**5. Maceration for improving melon spirit**

The effect of maceration in increasing and improving aroma intensity in a melon (*Cucumis melo* L.) spirit was studied.

Spirits and Liqueurs from Melon Fruits (*Cucumis melo* L.) 191

The concentration of total acids was lower in the macerated spirit, especially because of pronounced decreases in hexanoic acid, octanoic acid, and decanoic acid. Lauric acid levels increased considerably following maceration with both the sliced melon flesh and the

Furfural is produced by acid hydrolysis or during the heating of polysaccharides containing hexose or pentose fragments, the highest concentrations being found in alcoholic beverages (1-33 mg/kg). This compound is currently allowed, since it is naturally present in fruits and other foodstuffs (European Union, 2002). Furfural and hydroxyMethyl furfural increased substantially in the spirit macerated with the sliced melon flesh as compared to the control

The concentration of phenolic compounds was higher in the macerated spirit. This was particularly true for 4-Methyl-guaiacol, which increased appreciably in the spirit macerated with the sliced melon flesh. In wine this compound has a negative impact on aroma at levels higher than 4 mg/L but a positive impact at concentrations between 1.2 and 2.4 mg/L

Benzaldehyde, a carbonyl compound, decreased in the macerated spirit. Concentrations of this substance higher than 2-3 g/L are related to a bitter almond flavour in wines (Blaise &

Increased levels of acetoin (3-hydroxy-2-butanone) act as an indicator of oxidation of 2-3 butanediol during ageing (Jouret & Cantagrel, 2000). The increase was higher in the spirit

In the colour attribute determinations, the tristimulus values, to derive the rectangular (L\*, a\*, b\*), cylindrical (L\*, C\*, h\*) and the chromaticity (x, y, z) coordinates were used (CIE, 1986). The coordinates a\*, b\*, h\*, and C\* yielded the two-dimensional (CIELAB) colour space, where h\* is the angle formed with the a\* axis and C\* is the distance to the origin. Sample MSP1Y had the highest colour intensity (C\*) values, and in terms of chromaticity values, and MSP18 was similar to sample MF1Y. Variation in the angle (h\*) was minimal, with angles in the range of 90.14 to 94.46. All the samples fell in the region pale, with macerated sample MSP1Y exhibiting the highest colour intensity and thus falling closer to the region of lightness. Thus, maceration affected both lightness (L\*) and colour intensity (C\*), with the unmacerated spirit and the spirit macerated with the sliced melon flesh being

Figure 2 represents the differences in colour (E\*) between the macerated spirit and the control spirit and also lists the gradations in visual perception according to Schmidhofer, et al., (1994). The longest maceration time (1 year) exhibited higher E\* values than the shorter maceration times, and the three macerations carried out with the seeds+placenta exhibited

The results of sensorial analysis thus indicate that maceration did have an influence on the final product and that the panelists perceived distinct differences between the samples.The distillate macerated for 11 days (MF11) and the unmacerated control batch were significant

The preference tests failed to yield any preference for either the macerate spirit or the control sample. This result is ascribable; on the one hand, to the diverse make-up of the taste

higher E\* values than the macerations carried out with the sliced melon flesh.

seeds+placenta, possibly as a result of hydrolysis of Ethyl laurate.

spirit, but values were still within allowable limits.

(Etiévant, et al., 1989).

macerated with the sliced melon flesh.

paler than the spirit macerated with the seeds+placenta.

Bruns, 1986) .

at the 99.9-% level.

Double-distilled melon spirit (alcohol content 56 % v/v) was macerated using different melon parts (pieces of melon flesh, skinned and then sliced (MF) (72.5 % w/v), and seeds+placenta (MSP) (23.5 % w/v). Contact times were 8 days, 18 days, and one year. The solid percentage was selected based on literature data (Xandri, 1958) and personal communications. After steeping, the distillate was chilled and filtered and the solids pressed and 0.3 g/L bentonite was added for clarification.

Sugar levels in the double-distilled spirit macerated with sliced melon flesh (MF) were four to five times higher than in the spirit macerated with the seeds+placenta (MSP), possibly due to the higher solids content (72.5 % in the MF batches compared with 23.5 % in the MSP batches). Total sugar levels stayed practically constant throughout maceration, though after maceration for one year saccharose levels fell while glucose and fructose levels rose by 80 and 60 %, respectively, perhaps because of hydrolysis of the disaccharide.

Dry matter (DM) was much lower (0.10 g/L) in the unmacerated control batch (C) than in the macerated spirit. The spirit macerated with sliced melon flesh for one year (MF1Y) had the highest values (26.95 g/L), compared with the spirit macerated with the melon seeds+placenta also for one year (MSP1Y), (21.53 g/L), because of the larger amount of macerated solids.

The alcohol content (% v/v) decreased in the macerated double-distilled spirit, the decrease being more pronounced in the macerates made with the sliced melon flesh, possibly attributable to absorption by the melon tissue and the release of larger amounts of water from the pieces of melon flesh.

Respect to the major volatiles, Methanol levels decreased with maceration but in all cases were within the limits set by the legislation currently in force (European Union, 1989) The higher alcohols (HAs), namely, 1-Propanol, 2-Methyl-1-Propanol, 2-Methyl-1-Butanol, and 3-Methyl-1-Butanol, and the esters Ethyl lactate and Ethyl acetate fell appreciably during maceration. These decreases were much more pronounced in the spirit macerated with the sliced melon flesh (MF8 and MF18) than in the spirit macerated with the seeds+placenta (MSP8 and MSP18). Ethyl butyrate was not detected. 1-Butanol, which has a very pungent and heavy aroma (Nykänen & Suomalainen, 1983), was present in lower concentrations in the macerated spirits, and theirs decrease could have a positive influence on the aroma of the final spirit.

Levels of Benzyl alcohol, 1-Hexanol, t-3-nonenol, c-3-Hexen-1-ol, 2-Phenylethanol, and geraniol were higher in the batches macerated with the sliced melon flesh and the seeds+placenta. 2-Phenylethanol imparts an aroma of roses (Nykänen & Suomalainen, 1983), and the concentration of this compound was three to four times higher than in the control spirit. At high levels 1-Hexanol and c-Hexen-1-ol may contribute herbaceous aromas, while at low levels they may exert a positive influence (Bertrand, 1975, Orriols, 1992, 1994).

On the whole, the ester content in the macerated spirits were lower than in the control spirit, with substantial decreases in Ethyl caproate, Ethyl caprylate, Ethyl decanoate, and Ethyl laurate. At the same time, Ethyl palmitate decreased in the spirit macerated with the sliced melon flesh but remained constant in the spirit macerated with the seeds+placenta because of the quantities that leached out of the melon seed (Al-Khalifa, 1996) . Ethyl linoleate and Ethyl linoleoate were also related to the seeds and increased considerably in the spirit macerated with the seeds+placenta.

Double-distilled melon spirit (alcohol content 56 % v/v) was macerated using different melon parts (pieces of melon flesh, skinned and then sliced (MF) (72.5 % w/v), and seeds+placenta (MSP) (23.5 % w/v). Contact times were 8 days, 18 days, and one year. The solid percentage was selected based on literature data (Xandri, 1958) and personal communications. After steeping, the distillate was chilled and filtered and the solids pressed

Sugar levels in the double-distilled spirit macerated with sliced melon flesh (MF) were four to five times higher than in the spirit macerated with the seeds+placenta (MSP), possibly due to the higher solids content (72.5 % in the MF batches compared with 23.5 % in the MSP batches). Total sugar levels stayed practically constant throughout maceration, though after maceration for one year saccharose levels fell while glucose and fructose levels rose by 80

Dry matter (DM) was much lower (0.10 g/L) in the unmacerated control batch (C) than in the macerated spirit. The spirit macerated with sliced melon flesh for one year (MF1Y) had the highest values (26.95 g/L), compared with the spirit macerated with the melon seeds+placenta also for one year (MSP1Y), (21.53 g/L), because of the larger amount of

The alcohol content (% v/v) decreased in the macerated double-distilled spirit, the decrease being more pronounced in the macerates made with the sliced melon flesh, possibly attributable to absorption by the melon tissue and the release of larger amounts of water

Respect to the major volatiles, Methanol levels decreased with maceration but in all cases were within the limits set by the legislation currently in force (European Union, 1989) The higher alcohols (HAs), namely, 1-Propanol, 2-Methyl-1-Propanol, 2-Methyl-1-Butanol, and 3-Methyl-1-Butanol, and the esters Ethyl lactate and Ethyl acetate fell appreciably during maceration. These decreases were much more pronounced in the spirit macerated with the sliced melon flesh (MF8 and MF18) than in the spirit macerated with the seeds+placenta (MSP8 and MSP18). Ethyl butyrate was not detected. 1-Butanol, which has a very pungent and heavy aroma (Nykänen & Suomalainen, 1983), was present in lower concentrations in the macerated spirits, and theirs decrease could have a positive influence on the aroma of the final spirit.

Levels of Benzyl alcohol, 1-Hexanol, t-3-nonenol, c-3-Hexen-1-ol, 2-Phenylethanol, and geraniol were higher in the batches macerated with the sliced melon flesh and the seeds+placenta. 2-Phenylethanol imparts an aroma of roses (Nykänen & Suomalainen, 1983), and the concentration of this compound was three to four times higher than in the control spirit. At high levels 1-Hexanol and c-Hexen-1-ol may contribute herbaceous aromas, while at low levels they may exert a positive influence (Bertrand, 1975, Orriols,

On the whole, the ester content in the macerated spirits were lower than in the control spirit, with substantial decreases in Ethyl caproate, Ethyl caprylate, Ethyl decanoate, and Ethyl laurate. At the same time, Ethyl palmitate decreased in the spirit macerated with the sliced melon flesh but remained constant in the spirit macerated with the seeds+placenta because of the quantities that leached out of the melon seed (Al-Khalifa, 1996) . Ethyl linoleate and Ethyl linoleoate were also related to the seeds and increased considerably in the spirit

and 60 %, respectively, perhaps because of hydrolysis of the disaccharide.

and 0.3 g/L bentonite was added for clarification.

macerated solids.

1992, 1994).

macerated with the seeds+placenta.

from the pieces of melon flesh.

The concentration of total acids was lower in the macerated spirit, especially because of pronounced decreases in hexanoic acid, octanoic acid, and decanoic acid. Lauric acid levels increased considerably following maceration with both the sliced melon flesh and the seeds+placenta, possibly as a result of hydrolysis of Ethyl laurate.

Furfural is produced by acid hydrolysis or during the heating of polysaccharides containing hexose or pentose fragments, the highest concentrations being found in alcoholic beverages (1-33 mg/kg). This compound is currently allowed, since it is naturally present in fruits and other foodstuffs (European Union, 2002). Furfural and hydroxyMethyl furfural increased substantially in the spirit macerated with the sliced melon flesh as compared to the control spirit, but values were still within allowable limits.

The concentration of phenolic compounds was higher in the macerated spirit. This was particularly true for 4-Methyl-guaiacol, which increased appreciably in the spirit macerated with the sliced melon flesh. In wine this compound has a negative impact on aroma at levels higher than 4 mg/L but a positive impact at concentrations between 1.2 and 2.4 mg/L (Etiévant, et al., 1989).

Benzaldehyde, a carbonyl compound, decreased in the macerated spirit. Concentrations of this substance higher than 2-3 g/L are related to a bitter almond flavour in wines (Blaise & Bruns, 1986) .

Increased levels of acetoin (3-hydroxy-2-butanone) act as an indicator of oxidation of 2-3 butanediol during ageing (Jouret & Cantagrel, 2000). The increase was higher in the spirit macerated with the sliced melon flesh.

In the colour attribute determinations, the tristimulus values, to derive the rectangular (L\*, a\*, b\*), cylindrical (L\*, C\*, h\*) and the chromaticity (x, y, z) coordinates were used (CIE, 1986). The coordinates a\*, b\*, h\*, and C\* yielded the two-dimensional (CIELAB) colour space, where h\* is the angle formed with the a\* axis and C\* is the distance to the origin. Sample MSP1Y had the highest colour intensity (C\*) values, and in terms of chromaticity values, and MSP18 was similar to sample MF1Y. Variation in the angle (h\*) was minimal, with angles in the range of 90.14 to 94.46. All the samples fell in the region pale, with macerated sample MSP1Y exhibiting the highest colour intensity and thus falling closer to the region of lightness. Thus, maceration affected both lightness (L\*) and colour intensity (C\*), with the unmacerated spirit and the spirit macerated with the sliced melon flesh being paler than the spirit macerated with the seeds+placenta.

Figure 2 represents the differences in colour (E\*) between the macerated spirit and the control spirit and also lists the gradations in visual perception according to Schmidhofer, et al., (1994). The longest maceration time (1 year) exhibited higher E\* values than the shorter maceration times, and the three macerations carried out with the seeds+placenta exhibited higher E\* values than the macerations carried out with the sliced melon flesh.

The results of sensorial analysis thus indicate that maceration did have an influence on the final product and that the panelists perceived distinct differences between the samples.The distillate macerated for 11 days (MF11) and the unmacerated control batch were significant at the 99.9-% level.

The preference tests failed to yield any preference for either the macerate spirit or the control sample. This result is ascribable; on the one hand, to the diverse make-up of the taste

Spirits and Liqueurs from Melon Fruits (*Cucumis melo* L.) 193

diffusion out of the melon pieces and the larger proportion of macerated substrate. Analysis of each of the sugars separately revealed that due to hydrolysis of the disaccharide by the fruit enzymes, the saccharose content gradually decreased from day 8 on, while the glucose

According to main volatile components concentrations of esters and higher alcohols, of great

macerated with seeds and placenta and with melon pieces + seeds and placenta. Neither n-Butanol nor Ethyl butyrate were detected in any of the batches, a highly positive finding in that these compounds are deemed to produce off flavours when they are present in distillates. For the colour measurements, the value of L\* ranges from 0 for black to 100 for white and was extremely high for the spirits considered here, between 91.9 and 95.5, the highest value being recorded for the unmacerated distillate, which also had the lowest Chromaticity (C\*). C\* values increased with substrate content and maceration time, while lightness (L\*) decreased with

All the batches can be grouped close together in the greenish yellow quadrant, though in the region closer to yellow. The macerated with melon pieces were greenish in colour and those with seeds+placenta were yellow-orange in colour because of the higher proportion of

Finally, the results of the preference test carried out on all the macerated batches as part of the sensory analysis indicated a slight though non-significant preference by panelists for the

At the same time, production of an authentic melon liqueur was addressed. Three melon liqueurs were prepared (Hernández-Gómez et al., 2009). The alcohol content was adjusted to


Sensory analysis of the three liqueurs was based on preference tests in the same conditions as the sensory analyses described above. Nevertheless, none of the liqueurs was statistically preferred at the 95 % significance level, though the macerated melon liqueur received the highest scores. The product was novel and unrelated to the panelists' prior experience, and as a result while some of the panelists preferred the unmacerated distillate on account of its "clean" aroma, others preferred the macerated melon liqueur on account of its fruity aroma

Melon in the form of juice or paste without skin, constitutes an appropriate fermentation substrate, the sugars being consumed during fermentation to produce alcohol yields in

, were higher in the batches

organoleptic importance (Baro & Quiros-Carrasco , 1977)

maceration time and was unrelated to the amount of macerated substrate.

280 ml/L (28% v/v), and the saccharose content was 100 g/L (CAE, 1997).



and fructose contents rose.

carotenoids.

.

**6. Melon liqueur** 

kirsch over the raspberry and melon spirits.

The resulting liqueurs were thus:

alcohol content

and melon flavour.

**7. Conclusion** 

MSP8: seed+placente 8 days macerate; MSP18: seed+placente 18 days macerate; MSP1Y: seed+placente 1 year macerate; MF8: flesh 8 days macerate; MF18: flesh 18 days macerate; MF1Y: flesh 1 year macerate

Fig. 2. Colour differences (ΔE\*) between the macerated double-distilled spirit samples and the unmacerated control spirit (C) and perception levels as per Schmidhofer (1994).

panel and, on the other, to the fact that some panelists preferred a "flatter" spirit with traditional distillation aromas, while others preferred a spirit with an intensely "fruity" melon aroma.

#### **5.1 Optimization of maceration process**

A trial run at industrial scale is the final step in new product development. Maceration time and substrate are important factors, and the latter may include pieces of fruit of different sizes, seeds, placenta, or skins depending on the type of product being manufactured. For that preliminary trial macerations using melon (Hernandez-Gomez et al., 2005b) yielded positive results in terms of extraction of colour and typical melon aromas, suggesting that macerated spirit could be used to produce an authentic liqueur reminiscent of the fruit employed in production.

Different proportions of fruit were tested based on the results of previous maceration trials, taking unmacerated distillate and adding:


Maceration time was 16 d, and the sugar content was measured at 0, 4, 8, 12, and 16 days of maceration.

Respect to total sugars (fructose, glucose, and saccharose) at the end of the maceration period, the total sugar content in the macerated with melon pieces and seeds+placenta was double that of the batches macerated with seeds+placenta only basically as a result of diffusion out of the melon pieces and the larger proportion of macerated substrate. Analysis of each of the sugars separately revealed that due to hydrolysis of the disaccharide by the fruit enzymes, the saccharose content gradually decreased from day 8 on, while the glucose and fructose contents rose.

According to main volatile components concentrations of esters and higher alcohols, of great organoleptic importance (Baro & Quiros-Carrasco , 1977) , were higher in the batches macerated with seeds and placenta and with melon pieces + seeds and placenta. Neither n-Butanol nor Ethyl butyrate were detected in any of the batches, a highly positive finding in that these compounds are deemed to produce off flavours when they are present in distillates.

For the colour measurements, the value of L\* ranges from 0 for black to 100 for white and was extremely high for the spirits considered here, between 91.9 and 95.5, the highest value being recorded for the unmacerated distillate, which also had the lowest Chromaticity (C\*). C\* values increased with substrate content and maceration time, while lightness (L\*) decreased with maceration time and was unrelated to the amount of macerated substrate.

All the batches can be grouped close together in the greenish yellow quadrant, though in the region closer to yellow. The macerated with melon pieces were greenish in colour and those with seeds+placenta were yellow-orange in colour because of the higher proportion of carotenoids. .

Finally, the results of the preference test carried out on all the macerated batches as part of the sensory analysis indicated a slight though non-significant preference by panelists for the kirsch over the raspberry and melon spirits.
