Evaluation of novel green walnut liqueur as a source of antioxidants: Multi-method approach

Abstract

In this study, a novel green walnut liqueur (GWL) based on green walnuts, as the main ingredient, with the addition of fruits, aromatic plants, chocolate and honey, was produced at the pilot-scale. Antioxidant activity (AO) of the obtained GWL was determined using multiple AO assays in parallel, standard spectrophotometric (FC, DPPH and FRAP) and recently developed electrochemical ones, HydroxoPerhydroxoMercury(II) Complex and Mercury Reduction Antioxidant Power, and compared to similar commercial alcoholic beverages. Characterization of the GWL in terms of volatile and polyphenolic components was performed using SPE–GC–MS and HPLC–DAD–MS/MS, respectively. Sensory quality assessment was performed by experts in the field of sensory analysis of alcoholic beverages, using a scoring method. According to all AO assays applied, AO activity and total phenolics content of GWL were superior in comparison to commercial spirits considered. The volatile fraction of GWL was mainly composed of eugenol, α-terpineol and vanillin, while the most prevalent phenolic compounds were gallic (5.054 mg/L) and chlorogenic acid (1.307 mg/L) and flavonoids such as catechine (0.882 mg/L), quercetin (0.499 mg/L) and its sugar-conjugated derivatives, quercetin 3-O-glucoside (0.774 mg/L) and quercetin-3-O-rhamnoside (0.614 mg/L). Gallic acid is the major contributor to total AO activity, especially determined by DPPH and FRAP, followed by catechine, quercetin and chlorogenic acid. Among the terpenoids, contribution of eugenol to total AO activity is estimated as the highest. Excellent sensory quality (18.52 of maximum 20 scores) was ascribed to GWL. Consequently, the presence of bioactive compounds and high AO activity of GWL, in addition to high sensory quality score, indicates a high market potential for this high-value product.

Introduction

Fruits are plant-derived products that can be consumed in its raw form or after processing and conversion into various fruit-based products. Several fruit groups including temperate, tropical and subtropical are fundamental to all life and add value to earth's biodiversity. Due to being a rich source of bioactive and nutritive compounds, fruits are important components of a healthy diet (Serçe et al. 2010; Senica et al. 2019; Gecer et al. 2020).

Walnut fruits are highly beneficial for human health due to the presence of ω-3 fatty acids, vitamin E, plant sterols, and dietary fibers (Sánchez-González et al. 2017). Besides, the highest polyphenols content and antioxidant (AO) activity of walnuts among various nuts were reported earlier (Yang et al. 2009; Vinson and Cai 2012). The content of polyphenols in walnuts is influenced by ripeness of fruits, being the highest in June, with a decreasing trend through the process of ripening (Alamprese et al. 2005; Alamprese and Pompei 2005; Stampar et al. 2006). Polyphenols are known to have a protective role against many diseases including cardiovascular and metabolic disorders such as diabetes, various inflammation-related diseases and cancer, and exhibit neuroprotective and antiaging effects (Sánchez-González et al. 2017).

Walnut liqueur is a dark brown, bitter and appealing beverage, often served as an aperitif or being used as a tonic and digestive aid with beneficial effects against stomach trouble and digestion inconveniences (Alamprese et al. 2005). It is traditionally produced and sold in many countries worldwide. Consequently, there is a large number of recipes for the production of GWL, both for industry and homemade production. The most commonly known version of GWL is Nocino, coming from the Emilia Romagna region in Northern Italy. Similar drinks have been produced across Europe, whereas Serbian walnut liqueur, usually with the brandy as an extraction medium, is called Orahovača. Besides unripe walnuts, various herbs and spices can be used for flavor enrichment, including coffee (Slovenia), and citrus fruits (Croatia).

The astringency and bitterness of walnut liqueur are related to phenolic compounds (Jakopič et al. 2007). The influence of process variables on phenolic content and composition of green walnut liqueur was studied. Alamprese and Pompei (2005) showed that temperature and length of steeping in ethanol had little effect on the phenolic composition of the liqueur. The total phenolic content of walnut liqueur was found to increase with increasing ethanol concentration (Jakopič et al. 2007). The antioxidant activity of GWL was shown to be in good correlation with the total phenol content that according to some authors remained unchanged during the storage (Alamprese et al. 2005).

In this study, novel GWL of unique composition, based on green walnut fruits and additional ingredients, such as fruits, aromatic plants, chocolate and honey, was produced at a pilot-scale. The main aim was to assess AO activity of obtained GWL, as well as to compare it with similar commercial spirits popular on the Serbian market. In that purpose, commonly used spectrophotometric (FC, DPPH, and FRAP) and recently developed DC polarographic assays based on the decrease of the anodic limiting current of HydroxoPerhydroxoMercury(II) Complex (HPMC) formation (Sužnjević et al. 2011) (HPMC assay) and the cathodic current of Hg(II) reduction in alkaline solution (Sužnjević et al. 2015) (Mercury Reduction Antioxidant Power (MRAP) assay) were applied in parallel. Screening of the qualitative composition of volatile components, responsible for the flavour and aroma of the final liqueur, was performed by SPE–GC–MS, while polyphenolic content was analyzed by HPLC analysis. Finally, the sensory quality of GWL was assessed by experts, and compared to popular commercial drinks, belonging to a similar group of spirits.

Materials and methods

Chemicals and commercial spirits

Standards (HPLC-grade ≥ 98%) of quercetin, catechin, quercetin 3-O-glucoside, quercetin-3-O-rhamnoside, naringin, naringenin, apigenin, luteolin, kaempferol, rutin and phenolic acids (gallic, protocatechuic, p-hydroxybenzoic, chlorogenic, caffeic, p-coumaric, and ferulic acid), as well as Trolox (6-hydroxy-2,5,7,8-tetramethyl chromane-2-carboxylic acid) and TPTZ (2,4,6-tripyridyl-S-triazine) were obtained from Sigma-Aldrich, St Louis, USA. DPPH (2,2-diphenyl-1-picrylhydrazyl) was produced by Fluka AG, Buch, Switzerland. Folin–Ciocalteau reagent, sodium carbonate, sodium acetate trihydrate, acetic acid, hydrochloric acid, sodium hydroxide, acetonitrile and formic acid (MS grade) were obtained from Merck, Darmstadt, Germany. Medical-grade hydrogen peroxide solution of 35% v/v (volume per volume) (Belinka, Ljubljana, Slovenia) and corn ethanol 96% v/v (Alpis doo, Kovin, Serbia) were used. Water for standard solutions and dilutions was ultrapure water (TKA Germany MicroPure water purification system, 0.055 μS/cm). Syringe filters (13 mm, PTFE membrane 0.45 μm) were purchased from Supelco (Bellefonte, PA, USA).

Various herbal spirits of different origin, either domestic or imported, were purchased from local Serbian markets: Underberg (44% v/v of alcohol), Bitter 54 (35% v/v of alcohol), Bitter 55 (35% v/v of alcohol), Ramazzotti Amaro (28% v/v of alcohol), Zlatni pelin (28% v/v of alcohol), Jägermeister (35% vol of alcohol) and Gorki list (28% v/v of alcohol).

Green walnut liqueur preparation

For the GWL production, a traditional maceration procedure using a pilot-scale stainless steel container (250 L), located at the experimental field Radmilovac of the Faculty of Agriculture, University of Belgrade, was performed. Unripe walnut fruits (Juglans regia cultivar Cheinovo), less than 2 cm in diameters, were collected at the mentioned field in May 2017 (latitude 44° 45′ 17.5″ N, longitude 20° 35′ 01.6″ E).

About five thousand of whole green walnuts (25 kg) were used for the liqueur preparation. Additional ingredients were then added to enrich flavor and aroma of the final liqueur, namely: 18 kg of citrus fruits [oranges (Citrus sinensis) and lemons (Citrus limon) from the local supermarket, in the ratio 50:50 w/w (weight per weight), pre-treated with boiling water and cut into the quarters], 4 kg of dried fruits [1 kg of each dried figs (Ficus carica), dried cranberries (Vaccinium oxycoccos), dried plums (Prunus domestica) and dried sour cherry (Prunus cerasus), purchased at local health food store], 1 kg of dried aromatic plants [250 g of each vanilla pods (Vanilla planifolia), cinnamon pods (Cinnamomum verum), ground nutmeg (Myristica fragrans) and clove (Syzygium aromaticum), purchased at local health food store], 1 kg of 70% cacao dark chocolate bars (Factory of hocolate Simka, Vranje, Serbia), 1 kg of acacia honey (Gorjanović-Stakić Beekeeping Household, Stara Pazova, Serbia), and finally, 30 kg of commercial sugar. Water–ethanol solution (130 L; 60% v/v) was then added, and the mixture was left to macerate for 100 days, exposed to the sun. The total volume of 160 L of the extract was obtained by decanting and pressing the wet marc through the hydraulic press (INOP, Zemun, Serbia) at 50 bar. The obtained extract was diluted with water (40 L) up to the final alcohol strength of 25 ± 1.5%. Glycerol, as a viscous polyalcohol of a slightly sweet taste, was added in the amount of 0. 2 L/ 100 L to contribute to the body and fullness of the beverage. Caramel E150d was added as a colour enhancer. Obtained liqueur was finalized by filtration using filters Filtrox Fibrafix “AF6” and ‘‘AF31H’’, St. Gallen, Switzerland.

DC polarographic measurements of AO activity

The instrument used was the Princeton Applied Research (PAR) 174A Polarographic analyzer, Oak Ridge, USA. The DC polarographic i-E curves were recorded on Houston Instrument Omnigraphic 2000 X–Y recorder, USA (HPMC method) and by PC using XY recorder with National Instruments USB-6210 DAQ (L-Chem, Prague, Czech Republic, https://www.lchem.cz/zapisovac_en.html) (MRAP method). The dropping time of working dropping mercury electrode (DME) (with capillary characteristics of m = 2.5 mg/s at mercury reservoir height of 75 cm) was programmed on 1 s, while current oscillations were damped with low pass filter of instrument set at 3 s. Saturated calomel electrode (SCE) was used as a reference and platinum foil as an auxiliary electrode. The initial potentials for both HPMC and MRAP assays were 0.10 V versus SCE. The potential scan rate was 10 mV/s. As supporting electrolyte 20 mL of Clark and Lubs (CL) buffer (pH 9.8) with and without 5 mM H₂O₂ was used for HPMC and MRAP respectively. Before the first i–E curve recording and after every aliquot addition the electrochemical cell solutions were deaerated with pure nitrogen. To obtain linear dependence of limiting current (i_l) decrease on sample volume added, the aliquots of 20, 50 or 100 μL of undiluted samples were used. The slope of the initial, linear part, of i_l dependence on sample volume added (expressed in % per volume unit) was used as a measure of AO activity.

Spectrophotometric measurements of TPC and AO activity

Total phenolic content was determined by the Folin–Ciocalteu (FC) method described by Singleton et al. (1999). The DPPH antiradical activity was evaluated following the modified procedure described by Blois (1958). The FRAP assay was conducted according to the procedure described by Benzie and Strain (1996). The results for TPC were expressed as mg gallic acid equivalents (GAE) per liter of alcoholic drink using the calibration curve equation A(absorbance) = 0.011 × C(concentration) − 0.056, R² = 0.999. For DPPH and FRAP, mmol Trolox equivalents (TE) per liter of liqueur were calculated from the following calibration curve equations: %I (% of DPPH radical inhibition) = 126.0 × C − 7.344 (R² = 0.998) and A = 0.823 × C + 0.027 (R² = 0.995), respectively. All measurements were performed in triplicates.

Relative antioxidant capacity index (RACI)

RACI was calculated using results obtained by all AO assays, assigning equal weight to each of them as described by Petrović et al. (2016). A standard score was calculated according to the following equation:

$$\text{Standard score}=(x-\mu )/\sigma$$ 1

where x is the raw data, μ is the mean and σ is the standard deviation. The standard scores obtained for each sample and different assays, when averaged, gave a single unitless value termed RACI.

SPE–GC–MS analysis

The volatile profile of the GWL was analyzed using Thermo Fisher Focus GC (Carlsbad, USA) combined with Polaris Q mass spectrometer and TriPlusHS autosampler. Samples were prepared using solid-phase extraction (SPE). One mL of the sample was extracted using the Discovery DPA-6S SPE tubes. Extracted compounds were washed from the solid phase with methylene chloride. TR WAX-MS (30 m × 0.25 mm, 0.25 μm) capillary column was used, while analyzed samples were injected into GC through TriPlus AS autosampler (2 μL). The temperature program was: initial temperature 45 °C (8 min), then 8.0 °C/min to 230 °C (10 min). Injector, MS transfer line, and ion source temperatures were 250 °C, 200 °C, and 220 °C, respectively. Compounds were identified combining the NIST Standard Reference Data Program 1A and Wiley 275 Mass Spectral Library of authenticated standards. Helium was used as a carrier with a flow of 1.5 mL/min in split mode (30:1). Final results were expressed as a relative percentage (%) of compound in a total volatile fraction of the GWL.

Quantification of the phenolic compounds

For the separation and quantification of phenolic compounds, ultra-high performance liquid chromatography (Dionex Ultimate 3000 UHPLC with a diode array detector (DAD)) with a mass spectrometer (TSQ Quantum Access Max triple-quadrupole with heated electrospray ionization (HESI) (ThermoFisher Scientific, Basel, Switzerland) was used. The analytical separation was achieved on column Syncronis C18 (ThermoFisher Scientific, Bremen, Germany), with aqueous formic acid solution 0.1% (A) and acetonitrile (B) as components of the mobile phase. Details of gradient elution were given in Pantelić et al. (2017). Each polyphenol was quantitated by direct comparison with commercially available standards and the results were expressed as mg/L of liqueur.

Sensory quality assessment of obtained GWL and selected commercial spirits

The sensory quality rating of GWL and selected commercial spirits was performed by a sensory panel, consisting of researchers from the University of Belgrade experienced in alcoholic beverages quality judging. The spirits were presented to the judges monadically in random order. Overall sensory quality was assessed by evaluating five selected sensory characteristics: colour, clarity, distinction, odour (ortonasal olfaction) and flavour (Veljović et al. 2019), which were rated using category scales with score ranges 0–1, 0–1, 0–2, 0–6 and 0–10, respectively. The overall quality score, with a maximum score of 20, was calculated by adding the quality scores of the five individual characteristics. The quality of the spirits was ranked according to the following: excellent quality (quality score > 18), very good quality (16 < score < 18), good quality (14 < score < 16), poor/unsatisfactory quality (12 < score < 14), very poor quality (score < 12). All samples were evaluated in one replication.

Statistical analysis

Descriptive statistical analyses were performed using Microsoft Excel 2007 software for the means and the standard error of the mean calculation. The results are expressed in the form of mean ± standard deviation (SD). The evaluation of analysis of variance (ANOVA), and significant differences between samples are determined by post-hoc Tukey’s HSD test at 95% confidence limit, using StatSoft Statistica 12 software. Pearson coefficients were calculated using StatSoft Statistica 12 software.

Results and discussion

Within the scope of this study, a novel liqueur based on green walnuts was produced at the pilot-scale. Immediately after production, it was characterized in terms of AO activity, phenolic compounds, the content of volatiles, as well as sensory quality. For comparative purposes, similar spirits available at the Serbian market are analyzed in parallel.

Antioxidant activity of GWL and commercial spirits

Given the complexity and diversity of chemical reactions, occurring in different AO assays, especially in the case of food matrices containing different classes of antioxidants, each method for measurement of AO activity has its own limitations. Multiple methods have been employed to evaluate AO activity of samples, in order to obtain as much as a possible reliable insight into AO activity of analysed spirits and to enable more comprehensive comparison of their AO properties. Despite rapidness, accuracy, simplicity and low cost, electrochemical methods are still less commonly used than spectrophotometric ones. Here, two DC polarographic AO assays, HPMC and MRAP, have been applied in parallel with widely used spectrophotometric assays, FC, DPPH, and FRAP, to measure AO activity of GWL and similar commercial spirits. It is known that FC assay, commonly used to determine TPC, can be also considered as AO assay.

Also, relative antioxidant capacity index (RACI) was introduced (Table 1) as a ranking tool corresponding to the AO activity of analyzed spirits in a standardized form, being thus more reliable and more convenient for use by the food industry and scientists. RACI values, calculated by assigning equal weight to each AO assay result, enable direct comparison of antioxidant activity of the GWL and selected commercial spirits (Bałan and Tanumihardjo 2007; Petrović et al. 2016).

Table 1.

Antioxidant activity of green walnut liqueur (GWL) and commercial spirits measured by spectrophotometric assays (TPC, DPPH, and FRAP) and DC polarographic (HPMC and MRAP) assays, as well as relative antioxidant capacity index (RACI)

Spirits	TPC (mgGAE/L)	DPPH (mM TE)	FRAP (mM TE)	HPMC (%/mL)	MRAP (%/mL)	RACI
GWL	2385 ± 18f	13.99 ± 0.42 g	12.98 ± 0.09f	213 ± 10d	202 ± 13d	2.11
Underberg	1365 ± 13e	5.54 ± 0.10f	6.85 ± 0.03e	96 ± 4c	162 ± 21c	0.60
Bitter 54	661 ± 6d	3.08 ± 0.03e	2.41 ± 0.24d	92 ± 11c	128 ± 4b	0.02
Bitter55	515 ± 5c	2.62 ± 0.04d	1.88 ± 0.15c	80 ± 7c	130 ± 5b	− 0.09
Ramazzotti Amaro	649 ± 8d	2.04 ± 0.02c	1.23 ± 0.01b	77 ± 7c	47 ± 3a	− 0.38
Zlatni pelin	294 ± 22b	1.32 ± 0.01b	0.49 ± 0.01a	47 ± 5b	28 ± 3a	− 0.71
Jägermeister	542 ± 6c	2.02 ± 0.02c	1.31 ± 0.13b	52 ± 6b	26 ± 2a	− 0.58
Gorki list	249 ± 11a	0.78 ± 0.03a	0.66 ± 0.10a	13 ± 5a	31 ± 2a	− 0.83

Data are presented as the means of triplicate measurements ± standard deviation. Values with the same letter (a-f) are not statistically different at the p > 0.05 level (post hoc Tukey’s HSD test)

As it can be seen from the Table 1, the TPC determined by FC was found superior in GWL (2385 mg GAE/L). Almost two folds higher TPC was ascribed to GWL than Underberg and several-fold higher than all other analysed spirits. Such finding is in accordance with previously published results confirming that green walnuts liqueurs are a rich source of phenolics. The TPC of thirty samples of nocino liqueur samples, purchased in markets or prepared at home between 1977 and 2000, varied from 239 to 3884 mg catechine/L (Alamprese et al. 2005). Mrvcić et al. (2012) compared phenolic content among different spirits categories from the Croatian market and concluded that walnut liqueurs belong to the group with the highest TPC. GWL convincingly surpassed TPC of some red wines, with reported values 1724–1871 mg GAE/L (Paixao et al. 2007) and 1708–2314 mg GAE/L (Gorjanović et al. 2010b).

GWL also exhibited at least two times more potent scavenging activity towards artificial radical species DPPH and FRAP than Underberg and several times than other analysed spirits (Table 1). The order of AO activity of commercial spirits (Underberg, Bitter 54, Bitter 55, Jagermeister and Gorki list) has been found the same as reported in our previous study (Petrović et al. 2019). In addition to green walnut fruit, GWL contains additional herbs and spices, dark chocolate, honey and some fruits, forming the synergy of AOs, thus exhibiting strong AO activity.

Decrease of the anodic current of HydroxoPerhydroxoMercury(II) complex formation, in alkaline solutions of hydrogen peroxide, at the potential of mercury oxidation, occurred upon samples addition, is monitored by HPMC method. In Fig. 1, it can be observed that the decrease of the anodic current of HPMC formation is the most pronounced in the case of GWL (Fig. 1a), while Underberg (Fig. 1b) had two folds lower decrease.

Fig. 1.

Anodic polarographic curves obtained by gradual addition of a GWL and b Underberg in aliquots of 20 μL in 20 mL of buffered (CL buffer. pH 9.8) solution of 5 mmol/L H₂O₂. Inserts: % of decrease of HPMC anodic i_l versus volume of aliquots added

In Fig. 2a, the far greater effect of GWL (Fig. 2a) on the decrease of DC polarographic cathodic current of mercury(II) reduction than of the Gorki list (Fig. 2b) is shown.

Fig. 2.

Cathodic polarographic curves obtained by gradual addition of a GWL and b Gorki list in aliquots of 50 μL and 100 μL in 20 mL of buffered (CL buffer. pH 9.8) solution of 1 mmol/L HgCl₂. Inserts: percentage of Hg²⁺ cathodic limiting current (i_l) decrease versus volume of aliquots added

Decrease of anodic, as well as cathodic limiting currents (i_l), noticed upon gradual addition of samples, has been plotted against their volume (v). The slopes of the starting linear part of obtained plots were used as a measure of AO activity and are shown in inserts of Figs. 1 and 2.

HPMC was applied previously on various alcoholic beverages, such as red and white wines (Gorjanović et al. 2010a), different types of spirits (Gorjanović et al. 2010b), while MRAP assay, based on the decrease of DC polarographic cathodic current of mercury(II) reduction in the presence of AOs, has been applied here on alcoholic beverages for the first time.

High statistically significant (p < 0.01) correlation coefficients between spectrophotometric assays have been observed (0.982–0.992). HPMC correlated significantly positively (p < 0.01) with spectrophotometric assays (0.935–0.976), whereas MRAP was in lower, but still strong positive correlation with FC, DPPH, FRAP and HPMC (0.814–0.846). High correlations obtained validates the application of recently developed MRAP assay on complex samples such as spirits analyzed in this study. Also, a strong relationship between FC and other assays supports previous assumptions that phenolics are the main contributors to the total AO activity.

According to calculated RACI values, the following rank of order of analyzed spirits AO activity was observed: GWL (2.11) > Underberg (0.60) > Bitter 54 (0.02) > Bitter 55 (− 0.09) > Ramazzotti Amaro (− 0.38) > Jägermeister (− 0.58) > Zlatni pelin (− 0.71) > Gorki list (− 0.83). It’s therefore confirmed that, in comparison to selected commercial spirits, GWL is far superior in terms of AO activity, with RACI value even about four times greater than the following Underberg.

Phenolic constituents of GWL

Polyphenolic profile of GWL, obtained by HPLC analysis, is presented in Fig. 3. Multifold superiority of gallic acid in comparison to other detected individual phenolics can be observed, followed by catechine. The majority of studies dealing with green walnut liqueurs and extracts also found gallic acid as the predominant phenolic compound, and also detected some of the phenolic acids found in this study (chlorogenic, caffeic, protocatechuic, p-hydroxybenzoic, p-coumaric and ferulic acid) (Cosmulescu et al. 2014; Jakopič et al. 2007; Stampar et al. 2006). Cosmulescu et al. (2014) reported flavonoid catechin as the most abundant polyphenol in liqueurs prepared from green walnuts in three different ethanol concentrations. Several quercetin glycosides, including quercetin-3-O-glucoside and quercetin-3-O-rhamnoside, have been detected in the healthy tissue of walnut husks of six Juglans regia cultivars, whereas their content significantly increased after infection (Mikulic-Petkovsek et al. 2011). Naringin and naringenin can be found in citrus fruits (Raja Kumar et al. 2019), thus the presence of these flavonoids in GWL can be explained by the unique recipe which includes the addition of oranges and lemons. Chlorogenic acid is among the principal phenolic compounds contained in most herbs and spices, as well as p-coumaric and p-hydroxybenzoic acid (Shan et al. 2005).

Fig. 3.

Polyphenolic profile of GWL. Note: Data are expressed in mg/L of GWL and presented as the means of triplicate measurements ± standard deviation. GaA, gallic acid; C, catechin; ChA, chlorogenic acid; QUE-3-O-glc, quercetin 3-O-glucoside; QUE-3-O-rhm, quercetin-3-O-rhamnoside; NAR, naringin; QUE, quercetin; CaA, caffeic acid; PrA, protocatechuic acid; ISO-3-O-glc, isorhamnetin-3-O-glucoside; HbA, p-hydroxybenzoic acid; CoA, p-coumaric acid; NGN, naringenin; FeA, ferulic acid; AP, apigenin; LUT, luteolin; KAE, kaempferol; RUT, rutin

Based on the previously reported measurement of AO activity of individual compounds (Petrović et al. 2016), it can be concluded that the most prevalent gallic acid is by far the major contributor of FRAP and DPPH scavenging activity. Higher FRAP and DPPH scavenging activities were ascribed to gallic acid (10,829 and 3252 mM TE/mol) than to chlorogenic (1805 and 1170 mM TE/mol) and other benzoic or cinnamic acids detected, even of flavonoids catechine (1903 and 1402 mM TE/mol) and quercetin (4188 and 3088 mM TE/mol), respectively. According to MRAP, AO activity of gallic acid (27.8%/μmol) was higher than CGA (14.1%/μmol) but lower than catechine (50.9%/μmol) and quercetin (60.7%/μmol). Having in mind gallic acid content in GWL, it can be concluded that gallic acid is, anyway, the major contributor to AO activity measured by MRAP. Since it was not possible to measure AO activity of gallic acid using original HPMC assay, results of modified HPMC, developed to enable AO activity determination of samples less soluble in water (Stojićević et al. 2020), were considered to get an insight in the contribution of individual compounds present in GWL. According to modified HPMC, the contribution of gallic acid to total AO activity of GWL was estimated to be lower than in MRAP since the strongest AO activity was ascribed to quercetin, followed by twice as lower CGA and gallic acid. However, it can be concluded that gallic acid is the most important contributor to total AO activity, measured by all AO assays, followed by catechine, quercetin and CGA.

Volatile constituents of GWL

The results of GC–MS analysis of the GWL are shown in Table 2. The predominance of phenylpropanoid eugenol can be observed. Eugenol, a pleasant, spicy, clove-like flavor, is the major volatile component in clove buds (Eugenia caryophyllus), representing up to 95% of its total volatile fraction (Zachariah and Leela 2006). Also, nutmeg and cinnamon are rich sources of this aromatic compound (Li et al. 2013; Muchtaridi et al. 2010). GWL also contains considerable amounts of 4-terpineol (2.84%) and α-terpineol (9.19%), which are naturally contained in nutmeg (Myristica fragrans), as well as in citrus juices (Nguyen et al. 2009). Elemicin 1.60% and myristicin 0.85% may also originate from nutmeg seed (Muchtaridi et al. 2010), whereas vanillin (5.48%) is probably derived from the vanilla beans.

Table 2.

Volatiles composition of green walnut liqueur (GWL) obtained by SPE–GC/MS analysis

	Compound	%
1	Eugenol	76.90
2	α-terpineol	9.19
3	Vanillin	5.48
4	4-Terpineol	2.84
5	Elemicin	1.60
6	3,5-di-tert-buthyl-4-hydroxybenzaldehyde	0.87
8	Myristicin	0.85
9	2,4-Dimethylbenzaldehyde	0.48
10	Diethyl malate	0.28
11	Linalool	0.28
12	Linalyl acetate	0.25
13	Methyleugenol	0.21
14	2-Ethylbenzaldehyde	0.14
15	3-Allylsalicylaldehyde	0.12
16	Cinnamyl alcohol	0.09
17	Benzyl alcohol	0.06

Biological activity of terpenes, including antioxidant activity, has been demonstrated earlier, suggesting the realistic possibility of their contribution to the total AO activity of GWL. It was reported that the oxygenated-type of monoterpenes with phenolic structure, such as eugenol, found in GWL in major percentage, have the highest AO activity within the monoterpenes group (Gonzalez-Burgos and Gomez-Serranillos 2012). Scavenging activity of eugenol and its derivatives methyl and isomethyl eugenol towards DPPH and ABTS was higher than ascorbic acid (Sohilait and Kainama 2019). The hypothesis that eugenol reduces two or more DPPH radicals, despite the availability of only one hydrogen from a hydroxyl group was proposed as well as the formation of dimers (dehydrodieugenol) with two phenolic hydroxyl (Cortés-Rojas et al. 2014). Thus, a significant contribution of eugenol to scavenging activity of GWL towards DPPH might be supposed. Determination of AO activity of spices and their active principles by differential pulse voltammetry (Palma et al. 2014), as well as by DC polarography (modified HMPC assay) (Stojićević et al. 2020) confirmed that eugenol had the highest AO activity amongst terpenoids. Although much lower AO activity was ascribed to eugenol (0.50 × 10⁶%/mol) than to gallic acid and CGA (13.50 and 15.71 × 10⁶%/mol), and quercetin (38.50 × 10⁶%/mol) (Stojićević et al. 2020), its contribution to total AO activity measured by HPMC should not be neglected. Thus, the greatest percentage of eugenol in the total volatile fraction of GWL indicates its considerable contribution to total AO activity.

Sensory evaluation

The results of GWL sensory quality assessment, performed by experts in the field of alcoholic beverages, and comparison with selected commercial spirits are shown in Table 3. Differences between the score means for distinction and odour attributes of all spirits evaluated were not statistically significant (p > 0.05). Colour, clearness and taste of GWL were also scored high and not significantly different (p > 0.05) when compared to the majority of commercial spirits. The total sensory scores of all analyzed spirits were in the range from very good to excellent quality (16.43–18.66). Bitter 55, made of 55 herbs, spices, fruits, and Ganoderma lucidum, had the highest quality score (18.66) among all samples analyzed, followed by GWL with lower but non-significantly different (p < 0.05) total sensory score (18.52). Bitter 54, Amaro Ramazzotti, Jägermeister and Gorki list were also ascribed with high total scores (18.18–18.47), forming the non-significantly different (p < 0.05) group with above mentioned Bitter 55 and GWL. The lowest total sensory score (16.43), significantly lower (p < 0.05) among all analyzed spirits, was ascribed to Underberg, containing the highest content of ethanol (44% v/v). This result corroborates with the previous findings that ethanol influences considerably on sensorial perceptions of distilled beverages. The possible explanations appear to be associated with anesthetic qualities and stimulative effects on the trigeminal nerves of ethanol, which may diminish the other odor-active compounds in distilled beverages (Taylor et al. 2010).

Table 3.

Sensory quality assessment of green walnut liqueur (GWL) in comparison to similar commercial spirits

Samples	Colour (max 1 pts)	Clearness (max 1 pts)	Distinction (max 2 pts)	Odour (max 6 pts)	Taste (max 10 pts)	Total (max 20 pts)
GWL	0.96 ± 0.03b	0.94 ± 0.03b	1.92 ± 0.07a	5.40 ± 0.18a	9.30 ± 0.33c	18.52 ± 0.47bc
Underberg	0.82 ± 0.16a	0.79 ± 0.15a	1.83 ± 0.08a	4.99 ± 0.31a	8.00 ± 0.42a	16.43 ± 0.26a
Bitter 54	0.98 ± 0.02b	0.91 ± 0.07ab	1.96 ± 0.07a	5.42 ± 0.23a	9.20 ± 0.14c	18.47 ± 0.29c
Bitter 55	0.97 ± 0.02b	0.92 ± 0.07ab	1.93 ± 0.08a	5.46 ± 0.31a	9.38 ± 0.20c	18.66 ± 0.18c
Jägermeister	0.96 ± 0.04b	0.95 ± 0.03b	1.91 ± 0.07a	5.33 ± 0.22a	9.22 ± 0.04bc	18.37 ± 0.17c
Gorki list	0.96 ± 0.05b	0.96 ± 0.04b	1.94 ± 0.07a	5.30 ± 0.36a	9.02 ± 0.35bc	18.18 ± 0.37c
Amaro Ramazzotti	0.96 ± 0.04b	0.96 ± 0.02b	1.92 ± 0.07a	5.32 ± 0.21a	9.26 ± 0.09bc	18.42 ± 0.19c
Zlatni pelin	0.94 ± 0.06ab	0.94 ± 0.04b	1.92 ± 0.08a	5.32 ± 0.39a	8.56 ± 0.26b	17.68 ± 0.46b

Data are presented as the means of five evaluations of sensory experts ± standard deviation.Values with the same letter (a–c) are not statistically different at the p > 0.05 level (post hoc Tukey’s HSD test);

Based on the total sensory score, the excellent sensory quality of GWL, comparable and not significantly different (p > 0.05) from popular commercial similar spirits, was observed by the panel of experts, which indicates the market feasibility for this natural product with pronounced functional properties.

Conclusion

Liqueur obtained within the scope of this study from green walnuts with the addition of aromatic plants and fruits, chocolate and honey, represents a rich source of phenolic and volatile compounds, with prominent AO activity much higher than similar commercially available spirits. The multi-method approach (parallel application of various spectrophotometric and polarographic assays) enabled a more comprehensive evaluation of AO properties of obtained GWL. Integration of AO values, determined by mentioned methods based on different mechanisms, into one dimensionless quantity (RACI) enabled ranking that clearly showed the dominance of GWL concerning AO activity.

Gallic acid, found in the largest amount among all polyphenols analyzed, was followed by chlorogenic acid. Catechin, quercetin and its sugar-conjugated derivatives were also present in a considerable amount, while eugenol represented the most abundant volatile component present in GWL. Gallic acid is the most important contributor to total AO activity, first of all to FRAP and DPPH scavenging and then to reducing activity determined by MRAP and, finally, by HPMC. Contribution of flavonoids, well-known as highly efficient antioxidants, has been found more significant in AO activity determined by HPMC and MRAP than DPPH and FRAP. The most significant contribution to total AO activity amongst volatile compounds was ascribed to eugenol, whose participation in the strong AO potential of the GWL should not be neglected.

The presence of volatile and phenolic compounds influence the sensorial properties of GWL as well. The excellent sensory quality of GWL, comparable to the majority of selected commercial spirits, was confirmed by the scoring method. Considering the growing market demand for natural and high-quality products with a high content of bioactive phytochemicals, GWL, with its unique ingredient composition, can be considered a promising candidate for further development and production scale-up, with great potential and opportunity for commercialization.