lable at ScienceDirect

Food Control 89 (2018) 270e279

Contents lists avai

Food Control

journal homepage: www.elsevier .com/locate/ foodcont

Quality of limes juices based on the aroma and antioxidant properties

Martyna Lubinska-Szczygieł a, Anna R�o _za�nska a, Jacek Namie�snik a, Tomasz Dymerski a, *,Rajamohamed Beema Shafreen b, Moshe Weisz c, Aviva Ezra c, Shela Gorinstein c, **

a Gda�nsk University of Technology, Faculty of Chemistry, Department of Analytical Chemistry, Gda�nsk, 80-233, Polandb Molecular Nanomedicine Research Unit, Centre for Nanoscience and Nanotechnology, Sathyabama University, Chennai, 600119, TN, Indiac Institute for Drug Research, School of Pharmacy, Hadassah Medical School, The Hebrew University, Jerusalem, 9112001, Israel

a r t i c l e i n f o

Article history:Received 15 December 2017Received in revised form6 February 2018Accepted 7 February 2018Available online 7 February 2018

Keywords:Citrus hystrixCitrus aurantifoliaVolatilesAntioxidantsNMR shifts

Abbreviations: Polyph, polyphenols; GAE, gallic aequivalent; Flavan, flavanols; Flavon, flavonoids; Vitcyanins; CGE, cyanidin-3-glucoside equivalent; Chxanthophyllsþcarotenes; ABTS, 2, 2-Azino-bis (3-ethacid) diammonium salt; FRAP, Ferric-reducing/anCupric reducing antioxidant capacity; 1,1-diphenyl-trolox equivalent.* Corresponding author.** Corresponding author.

E-mail addresses: tomasz.dymerski@gmail.com (T.huji.ac.il, gorin@cc.huji.ac.il (S. Gorinstein).

https://doi.org/10.1016/j.foodcont.2018.02.0050956-7135/© 2018 Elsevier Ltd. All rights reserved.

a b s t r a c t

Kaffir (Citrus hystrix) and Key (Citrus aurantifolia) limes juices were investigated and compared. Twodimensional gas chromatography coupled with time-of-flight mass spectrometry (GC�GC-TOF-MS) wasapplied to assess the botanical origin of Kaffir and Key limes juices, based on volatile substances. Thebiggest differences in the contents of selected terpenes in Kaffir and Key limes occur in chemical com-pounds such as Limonene, Citral, Terpinen-4-ol. Limonene concentration is almost 8 times higher in theKey lime volatile fraction than in Kaffir lime. The difference in concentration of Citral in Kaffir lime isalmost 20mg/kg lower than in Key lime. Higher concentration of Terpinen-4-ol was noted in Kaffir limesamples and the content was almost 20 times higher. The concentrations of a-Pinene, Citronellal,Camphene, Nerol, trans-Geraniol and b-Pinene are at similar levels in the volatile fraction of both fruits.Bioactive substances (polyphenols, flavonoids, tannins and flavanols) and the values of antioxidant ca-pacities by four radical scavenging assays (DPPH, CUPRAC FRAP, ABTS) were determined and compared inwater and methanol extracts in Kaffir and Key limes juices. The bioactivity of Kaffir lime differ signifi-cantly in water extracts in comparison with Key lime juices. The 1H NMR shifts in methanol and chlo-roform extracts showed some differences in aromatic region between the two varieties of lime juices.Terpinen-4-ol for Kaffir lime and Citral for Key lime were used as potential markers. The GC�GC-TOF-MSallows better separation of substances originating from complex matrices than one-dimensional chro-matography, based on improved resolution, increased peak capacity and unique selectivity. The possiblefalsification of mentioned juices can be detected by the use of GC�GC-TOF-MS, antioxidant assays andNMR shifts.

© 2018 Elsevier Ltd. All rights reserved.

1. Introduction

Kaffir lime (Citrus hystrix) is one of the most popular fruits inThailand or Laos. Kaffir lime leaves are one of the most commonlyused Thai spices. Despite the leaves, the skin is also used for

cid equivalent; CE, catechinC, vitamin C; Anthoc, antho-lor, chlorophyll; XanþCarot,yl-benzothiazoline-6-sulfonictioxidant power; CUPRAC,2-picrylhydrazyl, DPPH; TE,

Dymerski), shela.gorin@mail.

culinary purposes, because of specific aroma. Both the leaves andthe skin contain many chemical compounds with a healthy effect.Limonene, a-Terpineol, 2b-Pinene, Terpinen-4-ol, g-Terpinene, a-Terpinene, and a-Terpinolene are common terpenes in leaves(Srisukh et al., 2012a,b). In turn, the content of the individual ter-penes in the skin were estimated: Limonene 40.65%, Terpinen-4-ol13.71%, a-Terpineol 13.20% (Srisukh et al., 2012a,b; Thanaboripat,Chareonsettasilp, & Pandee, 2006). Kaffir lime pulp and juice arenot consumed directly (Waikedre et al., 2010). However, they alsocontain many bioactive substances. Kaffir limes do not grow intemperate climate, and these fruits are also not imported into Eu-ropean countries. In Europe, the most popular and available be-tween lime varieties is Key lime (Citrus aurantifolia), which alsocontains many bioactive terpenes (Spadaro, Costa, Circosta, &Occhiuto, 2012).

The content of individual terpenes varies in the volatile fractions

M. Lubinska-Szczygieł et al. / Food Control 89 (2018) 270e279 271

of each above-mentioned fruits. It is extremely important todetermine terpenes in fruit products, because of their health-promoting effect or on the other site their excess can causehealth problems. One of themost popular food products made fromlimes is juice. Key lime juice is used as an additive to beverages orsauces, oppositely, Kaffir lime juice has sour and bitter taste andvery often is classified as an industrial waste. In many countries foreconomic reasons adulteration investigations of products contain-ing Key lime with Kaffir lime juice is provided. The major chemicalcompounds found in the Kaffir lime juices volatile fraction mayhave potential allergic effects (Rubel, Freeman, & Southwell, 1998)as well as a large number of antioxidants may induct allergic dis-eases (Allan, Kelly, & Devereux, 2010). Assessment of the authen-ticity of juices is also important for food industry. It preventsproducers from material losses due to contamination of the pro-duction line. Therefore, it is extremely important to develop ananalytical method to identify possible botanical origin of limes.

The applications of two-dimensional gas chromatography(GC�GC) and time of flight mass spectrometry (TOFMS) to analyzearoma of food products are shown in a number of reports (BoguszJunior et al., 2015; Dymerski et al., 2015; 2016). Two-dimensionalgas chromatography is useful tool to analyze fruit samples. Aromaprofile of the volatile fraction of apples, pears, and quince fruit wereperformed (Schmarr& Bernhardt, 2010). In turn, 3-methylbutan-1-ol, 3-methylbutan-1-ol acetate, 2-phenylethyl acetate and phenyl-ethyl alcohol were selected as compounds characteristic for bananasmell (Capobiango et al., 2015). Using GC�GC-TOFMS technique itwas also possible to quantify the volatile compounds of differentkinds of berries (Dymerski et al., 2015). Untargeted analysis wasalso performed after the postharvest and the storage of apples(Risticevic, Deell,& Pawliszyn, 2012). It was also possible to indicateterpenes in the samples of grapes (Banerjee et al., 2008; Rocha,Coelho, Zrostlíkov�a, Delgadillo, & Coimbra, 2007) and blueberries(Kupska, Chmiel, Jedrkiewicz, Wardencki, & Namie�snik, 2014).Strawberries growing in Australia have been distinguished due totheir botanical origin (Samykanno, Pang, & Marriott, 2013) anddifferent varieties of chili due were classified according to theirspecies (Bogusz Junior et al., 2015). Strawberries were also exam-ined in order to analyze profile of volatile fraction (Williams, Ryan,Olarte, Marriott, & Pang, 2005). Dymerski et al. (2016) classifiedsamples of cranberries, blueberries and cranberries. It is alsopossible to determine the pesticide residues in fruit samples(Zrostlíkov�a, Haj�slov�a, & Cajka, 2003).

The composition of the volatile fraction of essential oil ofC. aurantifolia was analyzed using GC-MS by Spadaro et al. (2012).Analysis of volatile fraction of Kaffir lime was performed using GC-MS technique. It was possible to select 15 major chemical respon-sible for the flavor of Kaffir lime (Kasuan et al., 2013). Nevertheless,there are no literature reports about authenticity markers of above-mentioned types of limes, including also the use of two-dimensional gas chromatography.

Similarly, the situation is revealed in case of studies concerningthe comparison of antioxidant activities of Kaffir and Key fruitjuices. There are only a few investigations, in which a total phenolicand flavonoid contents, ferric reducing antioxidant power (FRAP)and 1, 1-diphenyl-2-picryl hydrazyl (DPPH) radical scavenging ac-tivity were determined (Ghafar, Prasad, Weng, & Ismai, 2009). Thecharacterization of lime juices from the point of their antioxidantstatus is important. Therefore, the aim of this study was to compareKaffir and Key lime juices regarding to their aroma and antioxidantproperties. For this reason, the advanced analytical methods wereelaborated, with the use of two dimensional gas chromatographycoupled with time-of-flight mass spectrometry, 1H NMR spectros-copy and the investigation concerning antioxidant properties, usinga number of radical scavenging assays were included. According to

the best of our knowledge, there are no literature reports about thequantitative determination of selected terpenes of above-mentioned juices using spectrometric methods and there is a lackof information about comparison of these matrices in respect oftheir bioactivities and NMR shifts in the aromatic region. Such in-vestigations are very important for food control of the preparedlimes juices.

2. Materials and methods

2.1. Chemicals

Analytical terpene standards: a-Pinene, Limonene, Citronellal,Aromadendrene, Camphene, Linalool, Nerol, trans-Geraniol, b-Pinene, Terpinen-4-ol, Myrcene, g-Terpinene, a-Terpineol, Citral(Sigma-Aldrich, St. Louis, MO, USA) were used to prepare standardsolutions for calibration step. Methanol (Avantor PerformanceMaterials Poland S.A) was used as a solvent of these solutions.Trolox (6-hydroxy-2,5,7,8,-tetramethyl-chroman-2-carboxylicacid); 2,20-azobis-2-methyl-propanimidamide; FeCl3x6H2O;FolineCiocalteu reagent (FCR); Tris, tris (hydroxymethy1)amino-methane; lanthanum (III) chloride heptahydrate; CuCl2�2H2O; and2,9-dimethyl-1,10-phenanthroline (neocuproine), 1,1-diphenyl-2-picrylhydrazyl (DPPH), potassium persulfate, deuterated chloro-form (CDCl3), deuterated methanol-d4 (CH3OH-d4), and deuteriumoxide (D2O) were obtained from Sigma Chemical Co., St. Louis, MO,USA. 2, 4, 6-tripyridyl-s-triazine (TPTZ) was purchased from FlukaChemie, Buchs, Switzerland. All reagents were of analytical grade.Deionized and distilled water were used throughout.

2.2. Sample preparation

The objects of studywere the pulps of Kaffir lime (Citrus hysteria,Citrus hystrix) and Key lime (Citrus aurantifolia). The samples ofKaffir lime fruits were imported from Thailand where they hadbeen bought on the floating market in Taling Chan, which is locatedin the western part of Bangkok. Samples were transported toPoland in sealed plastic bags in portable fridge maintained at be-tween 10 and 15 �C. Key limes were bought in local distributionpoint in Poland. According to the seller's information, the countryof origin of the fruit was Brazil.

In order to prepare for analysis, the fruits were washed with tapwater and rinsed with distilled water. The fruit peel was thenseparated from the pulp and then squeezed to obtain the juices(Fig. 1). The next step was to weigh out 5.0± 0.1 g of sample unifiedcomposition in vials of 20mL and then 1mL of deionized water wasadded to the sample. The vials were closed with caps with siliconeTeflonmembrane. The procedure was repeated three times for eachspecies of lime, each time using a new fruit.

2.3. Isolation and enrichment of analytes

Solid phase microextraction was used to carry out isolation andenrichment of analytes. The extraction was conducted using thedivinylbenzene/carboxen/polydimethylsiloxane (DVB/CAR/PDMS)fiber with thickness of 50/30 mmand length of 2 cm (Sigma-Aldrich,St. Louis, MO, USA). The extraction was carried out at 40 �C for35min. After this step the thermal desorption of the analytes attemperature of 250 �C for 5min was provided. Between eachanalysis fiber was desorbed at 250 �C for 5min. Extraction step wasmade using a MPS autosampler (Gerstel Co., Mülheim, Germany).

2.4. Instrumentation

Two-dimensional gas chromatograph Agilent 7980 (Agilent

BA

D

C

Fig. 1. A, B, C, D, Kaffir lime, Kaffir juice, Key lime, Key juice.

M. Lubinska-Szczygieł et al. / Food Control 89 (2018) 270e279272

Technologies, Palo Alto, CA, USA) equipped with a liquid nitrogencooled two-stage cryogenic modulator and the dispenser, workingin spilt/splitless mode was used to carry out the analysis. Differenttypes of chromatography columns were chosen to provide properseparation according to the rule of orthogonality. Table 1 shows thecolumn parameters. Separation of substances was done by usingthe following temperature program: initial temperature of 40 �Cwas held for 3.5min, then a linear increase of temperature to 250 �Cat a rate of 5 �C/minwas applied. The final temperaturewas held for5min. The temperature program applied in the secondary ovenwasset up with 5 �C shift. Modulation period was set up to 4 s. As thecooling medium, the liquid nitrogen was used, and hydrogen ofpurity N 6.0 was utilized as a carrier gas.

The volumetric flow rate was 1mL/min. A single run time was43.5min. The injector worked in splitless mode at temperature of250 �C. Temperature of transfer line and ion source was also 250 �C.The voltage of detector was set up at 1600 V. The data werecollected over a mass range of m/z from 40 up to 500 with theacquisition rate of 125 spectra/s.

Table 1The parameters of chromatographic columns.

Parameters I dimension column

Type: capillaryLength: 30mInternal Diameter: 250 mmMaximum temperature: 325 �CTrade name of stationary phase: Equity 1 (Supelco, BellefoThe film thickness of the stationary phase 0.25 mm

2.5. Data analysis

To identify the chemicals time of fight mass spectrometerPegasus 4D produced by LECO (LECO Corp., St. Joseph, MI, USA) wasused. Processing of data was done automatically using chromato-graphic peak deconvolution algorithm implemented in the soft-ware ChromaTOF (LECO Corp., version 4.44.0.0). Tentativeidentification of analytes was made by comparing experimentalspectrawith the spectra included in NIST 11 andWiley libraries andby comparing calculated linear temperature-programmed reten-tion indices (LTPRIs) with literature values. LTPRI values werecalculated by performing analysis of C8eC20 n-alkanes. Positiveidentification was done using analytical terpenes standards.

2.6. Determination of bioactive compounds and total antioxidantcapacities (TACs)

Polyphenols were extracted with methanol and water (concen-tration 20mg/mL) during 1 h in a cooled ultrasonic bath. Totalpolyphenols (mg gallic acid equivalents (GAE)/g DW) were

II dimension column

capillary1.6m100 mm280 �C

nte, PA, USA) SGWAX (SGE Analytical Science, Austin, TX, USA)0.10 mm

M. Lubinska-Szczygieł et al. / Food Control 89 (2018) 270e279 273

determined by Folin-Ciocalteu method using spectrophotometer(Hewlett-Packard, model 8452A, Rockvile, USA) and measuring ob-tained absorbance after the complex reaction at wavelength of750 nm (Singleton, Orthofer, & Lamuela-Raventos, 1999). Anthocy-anins were determined by the measuring of absorbances of limeextracts (1 g of the defatted sample was extracted with 1mL ofacetonitrile containing 4% acetic acid) at 510 nm and 700 nm inbuffers at pH 1.0 and 4.5, and calculated using following equation:A¼ [(A510 e A 700) pH1.0-(A510 e A 700) pH4.5] with a molar extinctioncoefficient of cyaniding-3-glucoside of 29, 600. Results wereexpressed as milligrams of cyaniding-3-glucoside equivalent per100 g dw (Cheng & Breen, 1991). Total carotenoids (xantho-phyllsþcarotenes) were extracted with 100% acetone and deter-mined spectrophotometrically at different absorbances (nm) such asat 661.6, 644.8, and 470, respectively (Boyer, 1990). Flavonoids,extracted with 5% NaNO2, 10% AlCl3 x H2O and 1M NaOH, weremeasured at 510 nm. Total flavanols were estimated using the p-dimethylaminocinnamaldehyde method, and the absorbance wasmeasured at 640 nm (Feucht & Polster, 2001). The extracts ofcondensed tannins (procyanidins) with 4% vanillin solution inMeOHwere measured at 500 nm. (þ)Catechin served as a standard forflavonoids, flavanols and tannins as previously was described indetails (Leontowicz et al., 2016). Total ascorbic acid was determinedby CUPRAC assay inwater extract (100mg of lyophilized sample and5mL of water). The absorbance of the formed bis (Nc)-copper (I)chelate was measured at 450 nm (Ozyurek, Guclu, Bektasoglu, &Apak, 2007).

TACs were determined using the following methods:2, 2-Azino-bis (3-ethyl-benzothiazoline-6-sulfonic acid) dia-

mmonium salt (ABTS) method. ABTS radical cation was generatedby the interaction of ABTS (7mM/L) and K2S2O8 (2.45mM/L). Thissolution was diluted with methanol and the absorbance wasmeasured at 734 nm (Re et al., 1999).

Ferric-reducing/antioxidant power (FRAP): FRAP reagent (2.5mLof a 10mmol ferric-tripiridyltriazine solution in 40mmol HCl plus2.5mL of 20mmol FeCl3xH2O and 25mL of 0.3mol/L acetate buffer,pH 3.6) of 900 mL was mixed with 90 mL of distilled water and 30 mLof asparagus extract samples as the appropriate reagent blank andabsorbance was measured at 595 nm (Benzie & Strain, 1996).

1, 1-Diphenyl-2-picrylhydrazyl method (DPPH) solution (3.9mL,25mg/L) in methanol was mixed with the samples extracts(0.1mL). The reaction progress was monitored at 515 nm until theabsorbance was stable. The scavenging rate on DPPH radicals wascalculated (Brand-Williams, Cuvelier, & Berset, 1995).

Cupric reducing antioxidant capacity (CUPRAC): To the mixtureof 1mL of copper (II)-neocuproine and NH4Ac buffer solution,acidified and non acidified methanol extracts of lime (or standard)solution (x, in mL) and H2O [(1.1-x) mL] were added to make thefinal volume of 4.1mL and the absorbance was measured at 450 nm(Apak, Guclu, Ozyurek, & Karademir, 2004).

2.7. Sample extraction and 1H NMR analysis

Fine powder freeze dried material of 70mg of each sample wasadded with either 700 mL of CD3OD þ D2O (ratio 1:1) or 700 mL ofCDCl3. The suspension (in a 1.5mL Eppendorf tube) was ultra-sonicated at room temperature for 30min. And then, the suspen-sion was centrifuged at 13.000 rpm for 10min. The supernatantwas transferred into 5mL NMR tube and analyzed for its 1H NMR.CD3OD þ D2O aimed to extracts polar metabolites, while CDCl3extracted non polar metabolites. All NMR experiments wererecorded on Bruker 500 NMR spectrometer equipped with a 5-mmPABBO BB-probe head (499.953 for 1H shifts) at 25 �C. NMR dataprocessing was performed using MestReNova software (AbdulHamid et al., 2017; Drzewiecki et al., 2016).

2.8. Statistical and classification analysis

The results of quantitative analysis were expressed as meanvalue and standard deviation (SD) of three measures of concen-trations. Differences between groups were analyzed using two-wayanalysis of variance (ANOVA) followed by Duncan's new multiplerange test with a¼ 0.05. The analysis was carried out using STA-TISTICA 12 (StatSoft, Inc., Tulsa, Oklahoma, USA).

The peak areas obtained by GC�GC-TOF-MS analysis were usedto sample classifications. Orange Canvas Data Mining (Bioinfor-matics Lab, University of Ljubljana, Slovenia) was used to performSupport Vector Machine (SVM), Tree Classification (TC), NaïveBayes (NB) and Random Forest (RF) classifications with 2-foldcross-validation. The target class was the average over classes. Allthe classifiers were taken with their optimal settings.

3. Results and discussion

3.1. Composition of volatile substances

Detected chemical compounds were grouped according to theirchemical classes (Fig. 2).

Comparing the volatile fractions of both species of fruits, it canbe observed that the most numerous groups of chemical com-pounds presented in the volatile fraction, are terpenes. Theyrepresent nearly 88% of all volatile substances present in the Kaffirlime pulp, while in case of Citrus aurantifolia is about 53%. Kaffirlime pulp is therefore contains more aromatic compounds than Keylime. Additionally, high content of terpenes, which are consideredas bioactive chemical compounds, makes this fruit as a rich sourceof prohealth constituents. In addition, 30% of difference in terpenescontent explains the different odors of both fruits. It is well proventhat the aroma of citrus fruits is composed of complex mixture ofterpenes, which are chemical compounds whose main skeletonwas formed by the combination of five-carbon isoprene units(Sharon-Asa et al., 2003). They are therefore the main group ofKaffir lime compounds and have a complex of bioactive propertiessuch as antioxidant, antimicrobial or antiulcer effects (Al-Doghairi,El-Nadi, Elhag, & Al-Ayedh, 2004). Based on these properties, Kaffirlimes can be classified as a superfruit, which were characterized bypro-health properties backed up by scientific research, containedbioactive compounds, stand out in exotic origin and taste. Alcohols,esters are the groups of chemical compounds with the smallestcontribution in the composition of volatile fraction of Citrus Hys-teria, which do not exceed 1%. In Citrus aurantifolia their content involatile fraction is about 3%. Hydrocarbons are the next group andtheir percentage is more than 10%. Ketones represent 4% and al-dehydes only 2% of the total content of the headspace of Kaffir lime.In case of Citrus aurantifolia, the content of aldehydes is similar.Such a distribution of all compounds is responsible for the smell ofthe fruits. Characteristic, intense scent is caused by the high contentof terpenes compounds. Due to the very low content of carboxylicacids (<1%), the aroma of Kaffir lime has some irritating odor, whichcould cause some negative sensory perceptions during food con-trol. To compare the content of selected terpenes, an analysis ofCitrus aurantifolia, a fruit very popular and widely available inEurope, was also conducted. The object of the study was the juice ofthe above-mentioned lime fruits. The content of different terpenesfor Kaffir and Key limes based on the results of Table 2 differ be-tween them. In the volatile fraction of Kaffir lime,119 terpenes weredetected, and in case of Key lime - 87. Only chemical compoundswith the mass spectral match factor, similarity >800 wereconsidered.

The volatile fractions of both species of limes with the identifiedmajor chemical compounds are shown in Table 2. They were

Fig. 2. Distribution of volatiles by chemical classes for: A. Kaffir lime, B. Key lime.

M. Lubinska-Szczygieł et al. / Food Control 89 (2018) 270e279274

identified based on comparing spectra and LTPRI with literaturedata and the retention times were compared with retention time ofinternal standard. As it can be seen, all of the compounds belong tothe terpenes family.

It can be observed, that the biggest differences in the content ofselected terpenes in Kaffir and Key limes occur in case of chemicalcompounds such as Limonene, Citral, Terpinen-4-ol (Table 3).Limonene concentration is almost 8 times higher in the Key limevolatile fraction than in Kaffir lime. Extremely low content ofLimonene compound in Kaffir lime was also found in citrus fruits(Waikedre et al., 2010). In the case of Citral, the difference in con-centration of this compound in Kaffir lime is almost 20mg/kg lowerthan in Key lime. Higher concentration of Terpinen-4-ol was noted

in Kaffir lime samples and the content was almost 20 times higher.Terpinen-4-ol is the major chemical compound of volatile fractionof Kaffir lime. Terpinen-4-ol was selected as a major component ofCitrus hystrix essential oil (Waikedre et al., 2010). For the otherterpenes, differences in the contents are not statistically significant.The concentrations of a-Pinene, Citronellal, Camphene, Nerol,trans-Geraniol and b-Pinene are at similar levels in the volatilefraction of both fruits. In addition, in both cases the amount of thediscussed analytes does not exceed 5mg/kg. Terpenes, whosenumbers in the fraction of fruits don't reach 10 ppm, are Myrceneand g-Terpinene. The variation in the amount of the compoundsexplains the significant differences in taste and aroma of both fruitspecies. Among the determined terpenes, potential markers of Key

Table 2The major compounds identified in the volatile fraction of Kaffir and Key limes using GC�GC-TOF-MS.

No. Chemical compound RT1 [s] Average RT2 [s] Similarity Unique mass LTPRIlit LTPRIcalc

1 b-Pinene 862 1.3 854 93 962 9632 Sabinene 1386 2.3 939 93 958 9593 Citronellal 1114 1.8 886 69 1132 11314 Linalool 1060 2.3 745 71 1082 10835 a-Terpineol 1294 1.4 810 59 1289 12896 b-Citronellol 1238 2.8 928 69 1211 12097 Citronellyl acetate 1402 1.7 916 69 1335 13378 a-Copaene 1757 1.4 937 161 1353 135410 a-Cubebene 1434 1.23 870 161 1345 134811 b-Caryophyllene 1580 1.7 917 93 1421 142012 Limonene 945 1.5 936 93 1022 102413 Germacrene D 1536 1.2 907 161 1486 148514 a-Pinene 796 1.3 861 93 949 95015 Capmhene 812 1.3 951 91 953 95616 Terpinen-4-ol 1170 2.3 864 71 1163 106017 a-Terpinene 930 1.4 939 93 1010 100818 Myrcene 857 1.4 808 93 994 99319 a-Phellandrene 880 1.4 915 93 991 99220 a-Thujene 1286 3.7 899 93 923 92321 g-Terpinene 1042 1.7 914 93 1050 105222 b -Phellandrene 850 1.3 836 93 1031 103023 Citral 1250 2.4 885 69 1240 124124 Nerol 1242 3.1 923 69 1228 123125 Geraniol 1318 3.9 890 69 1233 123226 Aromadendrene 1720 1.8 851 91 1455 1456

RT 1 e first dimension retention time, RT 2 e second dimension retention time, LRIcalc e Linear Retention Index calculated; LRIlit e Linear Retention Index reported in theliterature for DB 1 or equivalent stationary phase.

Table 3Concentration of selected terpenes in the volatile fraction of Kaffir and Key limes.

No. Chemical compound R2 LOQ LOD Concentration± SD [mg/kg]

Kaffir lime Key lime

1 a-Pinene 0.995 1.08 0.36 3.07± 0.03 1.04± 0.012 Limonene 0.994 1.22 0.40 10.78± 0.17 50.5± 2.13 Citronellal 0.992 0.25 0.49 <LOQ 0.55± 0.014 Aromadendrene 0.999 0.47 0.15 1.00± 0.07 <LOQ5 Camphene 0.996 1.00 0,33 4.86± 0.67 3.38± 0.346 Linalool 0.990 1.69 0.56 20.13± 0.71 3.45± 0.097 Nerol 0.990 1.67 0.55 <LOQ 2.77± 0.108 trans-Geraniol 0.991 1.53 0.51 <LOQ 1.86± 0.109 b-Pinene 0.995 1.11 0.37 <LOQ 2.10± 0.1110 Terpinen-4-ol 0.997 0.86 0.28 44.79± 1.09 1.96± 0.0611 Myrcene 0.994 1.24 0.41 22.36± 0.95 24.89± 0.9612 g-Terpinene 0.993 1.32 0.44 25.01± 0.28 19.41± 0.6913 a-Terpineol 0.996 0.98 0.32 1.50± 0.06a 0.68± 0.02a

14 Citral 0.990 1.64 0.54 <LOQ 20.91± 0.60

LOQ - limit of quantitation, LOD-limit od detection, SD - standard deviation,Mean± SD of 3 measurements, Averages in rows marked with the same letters notdiffer significantly (P� 0.05), LOQ and LOD were calculated of the materiality levela¼ 0,05.

M. Lubinska-Szczygieł et al. / Food Control 89 (2018) 270e279 275

and Kaffir limes aroma were selected. It was chosen that 20-folddifference in content was used as a criterion for qualifying achemical compound to a group of potential indicators. In the case ofKaffir lime, Terpinen-4-ol was selected as a marker, with a contentof 44.79mg/kg. This is a chemical compound with a characteristicwoody aroma. In the case of Key lime, Citral is an indicator withfresh citrus scent. The presence of the two chemical compoundsmainly determines the smell of the fruits. Although Limonene'smost important ingredient in Key Lime was not chosen as a flavorfactor, because it is the most abundant chemical compound inmany citrus juices (Moufida et al., 2003). Citral was repeatedlylisted as one of the Key lime component (Cruz-Valenzuela, Tapia-Rodriguez, Vazquez-Armenta, Silva-Espinoza, & Ayala-Zavala,2015). The content of Citral in lime up to 5% was found and this

refers to essential oil (Costa et al., 2014; Spadaro et al., 2012). Theresults of quantitative determination of these chemical compoundsin lime juices have not been done so far. A quantitative analysis of 5citrus fruits shows that in each case Limonene is a major compo-nent, which confirms the correctness of the research. Key lime andKaffir lime were not found among the fruits which were tested andCitral content was not determined (Moufida et al., 2003).

The chromatographic peak area was used as an input data forchemometric analysis. Data set was preprocessed before classifi-cation. Preprocess was based on center by mean and scaling usingstandard deviation SD. Orange Canvas was used for training of fourclassifiers and for performance cross validation evaluation. Table 4shows the confusion matrices and evaluation results for usedclassification models, for results obtained by chromatographictechnique. Proportion of predicted are presented in confusionmatrices. All samples were correctly classified when 3 statisticalmodels were used: SVM, NB and RF. CT and RF models are based onsimple decision trees. SVM and NB models use more complex al-gorithms and therefore perform better. Based on results, the su-periority of the RF model over CT can be observed. This may be dueto the fact that CT is a single decision tree, while RF is a collection ofdecision trees. Table 4 with evaluated results contains informationabout area under curve (AUC), accuracy (CA), F1 score, precisionand recall (sensitivity). When values of all described parameters areequal 1.00, the model is perfect. These results were achieved inthree cases: SVM, NB and RF. In summary, the combination ofGC�GC-TOF-MS technique with chemometric analysis may be auseful tool for classifying lime juice in terms of botanical origin. Themost reliable results were achieved when 3 statistical models wereused (SVM, NB and RF).

3.2. Antioxidant properties of juices

The results of bioactive compounds in the juices of two limecultivars are presented in Table 5. In order to compare the changesin the amount of bioactive compounds, polyphenols and antioxi-dant capacities their values were determined in the pulp of the

Table 4Data obtained by using GC�GC-TOF-MS and four statistical models.

Confusion matrix

SVM Predicted CT Predicted K L Σ LK Σ

Act

ual K 1.00 0.00 3

Act

ual K 1.00 0.25 3

L 0.00 1.00 3 L 0.00 0.75 3 Σ 3 3 6 Σ 2 4 6

NB Predicted RF Predicted K L Σ LK Σ

Act

ual K 1.00 0.00 3

Act

ual K 1.00 0.00 3

L 0.00 1.00 3 L 0.00 1.00 3 Σ 3 3 6 Σ 3 3 6

Evaluation results for classification models Method AUC CA F1 Precision Recall SVM 1.000 1.000 1.000 1.000 1.000 CT 0.875 0.833 0.857 0.875 0.833 NB 1.000 1.000 1.000 1.000 1.000 RF 1.000 1.000 1.000 1.000 1.000

K – Kaffir lime samples, L – Key lime samples, SVM – Support Vector Machine, CT – Classification Tree, NB – Naïve Bayes, RF – Random Forrest Classification, AUC – Area Under Curve, CA – Accuracy, F1 – F1 score

Table 5Bioactive substances per g dry weight (DW) in (A) water (W) and methanol (M)extracts and (B) vitamin C, anthocyanins, chlorophylls and carotenoids of Kaffir andKey limes juices.

Indices Kaffir limeW Kaffir limeM Key limeW Key limeM

Polyph,mgGAE 23.16± 2.18a 15.79± 1.34b 11.93± 7.65c 12.78± 1.32bc

Flavan, mgCE 91.48± 8.54a 83.16± 6.54b 64.44± 5.43bc 51.18± 5.14c

Flavon, mgCE 0.62± 0.05b 1.41± 0.14a 0.36± 0.05c 0.65± 0.04b

Tannins,mgCE 1.92± 0.19bc 12.91± 2.11a 0.78± 0.07c 9.12± 0.85b

ABTS,mMTE, 161.64± 8.54a 75.43± 6.32bc 57.49± 5.56c 61.05± 7.65b

FRAP, mMTE, 28.50± 2.43a 20.01± 2.23b 14.68± 1.32c 16.20± 1.65bc

CUPRAC,mMTE 124.28± 5.43a 57.13± 5.21b 27.96± 2.67c 30.97± 2.45bc

DPPH, mMTE 33.29± 3.21a 14.83± 1.32b 12.03± 1.18c 17.15± 1.45ab

Indices Kaffir lime Key lime

Vit C,mgAsc 1.74± 0.17a 2.44± 0.11b

Anthoc, mgCGE/kg 63.45± 5.15a 47.32± 4.11b

Chlor a, mg 468.50± 12.12a 432.16± 10.98a

Chlor b, mg 260.08± 7.65a 245.76± 5.32a

Chlor aþb,mg 729.00± 16.54a 677.92± 12.54a

XanþCarot, mg 217.11± 5.32a 195.43± 4.43a

Mean± SD (standard deviation) of 5 measurements. Averages in rows marked withdifferent letters differ significantly (P< 0.05).

M. Lubinska-Szczygieł et al. / Food Control 89 (2018) 270e279276

fruits and also in the fresh prepared juices. The differences were notsignificant and the decrease was about 1e2%. The polyphenols inwater extract of Kaffir lime were twice higher than in Key lime, butin methanol extract were nearly equal. The bioactive compounds inKaffir lime such as polyphenols, flavanols, flavonoids, and tanninswere in water extract higher than in methanol of about 1.5 times.The antioxidant values were in similar correlation in 4 appliedmethods and the highest values showed ABTS and CUPRAC assaysin water fraction (Table 5A). It means that most of bioactive sub-stances were in the water extract. Key lime samples showeddifferent picture. The amount of polyphenols, flavanols, flavonoidsand tannins were slightly higher in methanol extract than in water.The corresponding values of antioxidants were in direct correlationwith polyphenols. Anthocyanins, chlorophylls and carotenoidswere similar in two limes, accept only the amount of vitamin C inKey lime (Table 5B). The results presented in Table 5 can be

compared only with a few data from cited reports such as Damian-Reyna, Gonzalez-Hernandez, Maya-Yescas, de Jesus Cortes-Penagos, and Del Carmen Chavez-Parga (2017). Total phenolics,total flavonoids, and ascorbic acid contents in Mexican sweet lime(Citrus limetta) juice were determined at two commercial maturitystages. The results of the above shown indices differ from thepresented in this report, but the lime variety was different withpolyphenols (725 ± 9.14mg/L), flavonoids (45.91± 1.00mg/L), andascorbic acid (222± 16mg/L). Results indicated that Mexican citricfruits were good sources of antioxidant agents. The results werecompared with Barros, Ferreira, and Genovese (2012). Four citrusspecies (C. sinensis, cvs. Pera and Lima; C. latifolia Tanaka cv. Tahiti;C. limettioides Tanaka cv. Sweet lime and C. reticulate, cv. Ponkan)were characterized in relation to pulps and peels contents ofascorbic acid, total polyphenols and antioxidant capacities. Theantioxidant capacity of citrus was correlated both to vitamin C andphenolics. Aside from citrus pulps, the peels are also good sourcesof bioactive compounds andminerals, and can be explored for theirhealth promoting values in food products.

In our previous report (Arancibia-Avila et al., 2012) was shownthe influence of different time durations of thermal processing onberries quality.

The antioxidant activity only of berries subjected to thermalprocessing for 10 and 20min did not differ from the non thermallyprocessed, showing high correlation between the total poly-phenols, flavanols and the antioxidant activities (Arancibia-Avilaet al., 2012). In another report for preservation of bioactive com-pounds long cold storage of smoothie-type ‘Hayward’ kiwifruitbeverages was suggested (Park et al., 2016). Both of the treatmentsas thermal and cold can be applied in the future research on thefreshly prepared kiwifruit and berries juices, but also on limes.

3.3. NMR shifts in juices

NMR spectra of two limes cultivars are shown on Fig. 3. Theassignments of 1H spectrum of limes chloroform and methanolextracts were obtained, where only peaks of aromatic regionsappeared between 6 and 10 ppm. The spectra were similar for two

Fig. 3. Comparison of 1H NMR spectra of limes juices in aromatic region: A1, A2, methanol extracts of Key and Kaffir limes; B1, B2, chloroform extracts of Key and Kaffir limes.

M. Lubinska-Szczygieł et al. / Food Control 89 (2018) 270e279 277

cultivars, but some slight differences were estimated. In the case ofchloroform or methanol, mostly the sugars were the main com-pounds and the aromatic part was the minor (ppm). The mixture ofwater and methanol is usually used for extraction and subsequentNMR analysis (Fig. 3): for Key lime (Fig. 3A1: 9.17; 9.13; 7.93; 7.21;7.15; 7.00; 6.69; 6.59; 6.45; 6.35; 6.22; 6.16; 5.91) and for Kaffir lime(Fig. 3A2: 8.30; 8.18; 8.08; 7.94; 7.87; 7.73; 7.52; 7.37; 7.15; 7.00;6.81; 6.72; 6.48; 6.45; 6.40; 6.22; 6.16; 6.04; 5.95; 5.91) ppm. In twovarieties of limes the common shifts appeared at 7.93; 7.15; 7.00;6.45; 6.16 and 5.91 ppm. All cited literature of fruits was dealingonly with aliphatic parts of juices. The aromatic range is shown in

Balan, Nicolescu, Stavarache, Ciobanu, and Deleanu (2013). Most ofthe shifts were between 6.21 and 8.07 for orange juice: 6.65; 6.91;7.07; 7.21; 7.32; 7.62; 7.72; and for grapefruit from 6.21 till 7.75 withthe middle ones at 6.62; 6.92; 7.21; 7.38; 7.63 and 7.75 ppm. Withorange and Key lime is a common shift in methanol fraction of7.21 ppm, and for Kaffir lime at 8.08 ppm. For grapefruit juice theshifts are 6.21; 6.62; 6.91; 7.23; 7.38; 7.63; 7.75 ppm. Key limeshowed a couple shifts common with grapefruit. In chloroformextract the following shift numbers appeared in Key lime (Fig. 3B1:9.61; 7.22; 7.09; 6.52; 5.36; 5.24), Kaffir lime (Fig. 3B: 8.21; 8.16;7.60; 7.20; 7.15; 6.99; 6.95; 6.31; 6.28; 5.59; 5.36). Common shifts

M. Lubinska-Szczygieł et al. / Food Control 89 (2018) 270e279278

for two varieties were at 5.36 ppm. It is known that phenoliccompounds are poorly soluble in water, but not for lime samples(Table 5, the amount of phenolic substances). Less peaks appearedin chloroform than in methanol. This result is obtained on the basisof the amount of polypenols extracted with chloroform is less thanwithmethanol. Previously it was described howwere characterizeddifferent kiwifruit cultivars. Using the NMR approach, it has beenpossible to identify primary as well as secondary metabolites ofdifferent fruits such as grape, orange, apple juice, mandarin, kiwi-fruits, mango, black raspberry, melon, watermelon, blueberry, andpeaches (Abdul Hamid et al., 2017; Balan et al., 2013; Drzewieckiet al., 2016; Sobolev et al., 2015). In NMR-based analyses, theextraction procedure is probably the most critical step aimed to thequantitative transfer of themetabolites from the solidmatrix. Peaksin the aromatic region indicate naringin and hesperidin, the mainflavanone glycosides in citrus fruits and from 5.10 ppm showedglucose, sucrose, fructose, and formate (Maltese, Erkelens, van derKooy, Choi, & Verpoorte, 2012; Rosa et al., 2015).

4. Conclusions

In this paper a usefulness of two dimensional gas chromatog-raphy coupled with time-of-flight mass spectrometry and NMRshifts were proven to assess the botanical origin of Kaffir lime andKey lime juices. Regarding to this, possible falsification ofmentioned juices can be detected. GC�GC-TOF-MS allows for betterseparation of substances originating from complexmatrices than incase of using one-dimensional chromatography, because ofimproved resolution, increased peak capacity and unique selec-tivity. On the basis of obtained results potential markers of botan-ical origin of limes were selected, namely Terpinen-4-ol for Kaffirlime and Citral for Key lime. Presented solution can be also treatedas an alternative to other analytical techniques used for deter-mining juice adulteration, such as HPLC or isotopic measurements.Furthermore, antioxidant property assessment of these juicesshows that it can be a new natural source for everydayconsumption.

Acknowledgments

Project “Antioxidant Power Series as a tool rational design andassessment of health promoting properties of functional food basedon antioxidant phytochemicals” (number of the application 2014/14/A/ST4/00640) financed by National Science Centre, Poland in aprogramme “MAESTRO 6”.

The authors are thankful to Dr. Elena Katrich (School of Phar-macy, Hebrew University of Jerusalem) for her technical assistancein determination of antioxidant status in limes.

References

Abdul Hamid, N. A., Mediani, A., Maulidiani, M., Abas, F., Park, Y.-S., Leontowicz, H.,et al. (2017). Characterization of metabolites in different kiwifruit varieties byNMR and fluorescence spectroscopy. Journal of Pharmaceutical and BiomedicalAnalysis, 138, 80e91.

Al-Doghairi, M., El-Nadi, A., Elhag, E., & Al-Ayedh, H. (2004). Effect of Solenostemmaargel on oviposition, egg hatchability and viability of Culex pipiens L. Larvae.Phytotherapy Research, 18, 335e338.

Allan, K., Kelly, F. J., & Devereux, G. (2010). Antioxidants and allergic disease: A caseof too little or too much? Clinical and Experimental Allergy, 40, 370e380.

Apak, R., Guclu, K., Ozyurek, M., & Karademir, S. E. (2004). Novel total antioxidantcapacity index for dietary polyphenols and vitamins C and E, using their cupricion reducing capability in the presence of neocuproine: CUPRAC method.Journal of Agricultural and Food Chemistry, 52, 7970e7981.

Arancibia-Avila, P., Namiesnik, J., Toledo, F., Werner, E., Martinez-Ayala, A. L., Rocha-Guzm�an, N. E., et al. (2012). The influence of different time durations of thermalprocessing on berries quality. Food Control, 26, 587e593.

Balan, M., Nicolescu, A., Stavarache, C., Ciobanu, M., & Deleanu, C. (2013). Fast NMRjuice identification based on sugars and other plant metabolites from fruits.

Revue Roumaine de Chimie, 58, 175e182.Banerjee, K., Patil, S. H., Dasgupta, S., Oulkar, D. P., Patil, S. B., Savant, R., et al. (2008).

Optimization of separation and detection conditions for the multiresidueanalysis of pesticides in grapes by comprehensive two-dimensional gas chro-matographyetime-of-flight mass spectrometry. Journal of Chromatography A,1190, 350e357.

Barros, H. R., Ferreira, T. A., & Genovese, M. I. (2012). Antioxidant capacity andmineral content of pulp and peel from commercial cultivars of citrus fromBrazil. Food Chemistry, 134, 1892e1898.

Benzie, I. F. F., & Strain, J. J. (1996). The ferric reducing ability of plasma (FRAP) as ameasure of antioxidant power: The FRAP assay. Analytical Biochemistry, 239,70e76.

Bogusz Junior, S., Março, P. H., Valderrama, P., Damasceno, F. C., Aranda, M. S.,Zini, C. A., et al. (2015). Analysis of volatile compounds in Capsicum spp. byheadspace solid-phase microextraction and GC � GC-TOFMS. AnalyticalMethods, 7, 521e529.

Boyer, R. F. (1990). Isolation and spectrophotometric characterization of photo-synthetic pigments. Biochemical Education, 18, 203e206.

Brand-Williams, W., Cuvelier, M. E., & Berset, C. (1995). Use of a free radical methodto evaluate antioxidant activity. Food Science and Technology, 28, 25e30.

Capobiango, M., Mastello, R. B., Chin, S.-T., De Souza Oliveira, E., De, Z., Cardeal, L.,et al. (2015). Identification of aroma-active volatiles in banana Terra spirit usingmultidimensional gas chromatography with simultaneous mass spectrometryand olfactometry detection. Journal of Chromatography A, 1388, 227e235.

Cheng, G. W., & Breen, P. J. (1991). Activity of phenylalanine ammonialyase (PAL)and concentrations of anthocyanins and phenolics in developing strawberryfruit. Journal of the American Society for Horticultural Science, 116, 865e869.

Costa, R., Bisignano, C., Filocamo, A., Grasso, E., Occhiuto, F., & Spadaro, F. (2014).Antimicrobial activity and chemical composition of Citrus aurantifolia (Christm.)Swingle essential oil from Italian organic crops. Journal of Essential Oil Research,26, 400e408.

Cruz-Valenzuela, M. R., Tapia-Rodriguez, M. R., Vazquez-Armenta, F. J., Silva-Espinoza, B. A., & Ayala-Zavala, J. F. (2015). Lime (Citrus aurantifolia) oils. InV. Preedy (Ed.), Essential oils in food preservation, flavor and safety (pp. 531e537).London: Academic Press.

Damian-Reyna, A. A., Gonzalez-Hernandez, J. C., Maya-Yescas, R., de Jesus Cortes-Penagos, C., & Del Carmen Chavez-Parga, M. (2017). Polyphenolic content andbactericidal effect of Mexican Citrus limetta and Citrus reticulata. Journal of FoodScience & Technology, 54, 531e537.

Drzewiecki, J., Latocha, P., Leontowicz, H., Leontowicz, M., Park, Y.-S., Najman, K.,et al. (2016). Analytical methods applied to characterization of Actinidia arguta,Actinidia deliciosa, and Actinidia eriantha kiwifruit cultivars. Food AnalyticalMethods, 9, 1353e1366.

Dymerski, T., Namie�snik, J., Leontowicz, H., Leontowicz, M., Vearasilp, K., Martinez-Ayala, A. L., et al. (2016). Chemistry and biological properties of berry volatilesby two-dimensional chromatography, fluorescence and Fourier transforminfrared spectroscopy techniques. Food Research International, 83, 74e86.

Dymerski, T., Namie�snik, J., Vearasilp, K., Arancibia-Avila, P., Toledo, F., Weisz, M.,et al. (2015). Comprehensive two-dimensional gas chromatography and three-dimensional fluorometry for detection of volatile and bioactive substances insome berries. Talanta, 134, 460e467.

Feucht, W., & Polster, J. (2001). Nuclei b of plants as a sink for flavanols. Journal ofBioscience, 56, 479e481.

Ghafar, M., Prasad, K., Weng, K., & Ismai, A. (2009). Flavonoid, hesperidine, totalphenolic contents and antioxidant activities from Citrus species. African Journalof Biotechnology, 9, 326e330.

Kasuan, N., Muhammad, Z., Yusoff, Z., Fazalul Rahiman, M. H., Taib, M. N., &Haiyee, Z. A. (2013). Extraction of citrus Hystrix d.c. (Kaffir lime) essential oilusing automated steam distillation process: Analysis of volatile compounds.Malaysian Journal of Analytical Sciences, 17, 359e369.

Kupska, M., Chmiel, T., Jedrkiewicz, R., Wardencki, W., & Namie�snik, J. (2014).Comprehensive two-dimensional gas chromatography for determination of theterpenes profile of blue honeysuckle berries. Food Chemistry, 152, 88e93.

Leontowicz, H., Leontowicz, M., Latocha, P., Jesion, I., Park, Y.-S., Katrich, E., et al.(2016). Bioactivity and nutritional properties of hardy kiwifruit Actinidia argutain comparison with Actinidia deliciosa ‘Hayward’ and Actinidia eriantha ‘Bidan’.Food Chemistry, 196, 281e291.

Maltese, F., Erkelens, C., van der Kooy, F., Choi, Y. H., & Verpoorte, R. (2012). Iden-tification of natural epimeric flavanone glycosides by NMR spectroscopy. FoodChemistry, 116, 575e579.

Moufida, S., & Marzouk, B. (2003). Biochemical characterization of blood orange,sweet orange, lemon, bergamot and bitter orange. Phytochemistry, 62,1283e1289.

Ozyurek, M., Guclu, K., Bektasoglu, B., & Apak, R. (2007). Spectrophotometricdetermination of ascorbic acid by the modified CUPRAC method with extractiveseparation of flavonoids e La (III) complexes. Analytica Chimica Acta, 588,88e95.

Park, Y.-S., Ham, K.-S., Park, Y.-K., Leontowicz, H., Leontowicz, M., Namiesnik, J., et al.(2016). The effects of treatment on quality parameters of smoothie-type ‘Hay-ward’ kiwi fruit beverages. Food Control, 70, 221e228.

Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M., & Rice-Evans, C. (1999).Antioxidant activity applying an improved ABTS radical cation decolorizationassay. Free Radical Biology & Medicine, 26, 1231e1237.

Risticevic, S., Deell, J. R., & Pawliszyn, J. (2012). Solid phase microextraction coupledwith comprehensive two-dimensional gas chromatography-time-of-flight mass

M. Lubinska-Szczygieł et al. / Food Control 89 (2018) 270e279 279

spectrometry for high-resolution metabolite profiling in apples: Implementa-tion of structured separations for optimization of sample. Journal of Chroma-tography A, 1251, 208e218.

Rocha, S. M., Coelho, E., Zrostlíkov�a, J., Delgadillo, I., & Coimbra, M. A. (2007).Comprehensive two-dimensional gas chromatography with time-of-flight massspectrometry of monoterpenoids as a powerful tool for grape origin traceability.Journal of Chromatography A, 1161, 292e299.

Rosa, A., Atzeri, A., Deiana, M., Scano, P., Incani, A., Piras, C., et al. (2015). Comparativeantioxidant activity and 1H NMR profiling of mediterranean fruit products FoodResearch International, 69 pp. 322e330).

Rubel, D. M., Freeman, S., & Southwell, I. A. (1998). Tea tree oil allergy: What is theoffending agent? Report of three cases of tea tree oil allergy and review of theliterature. Australasian Journal of Dermatology, 39, 244e247.

Samykanno, K., Pang, E., & Marriott, P. J. (2013). Chemical characterisation of twoAustralian-grown strawberry varieties by using comprehensive two-dimensional gas chromatography-mass spectrometry. Food Chemistry, 141,1997e2005.

Schmarr, H. G., & Bernhardt, J. (2010). Profiling analysis of volatile compounds fromfruits using comprehensive two-dimensional gas chromatography and imageprocessing techniques. Journal of Chromatography A, 1217, 565e574.

Sharon-Asa, L., Shalit, M., Frydman, A., Bar, E., Holland, D., Or, E., et al. (2003). Citrusfruit flavor and aroma biosynthesis: Isolation, functional characterization, anddevelopmental regulation of Cstps1, a key gene in the production of thesesquiterpene aroma compound valencene. The Plant Journal, 36, 664e674.

Singleton, V. L., Orthofer, R., & Lamuela-Raventos, R. M. (1999). Analysis of totalphenols and other oxidation substrates and antioxidants by means of Folin-Ciocalteu reagent. Methods in Enzymology, 299, 152e158.

Sobolev, A. P., Mannina, L., Proietti, N., Carradori, S., Daglia, M., Giusti, A. M., et al.

(2015). Untargeted NMR-based methodology in the study of fruit metabolites.Molecules, 20, 4088e4108.

Spadaro, F., Costa, R., Circosta, C., & Occhiuto, F. (2012). Volatile composition andbiological activity of key lime Citrus aurantifolia essential oil. Natural ProductCommunications, 7, 1523e1526.

Srisukh, V., Bunyapraphatsara, N., Pongpan, A., Tungrugsasut, W.,Puttipipatkhachorn, S., Oniam, W., Karawamitr, T., Bunsiriluk, S., &Thongbainoi, W. (2012a). Fresh produce antibacterial rinse from Kaffir lime oil.Journal of Pharmaceutical Sciences, 39, 15e27.

Srisukh, V., Tribuddharat, C., Nukoolkarn, V., Bunyapraphatsara, N.,Chokephaibulkit, K., Phoomniyom, S., et al. (2012b). Antibacterial activity ofessential oils from Citrus hystrix (makrut lime) against respiratory tract path-ogens. ScienceAsia, 38, 212e217.

Thanaboripat, D., Chareonsettasilp, S., & Pandee, K. (2006). Inhitory effect of Kaffirlime, Bitter cucumber and Tobacco extracts on the growth of Aspergillus flavus.Science Technology Journal, 6, 18e24.

Waikedre, A., Dugay, J., Barrachina, I., Herrenknecht, C., Cabalion, P., & Fournet, A.(2010). Chemical composition and antimicrobial activity of the essential oilsfrom new caledonian citrus macroptera and citrus hystrix. Chemistry andBiodiversity, 7, 871e877.

Williams, A., Ryan, D., Olarte, G. A., Marriott, P., & Pang, E. (2005). Analysis ofstrawberry volatiles using comprehensive two-dimensional gas chromatog-raphy with headspace solid-phase microextraction. Journal of ChromatographyB: Analytical Technologies in the Biomedical and Life Sciences, 817, 97e107.

Zrostlíkov�a, J., Haj�slov�a, J., & Cajka, T. (2003). Evaluation of two-dimensional gaschromatographyetime-of-flight mass spectrometry for the determination ofmultiple pesticide residues in fruit. Journal of Chromatography A, 1019, 173e186.