Professional Documents
Culture Documents
Impact of Temperature and Water Activity On The Aroma Composition and Flavor Stability of Pea Protein Isolates During Storage
Impact of Temperature and Water Activity On The Aroma Composition and Flavor Stability of Pea Protein Isolates During Storage
Function
View Article Online
PAPER View Journal
Published on 09 September 2020. Downloaded by Cornell University Library on 9/11/2020 7:51:59 AM.
Flavor stability of pea protein isolates (PPIs) during storage was investigated. Two commercial PPIs were
stored at three water activities (0.128–0.501) under refrigerated (7 °C) and accelerated (37 °C) tempera-
tures for 12 weeks. Eleven aroma compounds were monitored by gas chromatography-tandem mass
spectrometry (GC/MS/MS) and results revealed significant changes in the aroma concentrations among
the PPI samples during storage. In agreement with the chemical changes, significant differences in ortho-
nasal aroma profiles were demonstrated using a sensory difference-from-control test. The sample stored
under accelerated storage temperature (37 °C) and at the highest water activity showed the greatest
Received 29th April 2020, degree of aroma change. An aroma recombination sensory study indicated the generation of two specific
Accepted 3rd September 2020
compounds, 1-octen-3-ol and nonanal, along with the degradation of 2–4-decadienal resulted in
DOI: 10.1039/d0fo01100b sensory changes during storage indicating lipid oxidation was the main mechanism of flavor instability in
rsc.li/food-function the PPI samples.
ingredients by solvent or supercritical extraction has been 2.3 Water activity (aw) measurement
investigated and shown to reduce lipid oxidation volatile com- The water activities of the samples and salt solutions were
pounds, however, this can negatively affect certain functional measured with a Novasina Water Activity Meter (Novasina AG,
properties or may not be industrially feasible7,13,16–18 Switzerland) calibrated from 0.11–0.75 with standard solu-
Additional processing methods have been investigated includ- tions. Measurements were determined in triplicate.
ing alkaline extraction, solid-dispersion spray-drying, and
enzymatic treatment, which both improve functional pro-
2.4 Storage study
perties and reduce off-flavor perception or volatile
Published on 09 September 2020. Downloaded by Cornell University Library on 9/11/2020 7:51:59 AM.
composition19–21 Others have suggested utilizing pea dextrin PPI samples were stored at three different water activities (aw)
to mask certain specific lipid oxidation compounds.22 of 0.1, 0.3, and 0.5 to simulate different storage environments.
Flavor stability of protein powders during storage is also The initial aw of the PPIs as is was about 0.3 (Table 1), which
known to impact product performance in food applications.23 corresponds to the middle of the range for low moisture
PPIs are low moisture (water activity aw < 0.6), shelf-stable foods.24 For the lower aw sample, PPIs were lyophilized with a
ingredients24 and are typically stored in warehouses for Virtis Sentry 2.0 vacuum freezer at −60 °C and 100 mTorr for
extended periods of time under uncontrolled temperature and 15 min to achieve a low aw between 0.10–0.15. High aw
humidity conditions. A better understanding of the role of samples were placed under saturated salt solution of sodium
storage conditions on the off-flavor generation of pea protein bromide to achieve targeted aw of ∼0.5 (Table 1). In brevity, a
isolates and the specific compounds that impact flavor stabi- saturated salt solution of sodium bromide was prepared and
lity would support the industry’s ability to develop high quality layered into the bottom of glass desiccators. Three layers of
pea protein ingredients. The overall goal of this work was to aluminum pans were stacked within the desiccators on top of
investigate the molecular changes in the aroma profile of PPI a perforated platform. Flat wooden sticks were used to stack
stored under different water activity levels and at different the aluminum pans and allow for airflow between each of the
temperatures on flavor stability. Changes in the aroma layers. Fifty grams of each PPI were spread into each of the
composition were monitored by gas chromatography-tandem aluminum pans. The desiccators were sealed with vacuum
mass spectrometry (GC/MS/MS) analysis and changes in the grease and powders were stirred frequently (every 2–3 days)
orthonasal aroma perception were determined by sensory until equilibrium. The desiccators were stored at ambient
analysis. temperature (23 ± 2 desiccators) for 35 days to achieve equili-
bration at aw of 0.5.
After equilibration, samples were then placed in amber
glass bottles fitted with a PFTE lid and stored for 12 weeks
2. Materials and methods under refrigerated (7 ± 2 °C) and accelerated (37 ± 2 °C) con-
ditions. Accelerated conditions were chosen to mimic storage
2.1 Chemicals at ambient temperature (23 °C) for approximately 7 months.
Chemical grade sodium bromide anhydrous was purchased The estimated accelerated time is based on a temperature
from Sigma Aldrich (St. Louis, MO). Pure (>95%) commercial coefficient, Q10 equation.25 A Q10 of 2 is a typical value to esti-
standards of 4-heptanone, 1-pentanol, hexanal, methional, mate reaction rates in food, particularly dehydrated foods
2-pentyl furan, 2,5-dimethylpyrazine, 2-heptanone, heptanal, where the lowest expected Q10 is 1.5–2 for lipid oxidation.26,27
octanoic acid, isovaleric acid, (E,E)-2,4-nonadienal, (E,E)-2,4- After storage, samples were immediately placed in amber glass
decadienal, 4-hydroxybenzaldehyde, γ-nonalactone, 1-octen-3- bottles with PFTE lid and stored at −40 °C in a freezer until
ol, (E)-2-octenal, 2-isobutyl-3-methoxypyrazine, p-vinylguaiacol, further analysis.
4-methyl-5-thiazoleethanol, and maltol were purchased from
Sigma Aldrich. Nonanal was purchased from Alfa Aesar 2.5 Quantification of volatile aroma compounds using
(Haverhill, MA). Ethanol was purchased from Fisher Scientific dynamic headspace-gas chromatography/mass spectrometry
(Pittsburg, PA). The deuterated standard of 2-acetyl-1-pyrroline (DHS-GC/MS)
([2H3]-2AP) was synthesized in-house.
Twenty-one volatile aroma compounds previously identified in
UHT PPI beverages by Trikusuma et al.15 were monitored in
2.2 Pea protein isolates (PPI) the PPI samples over storage. Analyses were performed using
an Agilent 7890B GC (Agilent Technologies, Santa Clara, CA)
Two commercial PPIs were acquired within 12 weeks from
their manufacturing date. The PPIs contained between
80–85% w/w protein and the remaining content included <8% Table 1 Average water activity (aw) values (±standard deviation) for two
water, <8% fat, <5% fiber, and <6% ash by the supplier certifi- commercial pea protein isolates (PPIs) samples
cates of analysis. Prior to experiments, samples were stored at
−40 °C in glass amber bottles with a PFTE lined lid. For both Low aw Medium aw High aw
analytical and sensory analysis, samples were prepared at a 3% PPI A 0.128 ± 0.006 0.347 ± 0.003 0.501 ± 0.020
w/w suspension in water. PPI B 0.120 ± 0.004 0.337 ± 0.001 0.493 ± 0.004
equipped with a DB-Wax column (60 m × 0.25 mm ID × test, anchored with no difference (0), very little difference
0.25 μm film thickness; Agilent Technologies) and coupled to (1–2), little difference (3–4), medium difference (5–6), strong
an Agilent 7010B triple quadrupole mass spectrometer (MS) difference (7–8), and extreme difference (9–10). Panelist then
(Agilent Technologies). The GC/MS system was also equipped used Compusense Cloud v7.2 software (Compusense Inc.,
with a multipurpose sampler (MPS; Gerstel, Mulheim an der Guelph, Ontario, Canada) to collect the data.
Ruhr, Germany) dynamic headspace (DHS) unit (Gerstel), pro- 2.6.1. PPI samples. Aliquots (10 mL) of a 3% PPI (unaged
grammable temperature vaporization (PTV) inlet (CIS4, control, aged) suspension were placed in 60 mL amber bottles
Gerstel), and a thermal desorption unit (TDU; Gerstel). PPIs fitted with PFTE lids for orthonasal aroma analysis and placed
Published on 09 September 2020. Downloaded by Cornell University Library on 9/11/2020 7:51:59 AM.
were placed in suspension (3% w/w) in water and spiked with at room temperature for 30 min for equilibration.
an ethanolic internal standard solution of 4-heptanone 2.6.2. Sensory recombination samples. To evaluate the con-
(sample concentration of 50 µg L−1). Samples (2 mL) were tribution of aroma compounds to the overall sensory changes
placed into 20 mL glass vials for DHS volatile extraction. occurring during storage, recombination models were created
Samples were incubated and agitated for 10 min at 35 °C. by adding the compounds in water at their concentrations in
Following incubation, vials were purged with 200 mL of high control and aged samples in 10 mL aliquots in 60 mL amber
purity nitrogen gas at a flow rate of 50 mL min−1 to collect bottles with PTFE lids (Table 2). Based on the sensory DFC
volatile compounds on a Tenax adsorbent trap. The trap was results, the sample with the greatest sensory changes (PPI A at
subsequently dry purged for 4 min at a rate of 100 mL min−1 high aw and 37 °C) was selected as target for recombination.
and transferred to the TDU at 40 °C. The volatiles were ther- One model was prepared by adding all 11 compounds at their
mally desorbed from the trap at 250 °C for 3 min at a flow rate determined concentration prior to storage to mimic the
of 50 mL min−1. After desorption, the volatiles were cryofo- control sample (PPI A). Two models were prepared by changing
cused on the CIS maintained at −50 °C then ramped to 240 °C the concentration of the compounds that exhibited statistical
for 3 min and introduced into the column in splitless mode differences either from increasing or decreasing in concen-
for 0.5 min. The GC oven was programmed at 40 °C initially, tration during storage. Additionally, individual compound con-
ramped to 80 °C at 20 °C min−1, and then at 5 °C min−1 until tribution was evaluated by having panelists rate the size of the
240 °C, and was held for 1 minute at 240 °C. Helium carrier change induced by the change in concentration of each indi-
gas was set to a constant flow rate of 1.2 mL min−1. MS data vidual compound according to their levels in the aged
were collected in multiple reaction monitoring (MRM) mode samples.
adapted from Trikusuma et al.,15 obtained at 70 eV in electron 2.6.3 Data analysis. To evaluate change in volatile aroma
ionization (EI) mode with an acquisition rate of 50 scans per s. compounds, the effect of storage conditions on concentrations
Compounds were positively identified using mass spectra, was evaluated by one-way Analysis of Variance (ANOVA).
linear retention index (LRI) and injection of pure compound Whenever a significant effect of sample was observed, a Tukey
standards. LRI values were calculated with injection of post-hoc test was performed (α < 0.05) to determine differences
n-alkane from C6–C26 in the same condition as the samples. A between samples. For the sensory data, the effect of storage
standard solution of the compounds at a concentration corres- conditions on overall degree of difference was evaluated by
ponding to their odor threshold in water was analyzed as a two-way ANOVA (sample, panelist, and interaction). Whenever
reference. Compounds detected above odor threshold in at a significant effect of sample was observed, a Fisher LSD post-
least one storage condition were quantified using five-point
standard addition curves. Data were collected in triplicate.
Curves indicated good linearity (R2 > 0.95) and good Table 2 Aroma recombination model composition of unaged PPI
reproducibility. control and PPI aged for compounds significantly increasing or decreas-
ing during storage
2.6 Sensory analysis degree-of-difference
All sensory evaluations were conducted in compliance with the Unaged Aged PPI model of Aged PPI model of
PPI increasing decreasing
Institutional Review Board (#2019E0534, #2019E0743). A differ- model compounds compounds
ence-from-control (DFC) test was conducted according to the Compound (μg L−1) (μg L−1) (μg L−1)
methods outlined by Meilgaard et al. to determine if there was Hexanal 1188 1188 1188
a perceptible difference in orthonasal aroma between 2-Heptanone 35.6 144.7a 35.6
samples.28 Sixteen experienced sensory panelists (9 female, Heptanal 4.57 11.1a 4.57
2-Pentylfuran 71.5 159.1a 71.5
7 male, ages 24–45) were recruited participate. All sixteen Nonanal 15.6 26.5a 15.6
panelists participated in the initial evaluation of the control (E)-2-Octenal 4.47 4.47 4.47
and aged samples, 10 of which later evaluated the recombina- 1-Octen-3-ol 1.51 12.51a 1.51
2-Isobutyl-3- 0.026 0.026 0.026
tion samples. In each session, panelists were prompted to methoxypyrazine
smell an unaged, control sample and then were presented (E,E)-2,4-Nonadienal 1.15 1.15 0.53a
with four coded samples in a counterbalanced order, includ- (E,E)-2,4-Decadienal 13.3 13.3 3.0a
p-Vinylguaiacol 7.74 7.74 7.74
ing one blind control. Panelists used a visualized 10-point
a
paper scale to rate samples as they progressed through the Indicate compound concentration changed for aged models.
hoc multiple comparison test was performed. The panelist*re- aw. 2-Pentylfuran was present above odor threshold (approxi-
plicate interaction was not significant for any of the samples mately 10 to 30 times), thus suggesting contribution to the
indicating panelist performance and panelist’s replicates were fruity, green, and earthy notes of the protein aroma.
therefore considered independent. All statistical analyses were Nonanal is reported to derive from oleic acid and can be
performed with SPSS® 25 (IBM Corporation, Armonk, NY). impacted by differences between fatty acid profiles of the
protein brands.32,33 The longer chain aldehyde, nonanal,
showed an 8-fold increase in concentration in PPI A at the low
3. Results and discussion aw condition when stored at 7 °C, whereas no significant
Published on 09 September 2020. Downloaded by Cornell University Library on 9/11/2020 7:51:59 AM.
Table 3 Concentrations of aroma compounds in 3% pea protein isolate A aqueous samples (PPI A) over 12-week storagea
Storage at 7 °C
LRI (DB-wax) Compound Control Low aw (0.128) Medium aw (0.347) High aw (0.501) Odor thresholdb (μg L−1) Storage effect p-valuec
1095 Hexanal 1188 ± 215ab 990± 56ab 1419 ± 76b 774 ± 55a 4.5 0.028
1195 2-Heptanone 35.6 ± 0.8a 42.8 ± 0.7b 46.4 ± 1.4c 47.3 ± 2.4c 140 <0.001
1197 Heptanal 4.57 ± 0.38a 6.99 ± 0.46c 5.69 ± 0.12b 5.85 ± 0.15b 3 0.001
Storage at 37 °C
LRI (DB-wax) Compound Control Low aw (0.128) Medium aw (0.347) High aw (0.501) Odor thresholdb (μg L−1) Storage effect p-valuec
1095 Hexanal 1188 ± 215c 942 ± 68ab 803 ± 77a 1142 ± 210c 4.5 0.009
1195 2-Heptanone 35.6 ± 0.8a 135.6 ± 4.8b 154.1 ± 14.2b 144.7 ± 6.2b 140 <0.001
1197 Heptanal 4.57 ± 0.38a 13.09 ± 0.64b 12.89 ± 1.33b 11.14 ± 0.75b 3 0.002
1238 2-Pentyl-furan 71.5 ± 8.1a 177.5 ± 6.7b 165.9 ± 18.9b 159.1 ± 6.9b 6 0.002
1405 Nonanal 15.6 ± 0.4a 36.4 ± 3.1c 29.2 ± 2.3bc 26.52 ± 1.6b 1 0.004
1444 (E)-2-Octenal 4.47 ± 1.53 7.30 ± 1.38 7.28 ± 0.96 7.90 ± 0.48 3 Ns
1451 1-Octen-3-ol 1.51 ± 0.13a 9.30 ± 0.52b 11.4 ± 1.0bc 12.55 ± 0.82c 1 <0.001
1539 2-Isobutyl-3-methoxypyrazine 0.026 ± 0.001a 0.038 ± 0.007b 0.033 ± 0.003ab 0.034 ± 0.004ab 0.002 0.048
1717 (E,E)-2,4-Nonadienal 1.15 ± 0.09b 1.12 ± 0.09b 0.70 ± 0.09a 0.53 ± 0.01a 0.09 <0.001
1814 (E,E)-2,4-Decadienal 13.32 ± 1.59c 8.46 ± 0.95b 4.08 ± 0.64a 2.67 ± 0.30a 0.07 0.006
2138 p-Vinylguaiacol 7.74 ± 0.66 10.08 ± 0.68 7.15 ± 0.60 7.71 ± 1.90 3 Ns
a
Mean ± standard deviation. b Odor threshold value in water; retrieved from Leffingwell &Associates. c In each row, values with different letters are significantly different according to Tukey’s
HSD (α = 0.05), ns: not significant.
View Article Online
Paper
Food Funct.
Published on 09 September 2020. Downloaded by Cornell University Library on 9/11/2020 7:51:59 AM.
Paper
Food Funct.
Table 4 Concentrations of aroma compounds in 3% pea protein isolate B aqueous samples (PPI B) over 12-week storagea
Storage at 7 °C
LRI (DB-wax) Compound Control Low aw (0.120) Medium aw (0.337) High aw (0.439) Odor thresholdb (μg L−1) Storage effect p-valuec
Storage at 37 °C
LRI (DB-wax) Compound Control Low aw (0.128) Medium aw (0.347) High aw (0.501) Odor thresholdb (μg L−1) Storage effect p-valuec
1095 Hexanal 1750 ± 193a 2433 ± 230b 2214 ± 198ab 1587 ± 171a 4.5 0.022
1197 Heptanal 7.60 ± 0.24a 15.53 ± 0.98c 14.03 ± 0.26c 10.53 ± 0.23b 3 <0.001
1238 2-Pentyl-furan 64.9 ± 2.7a 113.1 ± 7.0b 99.1 ± 13.4b 93.2 ± 5.6a 6 0.009
1405 Nonanal 14.7 ± 1.6a 19.1 ± 0.3b 15.1 ± 0.8a 13.1 ± 0.8a 1 0.003
1444 (E)-2-Octenal 8.15 ± 1.57b 4.30 ± 0.38ab 1.14 ± 0.14a 7.23 ± 3.10ab 3 0.024
1451 1-Octen-3-ol 2.09 ± 0.15a 4.68 ± 0.69b 4.84 ± 0.79b 8.20 ± 1.10c 1 <0.001
1539 2-Isobutyl-3-methoxypyrazine 0.038 ± 0.005a 0.049 ± 0.002b 0.042 ± 0.003ab 0.038 ± 0.004ab 0.002 0.029
1717 (E,E)-2,4-Nonadienal 3.65 ± 1.09b 1.16 ± 0.05a 0.04 ± 0.03a 0.05 ± 0.03a 0.09 0.012
1814 (E,E)-2,4-Decadienal 60.69 ± 10.31b 15.55 ± 1.06a 5.86 ± 1.19a 2.60 ± 0.40a 0.07 <0.001
2138 p-Vinylguaiacol 22.61± 1.53c 6.57 ± 2.24b 3.48 ± 0.43ab 2.10 ± 0.48a 3 <0.001
a
Mean ± standard deviation. b Odor threshold value in water; retrieved from Leffingwell &Associates. c In each row, values with different letters are significantly different according to Tukey’s
HSD (α = 0.05), ns: not significant.
View Article Online
PPI A did not significantly change at low aw, but significantly During storage of peas the concentration of total volatile com-
decreases by about 2-fold in the medium and high aw con- pounds is higher at 37 °C than 4 °C.48
ditions. Similar changes were observed for (E,E)-2,4-nonadie- The potent pyrazine compound 2-isobutyl-3-methoxypyra-
nal in PPI B. All three aw levels resulted in the same significant zine (IBMP) with a green pepper aroma has been reported in
differences at 37 °C, with a significant decrease in concen- peas and other raw vegetables49,50 as well as in UHT processed
tration up to 60-fold in the medium aw sample. (E,E)-2,4- PPI.15 In the current study, IBMP concentration did not signifi-
Nonadienal is known to exhibit floral, brothy, and waxy odor cantly change over storage in PPI A nor in PPI B at 7 °C, but
notes, typically found in foods containing high linoleic acid did increase at elevated temperature for the low aw condition
Published on 09 September 2020. Downloaded by Cornell University Library on 9/11/2020 7:51:59 AM.
content and subjected to high temperature processing con- in both samples.51 IBMP has a very low odor threshold
ditions. This compound has been also identified in other (0.002 μg L−1) and was >13-fold above the threshold value in
spray-dried products, and losses could be a result of sub- the PPI samples supporting its contribution to the aroma
sequent reactivity of the double bonds to produce degradation attributes.
compounds.42 (E,E)-2,4-Decadienal showed similar changes in Levels of p-vinylguaiacol, a phenolic degradation compound
both PPIs. More specifically, in PPI A at 7 °C, (E,E)-2,4-decadie- with a smoky, vanilla-like aroma changed slightly throughout
nal concentration almost doubled under storage at low aw and storage of PPI A when stored at low temperature and low aw.
refrigerated conditions. Under accelerated conditions of 37 °C, However, in PPI B, significant decreases in concentration were
this compound significantly decreased at all three aw levels observed over storage at medium and high aw in the 7 °C con-
with greatest decreases of 3 to 5-fold at medium and high aw, dition and in all the conditions at 37 °C. This compound is a
respectively. In PPI B, the behavior of (E,E)-2,4-decadienal was ferulic acid degradation product found in plant materials.52 It
different at 7 °C; this compound significantly decreased at was shown that amino acid L-cysteine can inhibit p-vinylguaia-
medium and high aw. Whereas at 37 °C, changes similar to col formation in stored orange juice by up to 80%.53 Cysteine
PPI A were observed with large decreases suggesting it was is part of the amino acid profile of pea and may be preventing
degrading to other products. Decreases of (E,E)-2,4-decadienal ferulic acid degradation.54 A decrease in p-vinylguaiacol during
could result in the formation of hexanal;38 however, this storage suggests further degradation of this compound. For
increase was not observed as the level of hexanal were 25 to example, during beer storage, p-vinylguaiacol can be oxidized
100 times greater than the concentration of (E,E)-2,4-decadie- to form vanillin or hydrated to form apocynol.55 Alternatively,
nal in initial conditions. Another study using an isotope label- p-vinylguaiacol has been shown to interfere with the Maillard
ing experiment showed that (E,E)-2,4-decadienal can autoxi- reaction and may be decreasing due to formation of sugar-
dize to form (E,E)-2,4-decanoic acid, heptanal, 2-octenal, phenol adducts.56
2-nonenal, glyoxal, and malonaldehyde in soybean oil.43 These results can be compared to several other storage
1-Octen-3-ol is a vinyl alcohol with a potent mushroom studies of pea and other legume products. In one study, pea
aroma arising from linoleic acid oxidation in the presence of flour, concentrate, and starch were stored between 3.8–13.6%
enzymes.44 In the PPI samples, 1-octen-3-ol was present at con- moisture at 7, 21, and 30 °C for one year.57 Flavor changes as
centrations 2 to 10 times greater than threshold. In PPI A, compared to the control were correlated with higher moisture
1-octen-3-ol concentration exhibited significant increases levels and temperatures compared to the control. Sumner et al.
under all the aw conditions and at both 7 and 37 °C. Similarly, associated a decrease in linoleic acid content over storage that
in PPI B at 37 °C, 1-octen-3-ol concentration increased at all suggested oxidation or polymerization reactions.57
aw; however, at 7 °C the concentration of 1-octen-3-ol signifi- Furthermore, in peanuts stored at 0.19 and 0.60 aw for 7 weeks
cantly increased at medium aw. The generation of 1-octen-3-ol increased changes in the lipid oxidation volatiles were
was generally favored at higher aw over lower aw suggesting observed at 0.19 aw than 0.6 aw.58 In another study, fava bean
enzymatic pathways were involved in product formation. flours and concentrates were stored under accelerated storage
Ullrich and Grosch reported 1-octen-3-ol to be one of the conditions (38 °C, 65% relative humidity) and were compared
main products of linoleic acid oxidation that increases over to a year-old sample from the previous crop year.59 These
storage.45 Hoffman44 observed in soybean oil, that linoleic authors’ results showed that free fatty acids increased over
acid can degrade to form both (E)-2-octenal and 1-octen-3-ol, storage, although the characteristic dried pea flavor did not
but when pure linoleic acid was reacted only 1-octen-3-ol was change. Heat treatment was also investigated, which resulted
detected suggesting there may be other constituents that effect in enzyme inactivation and a decrease in dried pea flavor;
formation and stabilization in the matrix, such as enzymatic however, additional objectionable off-flavors also developed.
activity. Furthermore, 1-octen-3-ol can be further oxidized to
produce the more potent mushroom compound, 1-octen-
3-one,43 which was not quantified in this study. 3.2 Sensory evaluation and aroma recombination models
Review of the changes in the lipid oxidation products of the The overall perceived orthonasal change in the PPI samples
PPIs during storage in relation to aw were consistent with a during storage is shown in Fig. 1 and 2. Significant changes in
J-shaped relative reaction curve.46 As expected, higher tempera- aroma over storage ( p-value < 0.05) were observed for both
ture conditions resulted in more drastic changes during PPIs with the extent of changes depending on the PPI brand
storage, which is consistent with lipid oxidation in food.47 and the storage temperature.
Fig. 1 Average difference from control (DFC) ratings (n = 32) for 3% aqueous pea protein isolates PPI A (i and ii) and PPI B (iii and iv) for unaged and
aged samples under modified aw conditions at 7 °C and 37 °C for 12 weeks. Within each temperature condition, different letters indicate significant
difference according to LSD post-hoc (α = 0.05).
of the samples were also collected to give an indication of the (control) and two aged samples that reflected changes in the
nature of the chemical changes, such as intensity, aroma compounds that significantly increased or decreased in con-
characteristics, or a combination of the two. The comments centration during storage (see Table 2). The two recombination
from this study revealed that both aroma intensity and quality aged models of the increasing and decreasing compounds
differences were detected compared to the control. were rated at 3.2 and 3.0, respectively, corresponding to ‘little’
Aroma recombination models of the PPI A sample at high difference from the blind control (Fig. 2). Thus, the reported
aw condition for the unaged aged samples were subsequently aroma compositional changes induced a perceptible difference
analyzed (shown in Fig. 2) to determine the sensory relevance in orthonasal aroma.
Published on 09 September 2020. Downloaded by Cornell University Library on 9/11/2020 7:51:59 AM.
of the noted quantitative changes in the aroma compounds To further define the specific role of the seven individual
during storage (Table 3 and Fig. 1). The recombination models compounds that significantly changed in concentration during
contained a mixture of all compounds for the unaged sample storage on the reported flavor change of the PPI samples
Fig. 3 Average difference from control (DFC) ratings (n = 20) for recombination models of unaged PPI (blind control), (i) aged PPI for individual
compounds increasing in concentration (unaged PPI recombination model + 11, 109.1, 87.6, 6.53, 10.9 µg L−1 of 1-octen-3-ol, 2-heptanone, 2-pen-
tylfuran, heptanal, and nonanal, respectively) and (ii) aged PPI recombination model for individual compounds decreasing in concentration (unaged
PPI recombination model − 0.62, 10.3 µg L−1 of (E,E)-2,4-nonadienal and (E,E)-2,4-decadienal, respectively); different letters indicate significant
difference according to LSD post-hoc (α = 0.05).
(Fig. 1 and 2), the unaged PPI recombination model (Table 2, Conflicts of interest
11 compounds) was compared to seven different aged recombi-
nation models representing the concentration for each individ- There are no conflicts to declare.
ual compound (unaged PPI recombination model adjusted to
the concentration of each individual compound after storage)
and is shown in Fig. 3. Among the compounds that signifi-
cantly increased in concentration during storage (Fig. 3i), Acknowledgements
1-octen-3-ol and, to a lesser extent, nonanal had a significant
Published on 09 September 2020. Downloaded by Cornell University Library on 9/11/2020 7:51:59 AM.
19 Z. Gao, P. Shen, Y. Lan, L. Cui, J. B. Ohm, B. Chen and 39 L. Bott and E. Chambers IV, J. Sens. Stud., 2006, 21, 308–
J. Rao, Food Res. Int., 2020, 131, 109045. 321.
20 L. Fang, H. Xiang, D. Sun-Waterhouse, C. Cui and J. Lin, 40 M. Alaiz, R. Zamora and F. J. Hidalgo, J. Agric. Food Chem.,
J. Agric. Food Chem., 2020, 68, 1691–1697. 1995, 43, 795–800.
21 Y. Lan, M. Xu, J. B. Ohm, B. Chen and J. Rao, Food Chem., 41 M. Alaiz and J. Girón, J. Agric. Food Chem., 1994, 42, 2094–
2019, 278, 665–673. 2098.
22 S. Böttcher, U. Steinhäuser and S. Drusch, Food Chem., 42 K. Warner, W. E. Neff, W. C. Byrdwell and H. W. Gardner,
2015, 169, 492–498. J. Agric. Food Chem., 2001, 49, 899–905.
Published on 09 September 2020. Downloaded by Cornell University Library on 9/11/2020 7:51:59 AM.
23 B. J. Wright, S. E. Zevchak, J. M. Wright and M. A. Drake, 43 S. T. Michalski and E. G. Hammond, J. Am. Oil Chem. Soc.,
J. Food Sci., 2009, 74, S17–S29. 1972, 49, 563–566.
24 Q. Rao, A. K. Kamdar and T. P. Labuza, Crit. Rev. Food Sci. 44 G. Hoffmann, J. Am. Oil Chem. Soc., 1962, 39, 439–
Nutr., 2016, 56, 1169–1192. 444.
25 B. Fu and T. P. Labuza, Food Control, 1993, 4, 125–133. 45 F. Ullrich and W. Grosch, Z. Lebensm.-Unters. Forsch., 1987,
26 B. Fu and T. P. Labuza, in Quality in Frozen Foods, ed. 184, 277–282.
M. C. Erickson, Y. C. Hung, Springer, Boston, MA, 1997, 46 T. P. Labuza, Food Technol., 1980, 34, 36–41.
ch. 19, 377–415. 47 E. N. Frankel, Prog. Lipid Res., 1980, 19, 1–22.
27 T. P. Labuza, Application of Open Dating to Specific Foods. 48 S. Azarnia, J. I. Boye, T. Warkentin and L. Malcolmson,
1979. https://www.princeton.edu/~ota/disk3/1979/7911/ Int. J. Food Sci. Technol., 2011, 46, 2408–2419.
791112.PDF, (accessed April 26 2020). 49 K. E. Murray and F. B. Whitfield, J. Sci. Food Agric., 1975,
28 M. C. Meilgaard, G. V. Civille, T. B. Carr, Sensory Evaluation 26, 973–986.
Techniques, CRC Press, Taylor & Francis Group, 2016. 50 K. E. Murray, J. Shipton, F. B. Whitfield and J. H. Last,
29 M. A. Monsoor and A. Proctor, J. Food Sci., 2004, 69, C632–C636. J. Sci. Food Agric., 1976, 27, 1093–1107.
30 L. Heng, J.-P. Vincken, G. van Koningsveld, A. Legger, 51 K. E. Murray, J. Shipton and F. B. Whitfield, Chem. Ind.,
H. Gruppen, T. van Boekel, J. Roozen and F. Voragen, J. Sci. 1970, 27, 897–898.
Food Agric., 2006, 86, 1225–1231. 52 H. Peleg, I. Naim, U. Zehavi, R. L. Rouseff and S. Nagy,
31 D. J. Sessa and J. J. Rackis, J. Am. Oil Chem. Soc., 1977, 54, J. Agric. Food Chem., 1992, 40, 764–767.
468–473. 53 M. Naim, I. Zuker, U. Zehavi and R. L. Rouseff, Food Chem.,
32 J. C. Braddock, C. A. Sims and S. F. O’Keefe, J. Food Sci., 1993, 41, 1359–1361.
1995, 60, 489–493. 54 P. Leterme, T. Monmart and E. Baudart, J. Sci. Food Agric.,
33 M. M. Torres, M. L. Martínez and D. M. Maestri, J. Am. Oil 1990, 53, 107–110.
Chem. Soc., 2005, 82, 105–110. 55 N. Vanbeneden, D. Saison, F. Delvaux and F. R. Delvaux,
34 C. J. Bergman, J. T. Delgado, R. Bryant, C. Grimm, J. Agric. Food Chem., 2008, 56, 11983–11988.
K. R. Cadwallader and B. D. Webb, Cereal Chem., 2000, 77, 56 D. Jiang and D. G. Peterson, Phytochem. Rev., 2010, 9, 187–
454–458. 193.
35 C. W. Fritsch and J. A. Gale, J. Am. Oil Chem. Soc., 1977, 54, 57 A. K. Sumner, L. L. Whalley, G. Blankenagel and
225–228. C. G. Youngs, Can. Inst. Food Sci. Technol. J., 1979, 12, 51–
36 F. Shahidi and R. B. Pegg, J. Food Lipids, 1994, 1, 177–186. 55.
37 F. Shahidi, J. Yun, L. J. Rubin and D. F. Wood, Can. Inst. 58 K. A. Reed, C. A. Sims, D. W. Gorbet and S. F. O’Keefe, Food
Food Sci. Technol. J., 1987, 20, 104–106. Res. Int., 2002, 35, 769–774.
38 P. Schieberle and W. Grosch, J. Am. Oil Chem. Soc., 1981, 59 C. Hinchcliffe, M. McDaniel, M. Vaisey and N. A. M. Eskin,
58, 602–607. Can. Inst. Food Sci. Technol. J., 1977, 10, 181–184.