Evaluation of the Optimum Harvesting Maturity of Makhwaen Fruit for the Perfumery Industry

Harvesting makhwaen (Zanthoxylum myriacanthum Wall. ex Hook. f) fruits at the appropriate maturity is the key to ensure that the essential oil quality meets the need of consumers. In common practice, the fruits are usually harvested when their pericarps start to open and fruits are greenish-red in colour depending on the judgment of the farmers. This leads to inconsistencies in the essential oil quality. This research aims at characterising the aromatic profiles of makhwaen essential oil thereby for consumers to choose the quality that best fits their need and eventually identify the optimum harvesting index of the fruits. The effects of maturity states viz. 15, 36, 45 and 60 (MK15-60) days after fruiting on chemical and sensorial quality of the essential oil was evaluated. Fruit sizes ranged from ~3.3–3.7 mm and fruits appeared to dry initially when they reached 45 days. Essential oils were extracted from these fruits after they had been oven dried (60 ◦C) to the same moisture content, about 10%. The chemical profiles of the essential oil were different. L-limonene and sabinene were evaluated as key components for good quality essential oil and they were found to be higher in MK45 and MK60 (max = 139.04 μg·mL−1 and max = 146.27 respectively). NIR spectral patterns of pure extracted oil for every different harvesting time (of every different harvesting time of MK60 and MK36) were similar. Sensorial descriptive analysis by semi-trained panellists defined six terms for characteristics (woody, citrus, herb, sweet, pine and spice). The panels provided the highest rating score (15 numeric scale) of citrus and pine scents at MK45, while sweet and woody aromas were the highest at MK15. The spice scent was maximum when the fruits were harvested at 36 days after fruiting. From this study we suggest that the optimum harvesting index for the distinctive aroma of essential oil ought to be at late harvesting (45–60 days after fruiting). The findings contribute to our understanding of the harvesting maturity, which can also provide significant benefit for the perfumery industry, i.e., the optimum harvesting stage that imparts the essential oil with highest quality.


Introduction
The Zanthoxylum genus of the Rutaceae family comprises~250 plant species [1]. These aromatic crops can be found in tropical and temperate regions [2]. In Thailand, most Zanthoxylum species are known for not only uses as spices, but also to cure many ailments. For examples, Zanthoxylum acanthopodium DC. and value) was reported as the percentage of essential oil from dry plant material [16]. The oil was dried over anhydrous sodium sulphate and kept at 4 • C until analysis (usually within 3 days).

Chemical Composition Analysis
Essential oil samples (2 µL at the dilution of 1% v/v in dichloromethane with 0.003% w/v toluene as an internal standard) were injected in a split mode (1:20) [10]. The gas chromatography mass spectrometry (GC-MS) analysis was performed on using a Bruker -scion 436 GC a Rxi 5Sil MS (30 m × 0.25 mm; 0.25 µm film thicknesses). The temperature program oven temperature was held for 2 min at 60 • C and was enhanced to 150 • C at 3 • C·min −1 . Then, temperature enhancement was programmed up to 270 • C at 5 • C·min −1 and held at this temperature for 15 min. Other operating conditions were helium as the carrier gas with a flow rate of 1.1 mL·min −1 ; injector and detector temperatures were 300 • C, and split ratio, 1:50. Mass spectra were taken at 70 eV. The mass spectra and retention indices of essential oil components were identified by comparison to published literature as presented in the MS computer library, supplied by Adams [24]. The standard solution of C 8 -C 20 n-alkane (Fluka ® Analytical, Munich, Germany) in dichloromethane was also used for the calculation of retention indices (RI) [23]. The identification of the volatile compositions was by comparison with mass spectra in NIST 05.L and NIST 98.L libraries with >70% similarity. The compounds were confirmed by their RI as well as those from the literature [9]. The amount in µg·mL −1 of essential oil was calculated as relative to that of toluene internal standard.

NIR Analysis
The pure essential oils extracted were scanned to analyse the basic physical properties of essential oils from the makhwaen essential oil using a FT-NIR spectrometer (MPA FT-NIR Series Bruker Optics, Stuttgart, Germany), averaging 32 scans per spectrum at 8 cm −1 resolution to assess the makhwaen essential oil samples. Scanning was conducted in reflectance mode comparing the MPA integrated sphere and liquid probe against the Matrix-F with a fibre-coupled measurement head operating two tungsten halogen 12 V, 20 W light sources [25].

Sensory Analysis
A descriptive sensory analysis was carried out to profile the major odour characteristics of the essential oil. The sensory analysis was performed by five semi-trained panellists (2 male, 3 female, ages 20-37). During training sessions which lasted for 30 hours, the panellists smelled the different essential oil samples and discussed odour attributes [26]. From the discussion session and by using GC-MS analysed result, the descriptive terms were identified [10,27]. The odour attributes (Table 1), including citrus, herb, pine, spice, sweet and woody attributes, as well as general intensity and desirability were evaluated. The descriptive analysis was performed along a scoring line ranging from 0 (none) to 15 (very strong) in triplicates, and two samples were presented for each analysis [26,28].

Statistical Analyses
The data was statistically analysed using a comparison of the means of yield for essential oils evaluated by Duncan's multiple range test at 95% confidential level [30]. A principle component analysis (PCA) was used to identify the main sources of systematic variation in the sensory descriptive data [31].

Fruit Size, Weight and Colour
The size and weight of harvested makhwaen fruits of different maturity stages were measured ( Table 2). Fruit sizes and weight ranged from ≈3.4-3.7 mm and 0.016-0.034 g, and is consistent with the report from Suksathan et al. [21] who reported that the size of fruit was about 3-4 mm and 0.010-0.030 g. Harvest time was important for the accumulation of chemical substances in essential oils, as immature harvesting may cause the yield and accumulation of essential substances to be lower [32]. Variation of fruit colour from 15-60 days after initial fruiting are varied ( Figure 1). It can be observed that the colour of the fruits altered from light greenish (MK15) to reddish (MK45) and finally to brownish (MK60 or over mature). However, in some species of Zanthoxylum, such as Zanthoxylum bungeanum Maxim. [1] and Z. armatum [33], the colour of the fruits is clear when fruits are over mature.

Statistical Analyses
The data was statistically analysed using a comparison of the means of yield for essential oils evaluated by Duncan's multiple range test at 95% confidential level [30]. A principle component analysis (PCA) was used to identify the main sources of systematic variation in the sensory descriptive data [31].

Fruit Size, Weight and Colour
The size and weight of harvested makhwaen fruits of different maturity stages were measured ( Table 2). Fruit sizes and weight ranged from ≈3.4-3.7 mm and 0.016-0.034 g, and is consistent with the report from Suksathan et al. [21] who reported that the size of fruit was about 3-4 mm and 0.010-0.030 g. Harvest time was important for the accumulation of chemical substances in essential oils, as immature harvesting may cause the yield and accumulation of essential substances to be lower [32]. Variation of fruit colour from 15-60 days after initial fruiting are varied ( Figure 1). It can be observed that the colour of the fruits altered from light greenish (MK15) to reddish (MK45) and finally to brownish (MK60 or over mature). However, in some species of Zanthoxylum, such as Zanthoxylum bungeanum Maxim. [1] and Z. armatum [33], the colour of the fruits is clear when fruits are over mature.

Chemical Compounds
Forty-four volatile compounds were detected using GC-MS (Table 3). The chemical profiles of the essential oils from makhwaen fruits at different maturities were variable. Based on the results of GC-MS analysis, the main components of different harvesting time could be identified in the following sequence: MK15; sabinene (85.81 µg·mL −1 ), methyl (1-methylethyl) benzene (49.34 µg·mL −1 ) and β-phellandrene (45.22 µg·mL −1 ) essential oil, MK36; limonene (139.04 µg·mL −1 ), sabinene (118.89 µg·mL −1 ) and β-phellandrene (28.48 µg·mL −1 ) essential oil, MK45; limonene (140.56 µg·mL −1 ), sabinene (115.48 µg·mL −1 ) and L-phellandrene (52.41 µg·mL −1 ) essential oil and MK60; sabinene (146.27 µg·mL −1 ), limonene (135.64 µg·mL −1 ) and β-phellandrene (125.30 µg·mL −1 ) essential oil, respectively. The dominant component in MK60 was sabinene (146.27 µg·mL −1 ) and MK45 was L-limonene (140.65 µg·mL −1 ). Sabinene was also the major component in the essential oil of other Zanthoxylum species, including that found in Zanthoxylum rhoifolium L. and Z. myriacanthum [29,34]. Other literature suggests that L-limonene is the important volatile compound from the essential oil Agriculture 2019, 9, 78 5 of 10 of fresh Z. myriacanthum var. pubescens fruits [35], as well as Zanthoxylum schinifolium S.et Z., Z. bungeanum and Zanthoxylum piperitum (L.) DC. [9,36,37]. For the assessment of the essential oil quality for these particular species, sabinene and limonene as the representatives of woody and citrus aromas are then used as evaluating indicators [29]. Zhang et al. [1] noted that fruit maturity may influence the chemical compositions of the essential oil. The PCA between the volatile composition and the harvesting stage revealed two clustering groups. The first group (MK15 and MK60) had the dominant β-phellandrene representing a minty Agriculture 2019, 9, 78 6 of 10 aroma ( Figure 2). As with the experimental work of Jiang and Kubota [26], they reported that Japanese pepper (Z. piperitum) contained β-phellandrene as the main substance that provided a minty aroma. The second cluster was of the MK36 and MK45 with L-limonene (citrus aroma) as the most distinctive compound. This group represented the citrus aroma, for use as a flavouring agent and is found in the food flavour and fragrance aroma ingredient industry [36]. The optimum harvesting index of makwhaen fruits is therefore suggested to be 45 to 60 days after fruiting, as indicated by the greater contents of sabinene and limonene.

NIR Analysis
According to the NIR spectra analysis, oil samples were dominated by overtones and different combinations of C-H stretching and bending vibrations occurring between 9090-5555 cm −1 (1600-1800 nm) and 4545-4000 cm −1 (2200-2500 nm), respectively ( Figure 3). An NIR spectrum scan of makhwaen essential oil of different fruit maturities (MK15, MK36, MK45 and MK60) absorbed light at wavelengths of 8750-8000, 6500-7400, 6250 and 6000-5500 cm -1 , respectively, therefore illustrating similar light transmission patterns. Essential oil containing multiple chemical compositions usually presents a much more complex peak pattern than that of the pure chemical [18]. Moreover, NIR spectrum patterns are generally used to distinguish the different types of essential oils from different plant species [15]. For instance, there was a similarity in the NIR pattern of fresh and dried lavender essential oil, but when compared with the tea tree essential oil, the spectrum peaks were different [38]. Our results illustrated that even though there were variations in chemical compositions from the essential oils of different matured fruits, there was still the similarity in NIR pattern suggesting that these essential oils were of the same plant species with the corresponding chemical finger prints.

NIR Analysis
According to the NIR spectra analysis, oil samples were dominated by overtones and different combinations of C-H stretching and bending vibrations occurring between 9090-5555 cm −1 (1600-1800 nm) and 4545-4000 cm −1 (2200-2500 nm), respectively ( Figure 3). An NIR spectrum scan of makhwaen essential oil of different fruit maturities (MK15, MK36, MK45 and MK60) absorbed light at wavelengths of 8750-8000, 6500-7400, 6250 and 6000-5500 cm −1 , respectively, therefore illustrating similar light transmission patterns. Essential oil containing multiple chemical compositions usually presents a much more complex peak pattern than that of the pure chemical [18]. Moreover, NIR spectrum patterns are generally used to distinguish the different types of essential oils from different plant species [15]. For instance, there was a similarity in the NIR pattern of fresh and dried lavender essential oil, but when compared with the tea tree essential oil, the spectrum peaks were different [38]. Our results illustrated that even though there were variations in chemical compositions from the essential oils of different matured fruits, there was still the similarity in NIR pattern suggesting that these essential oils were of the same plant species with the corresponding chemical finger prints.

Sensory Analysis
The terms selection from GC-MS results, following descriptors viz. citrus, herb, pine, spice, sweet and woody, were selected ( Table 4). The citrus aroma is also a primary aromatic scent in many Zanthozylum spp. essential oils including that of Z. piperathum [36], Z. myriacanthum [9], Z. bungeanum [37] and Zanthoxylum xanthoxyloi [39]. From the results, the panellists gave citrus and pine scents a high rating score for MK45, while sweet and woody aromas were the highest scents at MK15. The harvesting at 36 days after fruiting of makhwaen fruit showed essential oil at the maximum value of spice scent. The panellists could significantly tell the different among the samples for woody, sweet and pine scents (Table 4). However, the descriptive scoring in citrus, herb and spice aroma was 8750-8000, 6500-7400, 6250, 6000-5500 cm-1 respectively, therefore, illustrating similar light 193 transmission patterns. Essential oil containing of various chemical compositions usually presents 194 much more complex peak patterns than that of pure chemical [18]. Harvesting maturity may affect 195 the quality and quantity of essential substances in essential oils. For example, changes of NIR spectra 196 pattern were observed in olive oil harvested from different maturity of "Gemlik" olive "Gemlik " 197 olive [39]. Our results illustrated that even there were variation chemical compositions from the  The terms selection of panellists from GC-MS results, following descriptors viz., citrus, herb, 205 pine, spice, sweet and woody were selected ( Table 4). The citrus aroma is also the primary aromatic 206 scents in many Zanthozylum spp. essential oil including that of Z. piperathum [36], Z. myriacanthum 207 [9], Z. bungeanum [37] and Z. xanthoxyloi [40]. From the results, the panellists described the high 208 rating score of MK45 was the citrus and pine scents, while sweet and woody aromas were the highest

Conclusions
The main volatile components of makhwaen essential oil were sabinene, L-limonene and β-phellandrene. The physical characteristics of makhwaen essential oil from different harvesting times using an NIR method illustrated similar light transmission patterns. The semi-trained panellists could distinguish the scent of essential oil from makhwaen fruit using the six attributes of  Mean scores (n = 5) for each attribute within a column with different superscript letters (a,b) are significantly different at α = 0.05 using Duncan's multiple comparison test.

Conclusions
The main volatile components of makhwaen essential oil were sabinene, L-limonene and β-phellandrene. The physical characteristics of makhwaen essential oil from different harvesting times using an NIR method illustrated similar light transmission patterns. The semi-trained panellists could distinguish the scent of essential oil from makhwaen fruit using the six attributes of citrus, herb, pine, spice, sweet and woody. Sensory information of essential oil from makhwaen fruits at 36 and 45 days after its initial fruiting had the highest pine aroma, and the citrus scent was dominant in all treatments. This report provides the patterns of aromatic profiles that can be used for future product development of the natural product industries.