contents of (Z)-hex-3-enal and (Z)-hex-3-en-1-ol and intermediate contents of (E)-dec-2-enal, (E)-dodec-2-enal, (E)-tridec-2-enal, and (Z)-tetradec-2-enal. In contrast, the lowest rated samples, mainly including only vacuum-microwave drying and convective drying at two temperatures, namely vacuum-microwave drying with a power of 480 W or convective drying at 50 • C for 120 min followed by convective finishing drying at 70 • C, were characterized by high contents of β-pinene and terpinolene.