Food Bioprocess Technol (2017) 10:1142–1153

DOI 10.1007/s11947-017-1888-1

ORIGINAL PAPER

Stabilization of Refrigerated Avocado Pulp:

Chemometrics-Assessed Antibrowning Allium and Brassica
Extracts as Effective Lipid Oxidation Retardants
M. C. Bustos 1 & M. F. Mazzobre 1 & M. P. Buera 1

Received: 28 March 2016 / Accepted: 9 February 2017 / Published online: 7 March 2017
# Springer Science+Business Media New York 2017

Abstract The effectiveness of Allium and Brassica extracts Introduction

to inhibit the evolution of lipids oxidation in avocado pulp
under refrigeration (storage at 4 °C) was studied. Onion, gar- From a nutritional point of view, avocado (Persea americana
lic, scallion, white cabbage, cauliflower, and Brussels sprouts Mill) is a highly caloric fruit, usually consumed in salads or
extract were tested as preserving agents in refrigerated avoca- prepared as guacamole. Avocado is a highly perishable fruit
do pulp. Allium extracts promoted almost a 60% retention of with a high metabolic rate resulting in a shelf-life of only 3–
the intrinsic anti-radical capacity of the pulps. Considering 4 weeks when the entire fruit is stored at the optimum temper-
secondary oxidation effects, extinction coefficient at 270 nm ature and relative humidity (Yahia and González-Aguilar
shows that all treated pulps (except those with scallion ad- 1998). The shelf-life of avocado pulp (about 5 days) is quite
dition) were acceptable at the 30th storage day short compared to that of the entire fruit, and is severely de-
(K270 < 0.22), but they were all significantly less oxidized termined by oxidative processes, which affect both lipid and
than the untreated samples (K270 = 1.8) (P < 0.05). Garlic- aqueous fractions (Balda et al. 2011). That means, control of
treated avocado showed the highest antioxidant effective- enzymatic browning during processing and storage of avoca-
ness, based on C=CH cis proportion (Icis = 108.3), while do is crucial to preserve the original appearance of fresh fruit.
samples with white cabbage extract presented the highest In addition, the quality decay promoted by enzyme mediated
C=CH trans (Itrans = 5.7) proportion after 30 days. The oxidative reactions, takes place especially in the aqueous frac-
PCA method was discriminant enough since 83.6% of the tion, where phenolic substrates are hydroxylated and then ox-
variance was explained by the first two principal compo- idized giving place to the formation of brown compounds
nents, allowing the samples to be grouped according to stor- (Robards et al. 1999).
age time and extract type. This study confirmed that the We have recently demonstrated that Allium and Brassica
addition of garlic, onion, and cauliflower extracts enhanced vegetables extracts inhibit browning of mushrooms and avo-
lipid antioxidant properties in refrigerated avocado pulps. cado slices (Bustos et al. 2014). Moreover, we studied the
effect of the same extracts on avocado pulp (Bustos et al.
2015) where we observed the inhibition of enzymatic brow-
Keywords Allium . Brassica . Lipid oxidation . Avocadopulp ning. Both species of vegetables have been also associated
with high antioxidant content that are known to scavenge free
radicals and singlet oxygen (Podsedek 2007; Prakash et al.
2007; Sikora et al. 2008). Regarding this, many antioxidants
are known to reduce hydroperoxides and prevent rancidity or
off-flavors in edible oils (Laguerre et al. 2007). Only few
* M. P. Buera
[email protected]
research works have taken into consideration the effect of
the applied antibrowning treatment on the lipid fraction stabil-
1
Dpto. de Industrias, Facultad de Ciencias Exactas y Naturales,
ity (Elez-Martínez et al. 2005; Elez-Martínez et al. 2007; Plaza
Universidad de Buenos Aires, Inte. Güirandes 100, et al. 2009). However, all of them applied lipid extraction
C1428EGA Buenos Aires, Argentina processes using high temperatures or solvents that clearly
Food Bioprocess Technol (2017) 10:1142–1153 1143

affect oil quality and lipids characteristics. In this regard, activity, centrifuged at 15585 g for 30 min at 4 °C (centrifuge
Moreno et al. (2003) found that avocado oil extraction with 5804 R, Eppendorf, Germany), and filtered on filter paper
solvents (hexane and acetone) increased trans fatty acids (20–25 μm, Whatman ECN-512-1026). Trehalose was added
proportion and Elez-Martínez et al. (2005) stated that heat to the liquid extracts at a final concentration of 15% w/v in
treatment was detrimental to the quality of avocado be- order to obtain a physically adequate and stable dry matrix
cause several undesirable reactions that turn into browning, (Schebor et al. 2010). In addition to the chemical inertness
flavor damage, and/or nutritional losses are accelerated as of this sugars, it has the particularly ability to form a glass
increasing temperature. with a high glass transition temperature compared to other
Allium and Brassica vegetables have been proposed as nat- disaccharides, which prevents the collapse of the freeze-
ural antioxidants sources. Onion and garlic have been reported dried matrix. This fact allows improving the handling, distri-
to inhibit oxidation as hydroxyl radical scavengers in several bution, and stability of the dried extract. Aliquots of the ex-
meat products while there is not such evidence for scallion tracts were distributed in plastic trays and frozen at −20 °C
(Aguirezábal et al. 2000, Tang and Cronin 2007; Yang et al. for 48 h, further cooled under liquid nitrogen and freeze-dried
2011a, b). The antioxidant activity of Allium plants was main- (ALPHA 1-4 LD2 Martin Christ Gefriertrocknungsanlagen
ly been attributed to a variety of sulfur-containing compounds GMB, Germany) for 48 h. Since we used 10% of the
and their precursors. Regarding Brassica vegetables the lipid freeze-dried extracts for obtaining the antioxidant effect, for
antioxidant capacity has been associated to glucosinolates each 100 g of avocado pulp we would need about 6 g of the
presence (Raitio et al. 2011; Jaiswal et al. 2011; Abu- raw vegetables, considering that 15% trehalose was added to
Ghannam and Jaiswal 2015). the extracts.
The oil content of avocado could reach as high as 15–30%
depending on the variety (Werman and Neeman 1987). Avocado Pulp Preparation and Storage Conditions
Avocado oil is highly valued because it has a high concentra-
tion of oleic monounsaturated fatty acids (representing almost Avocado var. Hass pieces of uniform size and color were
70% of total fatty acids) (Moreno et al. 2003). In regard, av- purchased at commercial maturity in a local market and im-
ocado is a bioactive-rich source of antioxidant vitamins, ca- mediately processed as previously reported (Bustos et al.
rotenoids, phenols, and phytosterols (Berasategi et al. 2012; 2015). Briefly, fruits were peeled and the pulp was pieced
Ding et al. 2007). and ground with a food processor HR 1372, 700 W (Philips
Nevertheless, to date, no research has integrated the Electronics, Slovenia). Avocado puree without any addition
study of the effect of minimal processing with the addition was used as control (untreated), and an additional control
of vegetable extracts (added during refrigeration storage sample treated with citric acid (CA) to decrease the pulp pH
with the aim to inhibit browning), with the lipid oxidation to 4.0 according to Soliva-Fortuny et al. (2001) was used. In
processes during storage. Therefore, in the present work our previous work we established that the addition of 10 w/w
the effectiveness of Brassica and Allium extracts to retard of vegetables extracts obtained in the Section BAntibrowning
the oxidation of avocado pulp lipid fraction (obtained un- Allium and Brassica Extracts Preparation^ was enough to
der refrigerated conditions), and its relationship with anti- promote an inhibitory effect on avocado puree browning
oxidant capacity were analyzed. (Bustos et al. 2015), also the addition of citric acid was nec-
essary to reach long self-life of puree. That means, citric acid
and 10% w/w of vegetables extracts were added to avocado
Materials and Methods pulps resulting in samples with garlic (G), onion (O) and scal-
lion (S), white cabbage (Wc), cauliflower (C), or Brussels
Antibrowning Allium and Brassica Extracts Preparation sprouts (Bs) addition. A reference antibrowning agent
(200 ppm ascorbic acid—AO) was also used for comparison
Fully mature vegetables from Allium family (garlic (Allium according to (Soliva-Fortuny et al. 2001). After homogeniza-
sativum), onion (Allium cepa), and scallion (Allium tion, avocado purees were distributed in polyethylene bags
fistulosum)) and from Brassica family (white cabbage (50 g) and lipid oxidation and antioxidant capacity were eval-
(Brassica oleracea var. viridis), cauliflower (Brassica uated on each aliquot, during storage at 4 °C for at least
oleracea var. botrytis), and Brussels sprouts (Brassica 30 days, which was an optimal period according to previous
oleracea var. gemmifera)) were washed, peeled when neces- results (Bustos et al. 2015).
sary (garlic and onion), and cut into pieces. The pieces were
homogenized and extracted under constant agitation in phos- Lipid Extraction from Avocado Pulp
phate buffer solution pH 6.0 (ratio 1:2 w/v) at 50 °C for 1 h as
previously described (Bustos et al. 2015). Briefly, the extracts At each evaluation time, 25 g of avocado pulp were mixed
were sterilized at 121 °C for 5 min to avoid further enzyme with 25 ml of McIlvaine buffer solution at pH 6.5 (1 M NaCl
1144 Food Bioprocess Technol (2017) 10:1142–1153

and 5% polyvinylpolypyrrolidone were added to buffer). The with hexane and allowed to dry at ambient temperature.
homogenate was centrifuged at 2000 g for 30 min at 4 °C Spectra acquired were processed using the Spectrum software
(centrifuge 5804 R, Eppendorf, Germany). Three phases were version 6.3.5 (PerkinElmer, Inc.). The signal heights of each
obtained after centrifugation: a hydrophobic liquid phase on spectrum at 3006, 2924, 2854, 1755, 1654, 1464, 1417, 1377,
the top (avocado oil), an insoluble phase (solid phase) in the 1240, 1163, 1119, 1099, 983, 967, and 723 cm−1 were record-
medium, and a hydrophilic liquid phase at the bottom (enzy- ed from each sample. The collected spectra were used for
matic aqueous extract) of the centrifuge tube, as was found by multivariate analysis.
Saha et al. (2013). The aqueous fraction (previously To compare oxidative states of the analyzed lipid fractions,
characterized by Bustos et al. 2015) was used for antioxidant two indices have been calculated according to Dréau et al.
capacity determination. The lipid fraction stability studies (2009) with the difference that we used the absorbance
were performed at 0 and 30 days of storage at refrigeration heights. The Icis was the index of C=CH cis decrease (Eq. 2)
conditions. and the Itrans the index of C=CH trans change (Eq. 3) that are
the most significant changes in oil aging.
Stability of the Lipid Fraction A 
3006
=ðA3006 þA2924 Þ T t

I cis ¼ A3006  ð2Þ
=ðA3006 þA2924 Þ T 0
Specific Extinction Coefficients at 232 nm (K232) and 270 nm  
(K270) A970
I trans ¼ Tt ð3Þ
A1465 þ A970
Specific extinction coefficients were calculated at 232 and
270 nm, in the lipid fraction. Briefly, 20 μL of extracted avo- Where Aλ is the absorbance read in the spectrophotometer
cado lipids were weighted and diluted to 10 ml with hexane, at each wavelength (λ). The wavelength selected were: 3006
after 5 min of agitation the absorption spectrum was recorded. corresponding to cis CH band, 2924 corresponding to -C-H
The following equation was employed according to Paz (CH2) stretching, 970 corresponding to -HC=CH- (trans)
Antolín & Molero Meneses (2000): bending, and 1465 corresponding to -C-H (CH2, CH3) bend-
ing. Indexes were calculated at 0 and 30 days of refrigeration.
Aλ
Kλ ¼ ð1Þ
C
Antioxidant Capacity of the Aqueous Fraction
Where Kλ is the specific extinction coefficient for each
wavelength (λ), Aλ is the absorbance read in the spectropho- Radical scavenging activity (ABTS) was measured by ABTS
tometer at each wavelength (λ), and C is the concentration of method following the procedure describes by Re et al. (1999).
lipids in g/100 ml. Specific extinction coefficients were calcu- Briefly, ABTS radical cation (ABTS+) was produced by
lated at 0 and 30 days of refrigeration. reacting 7 mM ABTS reactive (CAS 30931-67-0, Sigma-
Aldrich) with 2.45 mM potassium persulphate in the dark at
room temperature for 12 h before use. The ABTS+·solution
Attenuated Total Reflectance–Fourier Transform-Infrared
was diluted 1:2 with distilled water in order to obtain an ab-
Spectroscopy Analysis
sorbance of 0.700 at 734 nm. After addition of 50 μL of
aqueous fraction to 3.0 ml of diluted ABTS+ solution, absor-
Spectra of the lipid fractions were collected at room tempera-
bance readings were taken during 30 min using a spectropho-
ture using a FTIR spectrometer model spectrum 400
tometer UV-Visible Jasco V630 (Jasco Corporation, Japan) at
(PerkinElmer, Inc., Shelton CT, USA) equipped with a deu-
room temperature.
terated triglycine sulfate detector (DTGS). Each lipid fraction
Absorbance vs. time was plotted and bi-exponential fit was
extracted from avocado pulp samples were studied in a
applied using ORIGIN 8.0 software (Origin-Lab Corporation,
MIRacle single-reflection attenuated total reflectance (ATR)
Northhampton, MA, USA) and plateau value was considered
accessory (PIKE Technologies, Inc., Madison WI, USA) with
infinite absorbance (A∞m). ABTS radical scavenging rate was
a single-reflection diamond/ZnSe crystal at an incident angle
calculated as the following equation:
of 45°. The spectrum of each sample was obtained by taking


the average of 64 scans at a resolution of 4 cm−1, at 25 °C. The ABTS radical scavenging rate ð%Þ ¼ 100* 1− A∞ m =A0 ð4Þ
spectra were acquired between 550 and 4000 cm−1. A back-
ground spectrum was recorded in air (without sample) prior to Where A0 is the absorbance at time 0 without sample addi-
each spectrum measurement. Lipid fraction from avocado tion and A∞m is the sample absorbance when the stationary
pulp samples were directly applied to the ZnSe ATR accessory state, calculated through bi-exponential fitting, was reached.
and the crystal was thoroughly cleaned between each sample Ferric reducing ability (FRAP) of aqueous fraction was
by removing the previous sample with tissue and cleaning it determined by FRAP assay according to Pulido et al. (2000)
Food Bioprocess Technol (2017) 10:1142–1153 1145

using gallic acid as standard. Briefly, 900 μL of FRAP reagent oxidation products (Sinelli et al. 2007). The European
(2.5 ml of a tripyridyl-s-triazine (TPTZ) solution (10 mmol/L) Commission (Council Regulation N°1513/ 2001) as so as
in hydrochloric acid (40 mmol/L) and 2.5 ml of a FeCl3 solu- the Argentinean Codex (CAA 2012) established the limits of
tion (20 mmol/L) mixed with 25 ml of an acetate buffer olive oil oxidation as K232 ≤ 2.4 and K270 ≤ 0.22; that could be
(0.3 mol/L, pH 3.6), prepared freshly and warmed at 37 °C, taken as reference for avocado lipid fraction because of the
was mixed with 90 μL of distilled water and 30 μL of aqueous similarities in the fatty acid composition (Berasategi et al.
supernatant or water for the reagent blank. Readings were 2012). The products of primary oxidation, peroxides, and hy-
taken at the absorption maximum (595 nm) using a UV- droperoxides, absorb at 232 nm, while secondary oxidation
Visible spectrophotometer Jasco V630 (Jasco Corporation, products like aldehydes, ketones, and acids absorb at higher
Japan) equipped with a thermostatized auto-cell-holder. wavelengths (~270 nm) with dienes and trienes formed during
Temperature was maintained at 37 °C for up to 30 min. processing (Elez-Martínez et al. 2005; Paz Antolín and
Molero Meneses 2000).
Statistical Analysis The specific extinction coefficient K232 of lipid fractions
extracted from avocado pulp increased similarly in all sam-
All samples were prepared in duplicate and each replicate was ples, being the coefficients for ascorbic acid (AO), scallion
quantified in duplicate. (S), and Brussels sprouts (Bs) treated pulps acceptable during
Statistical analysis was conducted by using one-way anal- the 30 days evaluated (Fig. 1a). In addition, scallion-treated
ysis of variance (ANOVA), followed by the DGC means com- samples presented the lower value at beginning of the refrig-
parison test (Di Rienzo et al. 2002) to assess any differences eration storage, with no significant differences with untreated,
between group means using Infostat software (Di Rienzo et al. citric acid (CA) or AO avocado pulp (P > 0.05).
2012). This allowed the samples to be grouped according to Considering the secondary oxidation, K270 showed a great
descending levels of preference (A, B, and C) and with a inhibitory effect of vegetables extracts on secondary products
degree of significance of P = 0.05. formation in the samples, in comparison with those untreated,
Principal component analysis (PCA) and hierarchical clus- CA or AO samples (Fig. 1b). All avocado pulps with extracts
ter analysis (HCA) were also performed using the same soft- addition were acceptable at day 0, except for scallion-treated
ware. PCA was employed in order to analyze all data sets samples, which presented values higher than the limits
(corresponding to chemical determinations and to FTIR spec- established by legislation but significantly lower than the ob-
tra) for structuring the data and for reducing the dimensional- served in untreated, CA or AO pulp (P < 0.05). These obser-
ity of the spectral data. All replicates were used in calcula- vations indicate that the lipid fraction of these products
tions, but employed software shows average points for each underwent secondary oxidation and were unacceptable. In this
sample in final biplot. Cluster analysis or clustering is the task regard, garlic treatment (G) generated values slightly higher
of grouping a set of objects in such a way that objects in the than the limit of acceptability at 30 days of storage (Fig. 1b),
same group (called a cluster) are more similar (in some sense indicating that moderate amounts of vegetable extracts can
or another) to each other than to those in other groups (clus- effectively delay rancidity of avocado lipids. While we ob-
ters). It is a main task of exploratory data mining, and a com- served acceptable K270 values for most of the vegetable treated
mon technique for statistical data analysis. For cluster analy- samples during storage, Elez-Martínez et al. (2005) reported
ses, the variables included in the principal component analysis considerably higher and unacceptable values in oil extracted
were used and the Euclidean distance selected as a dissimilar- from Hass avocado with antioxidant addition. This discrepan-
ity distance. The number of clusters with the highest coeffi- cy is probably due to the temperatures higher than 50 °C
cient was finally decided as being the optimal (Bock 1996). employed for the extraction processes by those authors. In
the present work, the lipid fractions have been obtained at
4 °C in order to maintain the lipids properties, and the rela-
Results and Discussion tionship between their stability and the pulp treatments can be
performed.
Stability of the Lipid Fraction
Attenuated Total Reflectance–Fourier Transform-Infrared
Specific Extinction Coefficient at 232 nm (K232) and 270 nm (ATR-FTIR) Spectroscopy Analysis
(K270)
The spectra of the lipid fractions showed well resolved bands
One of the most common routine methods to evaluate the oil that could be assigned to different functional groups present in
deterioration is the determination of absorbance at 232 (K232) the lipid fraction (Table 1). Figure 2 depicts the spectrum of
and at 270 nm (K270) that measures the formation of conju- lipid fraction from untreated avocado pulp at day 0 in the
gated dienes and trienes due to the formation of secondary region of 4000–550 cm−1.
1146 Food Bioprocess Technol (2017) 10:1142–1153

Fig. 1 Extinction coefficient at 232 (a) and 270 nm (b) obtained for lipid letters differ significantly (p < 0.05). The horizontal lines represent the
fraction from avocado pulp samples with citric acid (CA), ascorbic acid limits established by the European Commission for the acceptability of
(AO), garlic (G), onion (O), scallion (S), white cabbage (Wc), cauliflower olive oil
(C), or Brussels sprouts (Bs) extract addition. Columns with different

During storage of avocado pulp under refrigeration lipid disappearance in the vibrations attributed to cis double bonds
fraction oxidation occurred, as reflected by the extinction co- (a decrease in the bands at 3006 and 1654 cm−1) and by isom-
efficients described in the previous section. Free radicals erization forming trans double bonds (changes in band at
formed during the initiation step react with oxygen to form 961 cm−1). We also noted the accumulation of CO2 (band
peroxy radicals, which remove an allylic hydrogen adjacent to 1755 cm−1), which could be a byproduct of the oxidation of
a double bond or combine to form a hydroperoxide. During the triglycerides. Considering these changes on cis and trans
the propagation phase, the newly formed alkoxy radicals can double bonds two indices were calculated, as proposed by
subsequently extract hydrogen or add to double bonds. Sinelli et al. (2007) and Dréau et al. (2009): Icis (which indi-
Peroxides accumulate, decompose, and produce secondary cates the decrease of the cis double bounds) and Itrans (which
oxidation products. Ketones, esters, and alcohols are formed indicates the proportion of the trans double bounds), which
as oxidation products. Isomerization of the double bond from are shown in Table 2.
cis to trans also occurs (Dréau et al. 2009). The spectrum of Due to the employed storage conditions (low temperature
the lipid fraction from avocado samples (Fig. 2) showed no and darkness), both indexes (Icis and Itrans) showed slight
accumulation of hydroxyl groups (absence of band at changes in the lipid fraction of avocado pulp. However, it
3530 cm−1). This feature was accompanied by the gradual can be observed that AO, garlic, scallion, and cauliflower
treated samples presented the higher Icis values, indicating that
Table 1 Peak assignment of avocado lipid fraction FTIR spectrum lipid fraction from these avocado pulps underwent a lower
degree of isomerization than white cabbage and Brussels
Peak (cm−1) Peak assignment
sprout-treated samples, which presented values even lower
3006 =C-H (cis) stretchinga,b than untreated or CA avocado. Similar Itrans values were found
2924 -C-H (CH2) stretching asyma,b,c for all lipid fractions, being Brassica extracts treated pulps the
2853 -C-H (CH2) stretching syma,b,c ones that presented high trans bonds generation in lipid frac-
1745 -C=O stretchinga,b,c tion, while untreated, CA and AO presented the lower values,
1654 -C=C- (cis) olefins stretchinga,c being Allium samples the one with an intermediate behavior.
1464 -C-H (CH2, CH3) bendinga,c While the Icis index reflected some differences in the samples
1417 =C-H (cis) bendinga,c treated or not with the different extracts, the Itrans index was
1377 -C-H (CH3) bendinga,c not sensitive enough to discriminate the treatment effect.
1240 -C-O, -CH2- stretching, bendinga,b,c
1163 -C-O, -CH2- stretching, bendinga,b,c Antioxidant Capacity of Aqueous Fraction of Avocado
1119 -C-O stretching a,c Pulps
1099 -C-O stretchinga,b
983 -HC=CH- (trans) bendingb,c Table 3 shows the ABTS radical scavenging rate (ABTS) and
967 -HC=CH- (trans) bendinga,b,c reducing capacity (FRAP) of aqueous fraction obtained from
723 -(CH2)n-, -HC=CH- (cis) bendinga avocado pulp treated or not with Allium and Brassica extracts
and conserved at 4 °C.
a
Guillén and Cabo (1997, 2000) Aqueous fraction extracted from untreated avocado pulp
b
Saha et al. (2013) showed a great antioxidant capacity. This result is in agree-
c
Quiñones-Islas et al. (2013) ment with that from Plaza et al. (2009) who stated that the
Food Bioprocess Technol (2017) 10:1142–1153 1147

Fig. 2 FTIR spectrum obtained

for lipid fraction extracted from
untreated avocado pulp

antioxidant activity of avocado is provided by hydrophilic It is remarkable that in the AO samples the antioxidant
phytochemicals, and they found no antioxidant effect of oil activity was almost completely conserved. Allium extracts
fraction (fatty acids and sterols). However, during storage, added to the pulps promoted a high retention of anti-radical
untreated avocado pulp presented an antioxidant activity loss capacity (almost 70% conserved). In spite of the great
of almost 50% after 30 days of storage, considering both anti- antibrowning effect previously observed for the scallion-
radical capacity and reducing power. Avocado pulp treated treated samples (Bustos et al. 2015), a great loss of reducing
with citric acid, as so as the untreated pulp showed the same power was observed during storage, which was similar to that
antioxidant capacity loss, considering both ABTS and FRAP of the citric acid (CA) system (Table 3). Brassica extracts,
methods (P > 0.05). Besides, the samples with onion and promoted different effects: in contrast to Allium, cabbage af-
cauliflower extracts addition showed a similar anti-radical ac- fected both capacities at the same level, while cauliflower
tivity at 30 days of avocado pulp storage under refrigeration, conserved a high level of both antioxidant indicators
but significantly lower than the ascorbic acid containing (AO) (Table 3). Recently, some researchers found that the inclusion
sample. On the other hand, Wc pulps presented the highest of a Brazilian fruit with high antioxidant activity promote a
anti-radical capacity loss (P < 0.05). The reducing power de- mayor antioxidant activity retention during storage of cheese
creased until no significantly different values in all untreated (Pereira et al. 2016a, b).
and treated avocado pulps, except for cauliflower extract that
conserved a similar reducing power to AO sample (Table 3). Table 3 Anti-radical scavenging rate (ABTS) and reducing power
(FRAP) of aqueous fraction from avocado pulp at 0 and 30 days of
refrigeration
Table 2 Indexes based on C=CH cis proportion (Icis) and the C=CH
Samplea ABTS (%) FRAP (mg GA/100 ml)
trans (Itrans) proportion in the lipid fraction of avocado pulp samples after
30 days of storage at 4 °C
Day 0 Day 30 Day 0 Day 30
Samplea Icis (×10−2) Itrans (×10−2)
Untreated 90.5 ± 0.7b 49.2 ± 0.4h 0.97 ± 0.07c 0.41 ± 0.01f
Untreated 98.9 5.0 CA 92.0 ± 0.8b 49.2 ± 1.4h 1.23 ± 0.10b 0.36 ± 0.04f
CA 97.5 5.1 AO 77.0 ± 0.9d 69.2 ± 1.2e 0.66 ± 0.02e 0.63 ± 0.03e
AO 101.7 4.6 G 84.6 ± 0.0c 55.9 ± 0.7g 0.89 ± 0.05d 0.41 ± 0.01f
G 108.3 5.1 O 87.7 ± 0.9b 62.1 ± 1.5f 1.03 ± 0.06c 0.40 ± 0.02f
O 97.7 4.9 S 89.8 ± 0.8b 59.8 ± 2.0f 1.09 ± 0.10c 0.35 ± 0.04f
S 100.6 5.1 Wc 77.9 ± 0.2d 26.4 ± 1.2i 0.97 ± 0.04c 0.34 ± 0.00f
Wc 92.3 5.2 C 93.1 ± 0.4b 62.2 ± 1.5f 0.86 ± 0.00d 0.65 ± 0.03e
C 101.6 5.3 Bs 93.1 ± 0.4b 47.9 ± 2.0h 1.03 ± 0.08c 0.46 ± 0.00f
Bs 93.7 5.7
Within the same parameter different letters differ significantly (p < 0.05)
a a
Avocado pulps with citric acid (CA), ascorbic acid (AO), garlic (G), Avocado pulps with citric acid (CA), ascorbic acid (AO), garlic (G),
onion (O), scallion (S), white cabbage (Wc), cauliflower (C), or onion (O), scallion (S), white cabbage (Wc), cauliflower (C), or Brussel
Brussels sprouts (Bs) extract addition sprouts (Bs) extract addition
1148 Food Bioprocess Technol (2017) 10:1142–1153

Multivariate Analysis: Principal Components Analysis bonds) during storage. This partial discrimination of
and Data Clustering samples through bands at 1654, 1755, and 3006 cm−1
related to cis and carbonyl bonds agrees with I cis
In order to reduce the dimensionality of the data set and values (Table 2). On the other hand, 961 cm−1 band
provide organized data from a high number of obtained corresponding to C=C trans bond presented the highest
values, PCA and HCA were applied (Gaze et al. 2015; load value (0.94) in the PC2 (Table 4), which separated
Matera et al. 2014; Xue et al. 2011). The samples treat- Wc and Bs treated samples at 30 days of refrigerated
ed with citric acid (which is a commonly employed storage from the other ones, due to their high proportion
additive in fresh vegetable systems) were included in of trans bonds indicating the high oxidation level of
the study in order to evaluate if the citric acid added those avocado pulps (Fig. 3a).
to all samples treated with vegetable extracts presented Since PCA using only FTIR spectral data was not
any significant effect. Since the effect was minimal, completely effective for discriminating oxidized and non-
these samples were omitted from PCA analysis for bet- oxidized lipid fractions from avocado pulp, an additional
ter focusing in the differences between untreated sample PCA was performed using extinction coefficients, antioxidant
or the one with ascorbic acid addition and those treated activity and FTIR data (Fig. 3b) taking into account all the
with vegetable extracts. experimental conditions. This second analysis allowed to ex-
A first PCA was performed on the spectral data in plain a total variance of 83.6% with a cophenetic correlation
order to evaluate which bands allowed to explain the of 0.957. The first component associated most of the variance
effect of extracts and ascorbic acid on lipid fraction in the data, indicating the successfully of PCA.
changes of avocado pulp compared with untreated sam- In Fig. 3b chemical indexes (FRAP, ABTS, and K232) are
ple. Differences between lipid fraction at 0 and 30 days the ones that showed loading values >0.50 and also better
of avocado pulp refrigerated storage were analyzed discriminated no oxidized (day 0) and oxidized (day 30) lipid
(Fig. 3a, b). From the results shown in Fig. 3a, it was fractions on PC1 which explained the 57.9% of variance. In
concluded that the PCA method was discriminant addition, both antioxidant indexes (ABTS and FRAP) corre-
enough, since 73.3% variance was explained in the first lated negatively with K232 and were more related to lipid frac-
two principal components. The analysis showed that the tions at day 0, that means no oxidized samples. Selected FTIR
spectral bands with the greatest discriminating power signals corresponding to C=C cis (3006 cm−1) and C=C trans
were those at 3006, 1755, 1654, and 961 cm−1 with a (961 cm−1) showed greater loading value on PC2 according to
cophenetic correlation of 0.926. That means these four results observed in previous PCA (Table 4). The second com-
bands capture the highest percentage of information ponent separated Bs and Wc samples at 30 days from the
from the FTIR spectrum, and hence, the analysis could others, also in agreement with high degree of isomerization
be reduced to these frequencies without considerable described in Icis and Itrans results. Considering lipid fraction
loss of results variance. extracted from avocado pulp stored 30 days at refrigeration,
Additionally, bands at 3006 and 961 cm−1 that corre- Allium treated samples were closely related to K232 while AO
spond to =C-H (cis) stretching and -HC=CH- (trans) appears to be more related to cis double bonds signal at
bending, respectively (Table 1) showed the highest var- 3006 cm−1.
iable loads (Table 4). The 3006 cm−1 band presented the It is remarkable that due to changes in band 3006 and
mayor load in PC1 and 961 cm−1 band in PC2. It is 961 cm−1 appeared perpendicular to antioxidant properties
remarkable that 1654, 1755, and 3006 cm−1 bands pre- from aqueous fraction and K232 means that variables did
sented the great load at PC1 which explains the 42.9% not correlated.
of variance (Fig. 3a). These bands, associated to cis and A second data analysis method, cluster analysis, was
carbonyl bonds (Table 1) discrimated all Allium treat- used to found similarities among samples and to produce
ments (G, S, and O) at day 0 and, C and AO samples a graphic display of how the samples were clustered
at 0 and 30 days from untreated samples, and from Bs (Matera et al. 2014). Cluster analysis is a method for join-
and Wc. This observation shows that Allium treated ing samples in a dendrogram (Btree diagram^), where sam-
samples at day 0 did not undergo oxidation during the ples with the highest correlations are grouped together
addition and homogenization of extracts in comparison while samples with small correlations are widely separated
to Bs and Wc, indicating that the quality of avocado (Jain et al. 2000). Figure 4 depicts the dendrogram obtained
pulp was better preserved during addition and homoge- from cluster analysis including all variables studied in pres-
nization of extract in comparison to other samples. In ent research, which means K232, K270, antioxidant proper-
addition, ascorbic acid (AO) and cauliflower treatments ties, and all FTIR spectral data. Considering in the dendro-
showed a great oxidation retardant effect with low de- gram a break point at the linkage distance around five, five
gree of isomerization (minimal modification in C=C cis classes of samples with similar oxidation profiles were
Food Bioprocess Technol (2017) 10:1142–1153 1149

Fig. 3 Biplot of first two

principal component analysis:
scores plot of variables and scores
plot of cases (samples treated and
untreated at 0 and 30 days of
storage) of lipid fraction from
avocado pulps with ascorbic acid
(AO), garlic (G), onion (O),
scallion (S), white cabbage (Wc),
cauliflower (C), or Brussels
sprouts (Bs) extract addition.
Fig. 3a shows the analysis per-
formed using spectral bands se-
lected for the greatest variance in
the discrimination of data and B
using chemical data and FTIR
spectral data. Colored dots indi-
cate principal component load for
each selected variable. Black
points represent the two principal
components obtained with the
mean responses associated to the
pulps treated and stored for
30 days and white points indicate
samples at day 0. All replicates
were used in calculations, but the
employed software shows aver-
age points for each sample in final
biplot

identified. Lipid fractions from avocado treated with ascor- treated with vegetables extracts and untreated sample gen-
bic acid were slightly different at both storage times, and erated lipid fractions at day 0 that were very similar,
the ones with the lowest degree of oxidation. Avocado pulp followed by those samples stored 30 days with the
1150 Food Bioprocess Technol (2017) 10:1142–1153

Table 4 Value of variables loading in the corresponding PCA analysis hydroperoxides formed by lipid peroxidation have a conjugat-
PC1 PC2 ed dienic system resulting from stabilization of the radical
state by double-bond rearrangement (Laguerre et al. 2007).
Variables analyzed in PCA including 4 selected FTIR intensity bands, These relatively stable compounds absorb in the UV range
time, and sample typea forming a shoulder on the main absorption peak of non-
Signal from FTIR at 961 cm−1 −0.03 0.94 conjugated double bonds (200–210 nm), so they can be mea-
Signal from FTIR at 1654 cm−1 0.77 0.02 sured more appropriately as the K232.
Signal from FTIR at 1755 cm−1 0.68 0.45 K232 presented significant negative Pearson coefficient
Signal from FTIR at 3006 cm−1 0.81 −0.37 with antioxidant indexes and bands associated to C=O bonds
Variables analyzed in PCA including 2 selected FTIR intensity bands, in FTIR spectra (Table 5). Similar coefficients were found for
time, sample type, and antioxidant capacity indexes (ABTS; FRAP,
these FTIR signals and each antioxidant activity measured on
K232)b
aqueous fraction. In the other hand, changes in C=C cis cor-
Signal from FTIR at 961 cm−1 0.04 0.74
related negatively with C-O stretching (1119 cm−1). No sig-
Signal from FTIR at 3006 cm−1 0.24 −0.62
nificant correlation was observed for trans double bonds sig-
ABTS 0.57 −0.08
nal at 961 or 3006 cm−1, probably due to the small changes
FRAP 0.56 0.10
registered for lipid fraction of avocado pulp samples, as
K232 −0.55 −0.21
discussed before.
a
Data corresponding to Fig. 3a
b
Data corresponding to Fig. 3b
Conclusion

exception of untreated sample that was separated alone in- Vegetables extracts added to avocado pulp promoted an in-
dicating the highest degree of oxidation. creased antioxidant activity in the aqueous fraction that clearly
Finally, Pearson correlation was applied to antioxidant prevented oxidation of the lipid fraction. All vegetable ex-
properties, extinction coefficients, and FTIR spectral data tracts added to the avocado pulp demonstrated to markedly
and results are shown in Table 5. Both antioxidant indexes: inhibit the formation of second oxidation products in the lipid
anti-radical scavenging rate (ABTS) and reducing activity fraction. We have previously demonstrated that garlic, scal-
(FRAP) of aqueous fraction from treated and untreated avo- lion, and cauliflower aqueous extracts prevent enzymatic
cado pulps correlated positively. Since free radicals can initi- browning reactions in the aqueous fraction of avocado. In
ate and propagate the lipid peroxidation cascade, the scaveng- the present work, of the mentioned extracts, cauliflower ex-
ing properties, related to the ABTS determination, are good tract showed also effective lipid antioxidant properties, as de-
markers of antioxidant activity (Nuutila et al. 2003). termined by K270 and Icis indexes that could be related to the
Extinction coefficient at 270 nm did not correlate with any observed maintenance of the radical scavenging and reducing
other variable studied, while many significant Pearson corre- power in the aqueous phase. The PCA and natural groupings
lations were found for the coefficient at 232 nm. This result (clustering), based on similarity revealed through analysis
could be explained due to the fact that over 90% of allowed defining the four FTIR spectral bands (3006, 1755,

Fig. 4 Dendrogram obtained

from cluster analysis using
Euclidean distance and
considering all variables
determined in avocado pulps with
citric acid (CA), ascorbic acid
(AO), garlic (G), onion (O),
scallion (S), white cabbage (Wc),
cauliflower (C), or Brussels
sprouts (Bs) extract addition.
Numbers between parentheses in-
dicate storage days
Table 5 Pearson correlation coefficients obtained for all experimental data

ABTS FRAP K232 K270 =C-H cis C=O C=C cis CH3 (1377 cm−1) C-O, CH2 C-O, CH2 C-O (1119 cm−−1) C-O (1028 cm−1) -HC=CH- trans
(3006 cm−1) (1755 cm−1) (1654 cm−1) (1239 cm−1) (1161 cm−1) (961 cm−1)

ABTS 1 0.89** −0.88** ns ns 0.61* ns 0.55* 0.57* 0.59* ns 0.60* ns

Food Bioprocess Technol (2017) 10:1142–1153

FRAP 1 −0.89** ns ns 0.59* ns ns 0.51* 0.60* ns 0.50* ns

K232 1 ns ns ns ns −0.55* −0.58* −0.51* ns −0.61* ns
K270 1 ns ns ns ns ns ns ns ns ns
=C-H cis 1 ns ns ns ns ns ns ns ns
(3006 cm−1)
C=O 1 ns 0.78** 0.80** 0.90** ns 0.74** ns
(1755 cm−1)
C=C cis 1 ns 0.56* ns −0.50* ns ns
(1654 cm−1)
CH3 1 0.68** 0.68** ns 0.73** ns
(1377 cm−1)
C-O, CH2 1 0.84** ns 0.71** ns
(1239 cm−1)
C-O, CH2 1 ns 0.62* ns
(1161 cm−1)
C-O 1 ns ns
(1119 cm−1)
C-O 1 ns
(1028 cm−1)
-HC=CH-trans 1
(961 cm−1)

ns no significative coefficient
*P < 0.05; **P < 0.001
1151
1152 Food Bioprocess Technol (2017) 10:1142–1153

1654, and 961 cm−1) which can discriminate between oxi- Universidad Nacional de Córdoba, Argentina. Retrieved from
http://www.infostat.com.ar.
dized and non-oxidized samples, and also on the type of ex-
Ding, H., Chin, Y. W., Kinghorn, A. D., & Ambrosio, S. M. D. (2007).
tract, which has important practical applications. Based on Chemopreventive characteristics of avocado fruit. Seminars in
previous research that investigated the potential of Allium Cancer Biology, 17, 386–394. doi:10.1016/j.semcancer.2007.04.003.
and Brassica by-products as natural antibrowning agents in Dréau, Y. L., Dupuy, N., Gaydou, V., Joachim, J., & Kister, J. (2009).
Study of jojoba oil aging by FTIR. Analysis, 642, 163–170. doi:10.
different matrices (Roldán et al. 2008; Zocca et al. 2010), the
1016/j.aca.2008.12.001.
employed extracts can be effectively recovered from vegeta- Elez-Martínez, P., Soliva-Fortuny, R. C., Gorinstein, S., & Martín-
ble residues from the industry, thus increasing the added value Belloso, O. (2005). Natural antioxidants preserve the lipid oxidative
of sub-products. stability of minimally processed avocado purée. Sensory and
Nutritive Quality of Food, 70(5), S325–S329.
Although sensory analysis using consumers and sensory
Elez-Martínez, P., Soliva-Fortuny, R. C., & Martín-Belloso, O. (2007).
profiling with trained panel will be necessary for the effective Oxidative rancidity in avocado purée as affected by α-tocopherol,
application of the antioxidant extracts (Horita et al. 2016), pres- sorbic acid and storage atmosphere. European Food Research and
ent results are a helpful start point for the development of min- Technology, 226, 295–300. doi:10.1007/s00217-006-0539-y.
European Commission. (2001). Council Regulation No 1513/2001.
imal processed vegetables, with the aim to preserve their qual-
Amending Regulations No 136/66/EEC and (EC) No 1638/98 as
ity and to increase potentially health-promoting attributes. regards the extension of the period of validity of the aid scheme and
the quality strategy for olive oil. Official Journal of the European
Communities, L, 201, 4–7.
Acknowledgements The authors acknowledge financial support from Gaze, L. V., Costa, M. P., Monteiro, M. L. G., Lavorato, J. A. A., Conte
UBACYT 20020100100397, ANPCYT (PICT 2008 0928), and Júnior, C. A., Raices, R. S. L., … Freitas, M. Q. (2015). Dulce de
CONICET PIP 100846 and 100486. Leche, a typical product of Latin America: Characterisation by phys-
icochemical, optical and instrumental methods. Food Chemistry,
169, 471–477. doi:10.1016/j.foodchem.2014.08.017.
Guillén, M., & Nerea, C. (1997). Infrared spectroscopy in the study of
References edible oils and fats. Journal of the Science of Food and Agriculture,
75(1), 1–11.
Abu-Ghannam, N., & Jaiswal, A. K. (2015). Blanching as a treatment Guillén, M., & Nerea, C. (2000). Some of the most significant changes in
process: effect on polyphenol and antioxidant capacity of cabbage. the Fourier transform infrared spectra of edible oils under oxidative
In V. Preedy (Ed.), Processing and impact on active components in conditions. Journal of the Science of Food and Agriculture, 80(14),
food (pp. 35–43). London: Academic Press. doi:10.1016/B978-0- 2028–2036.
12-404699-3.00005-6. Horita, C. N., Farías-Campomanes, A. M., Barbosa, T. S., Esmerino, E.
Aguirezábal, M. M., Mateo, J., Domínguez, M. C., & Zumalacárregui, J. A., Gomes, A., Bolini, H. M. A., … Pollonio, M. A. R. (2016). The
M. (2000). The effect of paprika, garlic and salt on rancidity in dry antimicrobial, antioxidant and sensory properties of garlic and its
sausages. Meat Science, 54, 77–81. doi:10.1016/S0309-1740(99) derivatives in Brazilian low-sodium frankfurters along shelf-life.
00074-1. Food Research International, 84, 1–8. doi:10.1016/j.foodres.2016.
Balda, P., Tonello, C., Peregrina, R., & de Celis, C. (2011). Industrial high 02.006.
pressure processing of avocado products: emerging trends and im- Jain, A. K., Murty, M. N., & Flynn, P. J. (2000). Data clustering: a review.
plementation in new markets. In Proceedings VII World Avocado ACM Computing Surveys, 31(3), 264–323.
Congress. Jaiswal, A., Rajauria, G., Abu-Ghannam, N., & Gupta, S. (2011).
Berasategi, I., Barriuso, B., Ansorena, D., & Astiasarán, I. (2012). Phenolic composition, antioxidant capacity and antibacterial activity
Stability of avocado oil during heating: comparative study to olive of selected Irish Brassica vegetables. Natural Products
oil. Food Chemistry, 132(1), 439–446. doi:10.1016/j.foodchem. Communications, 6, 1–6.
2011.11.018. Laguerre, M., Lecomte, J., & Villeneuve, P. (2007). Evaluation of the
ability of antioxidants to counteract lipid oxidation: existing
Bock, H. H. (1996). Probabilistic models in cluster analysis.
methods, new trends and challenges. Progress in Lipid Research,
Computational Statistics and Data Analysis, 23, 5–28.
46, 244–282. doi:10.1016/j.plipres.2007.05.002.
Bustos, M. C., Agudelo-Laverde, L. M., Mazzobre, M. F., & Buera, M. P. Matera, J. A., Cruz, A. G., Raices, R. S. L., Silva, M. C., Nogueira, L. C.,
(2014). The relationship between antibrowning, anti-radical and re- Quitério, S. L., et al. (2014). Discrimination of Brazilian artisanal
ducing capacity of Brassica and Allium extracts. The Journal of and inspected pork sausages: application of unsupervised, linear and
Food Studies, 3, 82–92. non-linear supervised chemometric methods. Food Research
Bustos, M. C., Mazzobre, M. F., & Buera, M. P. (2015). Stabilization of International, 64, 380–386. doi:10.1016/j.foodres.2014.07.003.
refrigerated avocado pulp: effect of Allium and Brassica extracts on Moreno, A., Dorantes, A. L., Galíndez, J., & Guzmán, R. I.
enzymatic browning. LWT - Food Science and Technology, 61(1), (2003). Effect of different extraction methods on fatty acids,
89–97. doi:10.1016/j.lwt.2014.11.026. volatile compounds, and physical and chemical properties of
CAA (2012). Alimentos grasos, aceites alimenticios. Chapter 12, Art. av o c a d o (P er s e a a m e r i ca n a M i ll . ) o i l . J o u r n a l o f
535. Resolución Conjunta SPReI N° 64/2012 y SAGyP N° 165/ Agricultural and Food Chemistry, 51, 2216–2221.
2012. In CAA (Código Alimentario Argentino). Nuutila, A. M., Puupponen-Pimiä, R., & Aarni, M. (2003). Comparison
Di Rienzo, J., Guzmán, A., & Casanoves, F. (2002). A multiple- of antioxidant activities of onion and garlic extracts by inhibition of
comparisons method based on the distribution of the root node dis- lipid peroxidation and radical scavenging activity. FoodChemistry,
tance of a binary tree. Journal of Agricultural and Food Chemistry, 81, 485–493.
7(2), 129–142. Paz Antolín, I., & Molero Meneses, M. (2000). Aplicacion de la
Di Rienzo, J., Casanoves, F., Balzarini, M. G., Gonzalez, L., Tablada, M., espectrofotometria UV visible al estudio de la estabilidad termica
& Robledo, C. W. (2012). Infostat. Grupo InfoStat, FCA, de aceites vegetables comestibles. Grasas y Aceites, 51(6), 424–428.
Food Bioprocess Technol (2017) 10:1142–1153 1153

Pereira, E. P. R., Cavalcanti, R. N., Esmerino, E. A., Silva, R., Guerreiro, oils from five Cameroon cultivars extracted with a frendly
L. R. M., & Cunha, R. L. (2016a). Effect of incorporation of anti- enviromental process. Enviromental Engineering and Management
oxidants on the chemical, rheological, and sensory properties of Journal, 12, 97–103.
probiotic petit suisse cheese. Journal of Dairy Science, 99, 1–11. Schebor, C., Mazzobre, M. F., & Buera, M. P. (2010). Glass
doi:10.3168/jds.2015-9701. transition and time-dependent crystallization behavior of de-
Pereira, E. P. R., Faria, J. A. F., Cavalcanti, R. N., Garcia, R. K. A., Silva, hydration bioprotectant sugars. Carbohydrate Research,
R., Esmerino, E. A., … Cruz, A. G. (2016b). Oxidative stress in 345(2), 303–308.
probiotic Petit Suisse: is the jabuticaba skin extract a potential op- Sikora, E., Cieślik, E., Leszczynska, T., Filipiak-Florkiewicz, A., &
tion? Food Research International, 81, 149–156. doi:10.1016/j. Pisulewski, P. M. (2008). The antioxidant activity of selected cru-
foodres.2015.12.034. ciferous vegetables subjected to aquathermal processing. Food
Plaza, L., Sánchez-Moreno, C., Pascual-Teresa, S., de Ancos, B., Cano, Chemistry, 107, 55–59. doi:10.1016/j.foodchem.2007.07.023.
M. P., Pacual-Teresa, S., … Cano, M. P. (2009). Fatty acids, sterols, Sinelli, N., Cosioa, M. S., Gigliotti, C., & Casiraghi, E. (2007).
and antioxidant activity in minimally processed avocados during Preliminary study on application of mid infrared spectroscopy
refrigerated storage. Journal of Agricultural and Food Chemistry, for the evaluation of the virgin olive oil Bfreshness^.
57, 3204–3209. AnalyticaChimica Acta, 598, 128–134. doi:10.1016/j.aca.2007.
Podsedek, A. (2007). Natural antioxidants and antioxidant capacity of 07.024.
Brassica vegetables: a review. LWT - Food Science and Soliva-Fortuny, R. C., Elez-Martínez, P., & Martín-Belloso, O. (2001).
Technology, 40, 1–11. doi:10.1016/j.lwt.2005.07.023. Evaluation of browning effect on avocado purée preserved by com-
Prakash, D., Singh, B. N., & Upadhyay, G. (2007). Antioxidant and free bined methods. Innovative Food Science and Emerging
radical scavenging activities of phenols from onion (Allium cepa). Technologies, 1, 261–268.
Food Chemistry, 102, 1389–1393. doi:10.1016/j.foodchem.2006.
Tang, X., & Cronin, D. A. (2007). The effects of brined onion extracts on
06.063.
lipid oxidation and sensory quality in refrigerated cooked turkey
Pulido, R., Bravo, L., & Saura-Calixto, F. (2000). Antioxidant activity of
breast rolls during storage. Food Chemistry, 100(2), 712–718.
dietary polyphenols as determined by a modified ferric reducing/
Werman, M. J., & Neeman, I. (1987). Avocado oil production and chem-
antioxidant power assay. Journal of Agricultural and Food
ical characteristics. Journal of Americal Oil Chemist Society, 64,
Chemistry, 48, 3396–3402.
229–232.
Quiñones-Islas, N., Meza-Márquez, O. G., Osorio-Revilla, G., &
Gallardo-Velazquez, T. (2013). Detection of adulterants in avocado Xue, J., Lee, C., Wakeham, S. G., & Armstrong, R. A. (2011).
oil by Mid-FTIR spectroscopy and multivariate analysis. Food Organic geochemistry using principal components analysis
Research International, 51(1), 148–154. (PCA) with cluster analysis to study the organic geochemis-
Raitio, R., Orlien, V., & Skibsted, L. H. (2011). Storage stability of cau- try of sinking particles in the ocean. Organic Geochemistry,
liflower soup powder: the effect of lipid oxidation and protein deg- 42(4), 356–367. doi:10.1016/j.orggeochem.2011.01.012.
radation reactions. Food Chemistry, 128(2), 371–379. Yahia, E. M., & González-Aguilar, G. (1998). Use of passive and semi-
Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M., & Rice- active atmospheres to prolong the postharvest life of avocado fruit.
Evans, C. A. (1999). Antioxidant activity applying an improved LWT—Food Science and Technology, 606, 602–606.
ABTS radical cation decolorization assay. Free Radical Biology Yang, H. S., Lee, E. J., Moon, S. H., Paik, H. D., Nam, K., & Ahn, D. U.
and Medicine, 26(9/10), 1231–1237. (2011a). Effect of garlic, onion, and their combination on the quality
Robards, K., Prenzler, P. D., Tucker, G., Swatsitang, P., & Glover, W. and sensory characteristics of irradiated raw ground beef. Meat
(1999). Phenolic compounds and their role in oxidative processes in Science, 89(2), 202–208.
fruits. Food Chemistry, 66, 401–436. Yang, H. S., Lee, E. J., Moon, S. H., Paik, H. D., & Ahn, D. U. (2011b).
Roldán, E., Sánchez-Moreno, C., de Ancos, B., & Cano, M. P. (2008). Addition of garlic or onion before irradiation on lipid oxidation,
Characterisation of onion (Allium cepa L.) by-products as food in- volatiles and sensory characteristics of cooked ground beef. Meat
gredients with antioxidant and antibrowning properties. Food Science, 88(2), 286–291.
Chemistry, 108(3), 907–916. Zocca, F., Lomolino, G., & Lante, A. (2010). Antibrowning potential of
Saha, F. B. U., Kansci, G., Lazar, I. M., Lazar, G., Fokou, E., & Etoa, F. Brassicacaea processing water. Bioresource Technology, 101(10),
X. (2013). ATR-FTIR characterization and classification of avocado 3791–3795.