Changes in the Key Odorants and Aroma Profiles of Hamlin and Valencia Orange Juices Not from Concentrate (NFC) during Chilled Storage

Article abstract of DOI:10.1021/acs.jafc.8b02257

Application of the aroma extract dilution analysis (AEDA) on the volatiles isolated by extraction/SAFE distillation from NFC (not from concentrate) juice from Hamlin oranges revealed 51 odor-active constituents in the flavor dilution (FD) factor range of 8 to 8192 among which vanillin, wine lactone, and (R)-linalool appeared with the highest FD factors. The AEDA applied on the volatile fraction of the same batch of juice stored at 0 °C for 10 months under aseptic conditions showed clear changes in the aroma profile as well as in the FD factors of key odorants. The reduction in the intensity of the citrus-like, pungent, green odor attributes in the aroma profile correlated with the loss of 1-penten-3-one, acetaldehyde, and (Z)-3-hexenal and a clear decrease in hexanal, octanal, nonanal, decanal, and (E,E)-2,4-decadienal. Quantitation done by stabile isotope dilution assays followed by a calculation of odor activity values (ratio of concentration to odor thresholds in citrate buffer) confirmed that the quick loss of 1-penten-3-one and acetaldehyde already within a few weeks and a significant reduction in nearly all aldehydes over the storage time of 10 months were responsible for the changes in the overall aroma profile of the juice. The same approach applied on Hamlin juice from the next harvest year as well as on chilled stored NFC juice from Valencia oranges confirmed the results for another harvest year and another orange variety.

Full text of DOI:10.1021/acs.jafc.8b02257

Subscriber access provided by Kaohsiung Medical University
Chemistry and Biology of Aroma and Taste
Changes in the Key Odorants and Aroma Profiles of Hamlin and Valencia
Orange Juices not from Concentrate (NFC) During Chilled Storage
Ines Sellami, Veronika Mall, and Peter Schieberle
J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.8b02257 • Publication Date (Web): 11 Jun 2018
Downloaded from http://pubs.acs.org on June 12, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 51
Journal of Agricultural and Food Chemistry
1
Changes in the Key Odorants and Aroma
Profiles of Hamlin and Valencia Orange
Juices not from Concentrate (NFC) During
Chilled Storage
Ines Sellami#, Veronika Mall# and Peter Schieberle§*
# Deutsche Forschungsanstalt für Lebensmittelchemiea,
LiseꢀMeitnerꢀStrasse 34, Dꢀ85354 Freising, Germany
§ Technische Universität München, Department für Chemie,
LiseꢀMeitnerꢀStrasse 34, Dꢀ85354 Freising, Germany
*Corresponding Author
aThe institute was renamed as Leibniz Institute for Systems Biology at the Technical
University Munich
Phone: +49 871 97698160
Eꢀmail: peter.schieberle@ch.tum.de
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 2 of 51
2
1
2
ABSTRACT. Application of the aroma extract dilution analysis (AEDA) on the volaꢀ
tiles isolated by extraction/SAFE distillation from NFC (not from concentrate) juice
from Hamlin oranges revealed 51 odorꢀactive constituents in the flavor dilution (FD)
factor range of 8 to 8192 among which vanillin, wine lactone and (R)ꢀlinalool apꢀ
peared with the highest FD factors. The AEDA applied on the volatile fraction of the
same batch of juice stored at 0 °C for ten months under aseptic conditions showed
clear changes in the aroma profile as well as in the FD factors of key odorants. The
reduction in the intensity of the citrusꢀlike, pungent, green odor attributes in the aroꢀ
ma profile correlated with the loss of 1ꢀpentenꢀ3ꢀone, acetaldehyde and (Z)ꢀ3ꢀ
hexenal and a clear decrease in hexanal, octanal, nonanal, decanal and (E,E)ꢀ2,4ꢀ
decadienal. Quantitation done by stabile isotope dilution assays followed by a calcuꢀ
lation of odor activity values (ratio of concentration to odor thresholds in citrate buffer)
confirmed that the quick loss of 1ꢀpentenꢀ3ꢀone and acetaldehyde already within a
few weeks, and a significant reduction in nearly all aldehydes over the storage time of
10 months was responsible for the changes in the overall aroma profile of the juice.
The same approach applied on Hamlin juice from the next harvest year as well as on
chilled stored NFC juice from Valencia oranges confirmed the results for another harꢀ
vest year and another orange variety.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
KEY WORDS: NFC orange juice, storage, aroma extract dilution analysis, stable isoꢀ
tope dilution assay, odor activity value; 1ꢀpentenꢀ3ꢀone; acetic acid
ACS Paragon Plus Environment

Page 3 of 51
Journal of Agricultural and Food Chemistry
3
22 INTRODUCTION
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Orange trees (Citrus sinensis, L. Osbeck) are one of the oldest cultivated plants
and the fruit and juice production constantly increased due to the characteristic flavor
and beneficial nutrient content. Today, orange juice ranks among the world`s most
produced fruit juices, and the perꢀcapitaꢀconsumption in Germany was about 8 Liter
in 2014.1 Two types of orange juices are prepared commercially; juice from concenꢀ
trate (CJ) which undergoes a thermal treatment to evaporate the major part of the
water as well as the aroma fraction, and NFC (not from concentrate) juice, which is
only pasteurized and then chilled stored before consumption. Although freshly
squeezed orange juice is said to have a superior aroma compared to both, CJ and
NFC juice, in particular the overall aroma of NFC juices comes close to the aroma of
freshly squeezed juices.2
Investigations on the volatile fraction of orange juices were already started about
100 years ago,3 and until now more than 400 volatiles have been identified.4 Next to
(R)ꢀlimonene, (R)ꢀαꢀpinene and myrcene, the esters (S)ꢀethyl 2ꢀmethylbutanoate,
ethyl butanoate and ethyl 2ꢀmethylpropanoate as well as the aldehydes (Z)ꢀ3ꢀ
hexenal, hexanal and acetaldehyde were previously characterized 5,6 among the key
aroma compounds in freshly squeezed orange juice using the Sensomics concept.
But, the aroma of freshly squeezed orange juice is not stable, and a significant
change in the overall aroma profile rapidly occurs during processing and storage.7
For example, enzymatic reactions, thermal reactions during pasteurization and botꢀ
tling8ꢀ12 are wellꢀknown to induce changes in the aroma composition.
In addition, several aroma compounds are newly formed or generated either durꢀ
ing storage in glass bottles in the trade or during chilled storage before bottling. Sevꢀ
eral investigations have already been carried out to understand the formation of offꢀ
flavors.7,9,13,14 Kirchner and Miller9 were among the first to study the changes in the
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 4 of 51
4
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
volatiles between fresh and stored juices and found an increase in, e.g. carvone, aceꢀ
tic acid and αꢀterpineol, while esters, aldehydes and terpene hydrocarbons deꢀ
creased. Tatum et al.13 later on suggested 2ꢀmethoxyꢀ4ꢀvinylphenol and 4ꢀhydroxyꢀ
2,5ꢀdimethylꢀ3(2H)ꢀfuranone to be responsible for the aged offꢀflavor formed after
storage at 35 °C. Additionally, it is long known that the concentration of αꢀterpineol
increases significantly depending on the storage temperature.9,13ꢀ16
However, most studies did not use approaches directly combining sensory evaluaꢀ
tion and analytical measurements, such as gas chromatographyꢀolfactometry (GCꢀO)
or odor activity value calculations. In a first study trying to confirm the above menꢀ
tioned compounds as marker substances for juice storage using the Sensomics conꢀ
cept16 the authors did not confirm the influence of these compounds on the formation
of offꢀflavors during storage of juice from concentrate in glass bottles at 20 °C. Inꢀ
stead, a drastic decrease in acetaldehyde, octanal and decanal was found and it was
suggested that the increase in concentration of certain juice compounds together with
the decrease of key juice odorants is responsible for the overall changes in the aroꢀ
ma of the orange juice during storage.16
But, to date, no comprehensive approach using the Sensomics concept including
systematic sensory experiments has been used to clarify the molecular reason for
changes in the overall aroma of chilled stored NFC orange juice. Therefore, it was
the aim of this study to clarify the molecular background of aroma changes occurring
during chilled storage of NFC orange juices from the two varieties, Hamlin and Vaꢀ
lencia, by means of the Sensomics concept. The particular focus was on compounds
degrading during storage in order to establish a basis for further studies on the pathꢀ
ways leading to their instability even under the mild conditions of a chilled storage.
ACS Paragon Plus Environment

Page 5 of 51
Journal of Agricultural and Food Chemistry
5
73 MATERIALS AND METHODS
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
Materials. Orange juices were produced and stored by a commercial juice supplier
(Tropicana Products Inc). Fruits from the same batch of either Hamlin (Ham) or Vaꢀ
lencia (Val) oranges were extracted and stored in aseptic tanks at 0 °C. Samples
were taken every month for a total period of 10 months, filled in aseptic bottles, sent
frozen to Germany and stored at – 60 °C prior to analysis.
Reference Odorants. Reference compounds were obtained from the commercial
sources given in parentheses: acetaldehyde, (E,E)ꢀ2,4ꢀdecadienal, decanal, ethyl 2ꢀ
methylbutanoate, ethyl 3ꢀhydroxyhexanoate, ethyl hexanoate, ethyl propanoate, hexꢀ
anal, (Z)ꢀ3ꢀhexenal, 4ꢀhydroxyꢀ2,5ꢀdimethylꢀ3(2H)ꢀfuranone, 2ꢀhydroxyꢀ3ꢀbutanone,
3ꢀisopropylꢀ2ꢀmethoxyꢀpyrazine, (R)ꢀlimonene, 3ꢀ(methylthio)propanal, methyl 3ꢀ
hydroxyhexanoate, 2ꢀ and 3ꢀmethylbutanoic acid, myrcene, (E,Z)ꢀ2,6ꢀnonadienal,
octanal, 1ꢀpentenꢀ3ꢀone, phenylacetic acid, αꢀpinene, αꢀterpineol, vanillin and 2ꢀethylꢀ
3,5ꢀmethylpyrazine, 1,8ꢀcineol, ethyl butanoate and linalool (SigmaꢀAldrich Chemie,
Taufkirchen, Germany). Carvone, nootkatone, 1ꢀoctenꢀ3ꢀone and 2ꢀmethoxyꢀ4ꢀ
vinylphenol (Lancaster, Mühlheim, Germany). Acetic acid and 2,3ꢀbutanedione
(Merck, Darmstadt, Germany). Nonanal and βꢀionone (Roth, Karlsruhe, Germany).
The following reference odorants were synthesized as reported in the literature:
transꢀ4,5ꢀepoxyꢀ(E)ꢀ2ꢀdecenal17
and
3α,4,5,7αꢀtetrahydroꢀ3,6ꢀdimethylꢀ2(3H)ꢀ
benzofuranone (wine lactone).18
Chemicals Ascorbic acid, citric acid (water free), ethanol, fructose, glucose, lithiꢀ
um aluminium hydride, transꢀpꢀmenthꢀ2ꢀene, potassium peroxomonosulfate, potassiꢀ
um acetate, potassium ferrocyanide, sucrose, sodium thiosulfate, pꢀtoluene sulfonic
acid and zinc sulfate were purchased from SigmaꢀAldrich Chemie (Taufkirchen, Gerꢀ
many). Aceton (suprasolv), acetic anhydride, methyl octanoate, silica gel, sodium
chloride, sodium hydrogen carbonate, anhydrous sodium sulfate, and sulfuric acid
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 6 of 51
6
99
were from Merck (Darmstadt, Germany). Diethyl ether and dichloromethane (Merck)
100 were freshly distilled before use.
101 Isotopically Labeled Internal Standards. These were synthesized as previously
102 reported: [2H4]ꢀcarvone and [2H4]ꢀ decanal,15 [2H4]ꢀoctanal,20 [2H6]ꢀαꢀterpineol,6 [2H2]ꢀ
103 linalool,21 [2H3]ꢀ(E,E)ꢀ2,4ꢀdecadienal,22 [2H3]ꢀethyl 2ꢀmethylbutanoate,23 [2H3]ꢀ
104 vanillin,24 [2H3]ꢀethyl hexanoate,25 [2H4]ꢀnonanal,26 [2H2]ꢀbutanoic acid,27 [2H3]ꢀβꢀ
105 ionone,28 [13C4]ꢀ2,3ꢀbutanedione,29 [2H2]ꢀethyl butanoate,30 [2H3]ꢀ2ꢀmethoxyꢀ4ꢀvinylꢀ
106 phenol,31 [2H4]ꢀ(Z)ꢀ3ꢀhexenal32 and [2H4]ꢀhexanal.33
107
Most of the syntheses were performed on a micro scale basis. Thus, concentraꢀ
108 tions were determined by GC/FID as follows: An FIDꢀresponse factor was calculated
109 by analyzing defined amounts of the respective unlabeled compound and methyl ocꢀ
110 tanoate as internal standard. Using the same internal standard and the FID response
111 factor determined with the unlabeled compound, the concentration of the labeled
112 compound was calculated on the basis of a defined volume of the respective stock
113 solution.
114
[13C2]ꢀAcetaldehyde, [2H3]ꢀacetic acid, [13C8]ꢀoctanoic acid and [2H2]ꢀ1ꢀpentenꢀ3ꢀol
115 were obtained from SigmaꢀAldrich (Taufkirchen, Germany).
116
117
Syntheses.
Synthesis of [2H2]ꢀ1ꢀpentenꢀ3ꢀone. To obtain the target compound [2H2]ꢀ1ꢀpentenꢀ
118 3ꢀol was oxidized with freshly prepared Dess Martin periodinane [1,1,1ꢀtriacetoxyꢀ1,1ꢀ
119 dihydroꢀ1,2ꢀbenziodoxolꢀ3ꢀ(1H) one].
120
Preparation of Dess Martin periodinane.34 2ꢀIodobenzoic acid (40 mmol) and poꢀ
121 tassium peroxomonosulfate (60 mmol) were suspended at 70 °C in distilled water
122 (300 mL) and stirred for 3 h. After cooling to 5 °C, the reaction mixture was stirred for
123 another 90 min. The white crystals formed were filtered off, washed with distilled waꢀ
124 ter (240 mL) followed by acetone (80 mL) and finally dried in a stream of nitrogen. A
ACS Paragon Plus Environment

Page 7 of 51
Journal of Agricultural and Food Chemistry
7
125 mixture of the reaction product obtained (35 mmol), was refluxed with acetic acid anꢀ
126 hydride (50 mL) and pꢀtoluene sulfonic acid (0.6 mmol) for 2 h at 80 °C. After cooling
127 to 0 °C, the crystals formed were filtered off and washed with diethyl ether (50 mL).
128 Dess Martin periodinane was stored under argon prior to use.
129
Oxidation of [2H2]ꢀ1ꢀpentenꢀ3ꢀol. The alcohol (2.85 mmol in 10 mL anhydrous diꢀ
130 chloromethane) was added within 10 min to a solution of Dess Martin periodinane
131 (3.42 mmol in 10 mL anhydrous dichloromethane), and the solution was stirred for 1
132 h at room temperature. Diethyl ether (50 mL) and an aqueous sodium thiosulfateꢀ
133 solution (100 mL; 1 mol/L, saturated with sodium hydrogen carbonate) was added
134 and the reaction mixture was stirred for 10 min. The aqueous layer was separated
135 and extracted with diethyl ether (total volume: 90 mL). The combined organic layers
136 were subsequently washed with sodium thiosulfate (50 mL; 1 mol/L, saturated with
137 sodium hydrogen carbonate), an aqueous saturated sodium hydrogen carbonate soꢀ
138 lution (100 mL) and finally with distilled water (100 mL). The solution was then dried
139 over sodium sulfate and the target compound was characterized by mass spectromeꢀ
140 try.
141 MS (EI): m/z (%): 57 (100), 56 (92), 86 (83), 58 (50), 85 (46), 82 (21), 48 (15).
142 MS (CI): 87 (M+1, 100).
143
Synthesis of [13C8]ꢀoctanal. The compound was prepared by reduction of 13C8ꢀ
144 octanoic acid to [13C8]ꢀOctanol followed by an oxidation into the target compound usꢀ
145 ing Dess Martin periodinane.
146
Synthesis of [13C8]ꢀoctanol. [13C8]ꢀoctanoic acid (0.7 mmol in 20 mL diethyl ether)
147 was slowly added to a solution of lithium aluminium hydride (0.6 mmol in 50 mL diꢀ
148 ethyl ether) and the mixture was refluxed for 1 h. Ice water was slowly added until the
149 hydrogen formation stopped, and sulfuric acid (10%) was added dropwise to dissolve
150 the aluminium hydroxide formed. The aqueous layer was extracted with diethyl ether
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 8 of 51
8
151 (150 mL), the combined organic layers were washed with an aqueous saturated soꢀ
152 dium chloride solution (50 mL) and finally dried over sodium sulfate.
153
Synthesis of [13C8]ꢀoctanal. The alcohol was oxidized with Dess Martin periodinane
154 as described above for 1ꢀpentenꢀ3ꢀol. To remove impurities, the aldehyde was puriꢀ
155 fied by chromatography on silica gel using an nꢀpentane/diethyl ether gradient. The
156 [13C8]ꢀoctanal was eluted with nꢀpentane/diethyl ether (9:1; 100 mL) and was characꢀ
157 terized by mass spectrometry.
158 MS (EI): m/z (%): 49 (100), 44 (68), 60 (66), 42 (33), 59 (30), 90 (16), 74 (15), 106
159 (4), 118 (2).
160 MS (CI): 136 (M+1, 100).
161
Isolation of the Volatiles. Orange juice (100 mL) was extracted with diethyl ether
162 (total volume: 200 mL) by vigorous stirring for 2 h, and the volatiles were isolated by
163 solvent assisted flavor evaporation (SAFE).35 The distillate was separated into an
164 acidic fraction (AF) and a fraction containing the neutral/basic volatiles (NBF).36 Each
165 fraction was concentrated at 40 °C to obtain a final volume of about 250 ꢁL using a
166 Vigreux column (50 cm x 1.5 cm i.d.), and a micro distillation device.37
167
High Resolution Gas Chromatography/Olfactometry (HRGC/O), Aroma Ex-
168 tract Dilution Analysis (AEDA) and High Resolution Gas Chromatography/Mass
169 Spectrometry (HRGC/MS). The SAFE distillates were injected by the cold on column
170 injector onto a GC column installed in a Carlo Erba Instruments gas chromatograph
171 HRGC 8000 (Hofheim, Germany). The volatiles were separated with a helium flow
172 rate of 1.9 mL per min on a J&W Scientific DBꢀFFAP column (30 m x 0.32 mm i.d.,
173 0.25 ꢁm film thickness) (Folsom, USA) using the following parameters: 2 min held at
174 40 °C, then increased at 20 °C/min to 60 °C; held 2 min at 60 °C, then increased at 6
175 °C/min to 180 °C. Finally, the temperature was increased to 230 °C and held for 5
176 min. HRGCꢀO was performed by dividing the effluent 1:1 at the end of the capillary
ACS Paragon Plus Environment

Page 9 of 51
Journal of Agricultural and Food Chemistry
9
177 column using a Yꢀshaped glass splitter (Chromatographie Handel Müller, Germany)
178 and two deactivated, nonꢀcoated fusedꢀsilica capillaries of 50 cm each. One was
179 connected with the flame ionization detector (250 °C) and the other with the sniffing
180 port (190 °C). The extract was diluted 1:1 until no odorꢀactive compound could be
181 smelled. Thus, the flavour dilution factor indicates the last dilution in which an odorant
182 could be smelled.
183
The parameters used for the second column, a J&W Scientific DB 5 (30 m x 0.32
184 mm i.d., 0.25 ꢁm film thickness) (Folsom, USA) were as follows: 2 min held at 40 °C,
185 then increased at 9 °C/min to 50 °C; held for 2 min and increased at 6 °C/min to 180
186 °C, Finally the temperature was increased at 9 °C/min to 240 °C and held for 5 min.
187
Linear retention indices (RI) of the compounds were calculated using a series of nꢀ
188 alkanes (C6−C26 for DBꢀFFAP and C6−C18 for DBꢀ5) as previously described.36
189
HRGC/MS was performed by means of a HewlettꢀPackard gas chromatograph
190 5890 series II (Waldbronn, Germany) connected to a Finnigan sector field mass
191 spectrometer type MAT 95 S (Bremen, Germany) using the capillaries described
192 above. Mass spectra were generated in the electron impact mode (MSꢀEI) at 70 eV
193 and in the chemical ionization mode (MSꢀCI) at 115 V using isobutane as the reacꢀ
194 tant gas.
195
196
Quantitation by Stable Isotope Dilution Assays.
Gas Chromatography/Ion Trap Mass Spectrometry (GC/ITMS). Orange juice (5
197 to 1000 mL, depending on the concentration of the respective odorant determined in
198 preliminary experiments) was spiked with the respective isotopically labeled internal
199 standards either dissolved nꢀpentane or diethyl ether. The amount of the standard
200 was chosen in a similar concentration as the analyte, and the samples were equiliꢀ
201 brated for 30 min with stirring. After extraction with dichloromethane (10 to 1000 mL,
202 depending on the initial volume of the juice used), the volatiles and the labeled interꢀ
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 10 of 51
10
203 nal standards were isolated by SAFE distillation.35 After drying over sodium sulfate
204 and concentration to about 200 uL, the resulting solutions were analyzed by GC/MS.
205 Quantitation of major compounds was performed using a Varian 431 gas chromatoꢀ
206 graph (Darmstadt, Germany) equipped with the DBꢀFFAP column coupled to a Variꢀ
207 an mass spectrometer 220 (Darmstadt, Germany). Mass spectra were generated by
208 MSꢀCI using methanol as the reactant gas.
209
Two dimensional High Resolution Gas Chromatography-Gas Chromatog-
210 raphy-Ion Trap-Mass Spectrometry (GCxGC/ITMS). For the quantitation of trace
211 aroma compounds, two dimensional GCxGC/mass spectrometry was applied, using a
212 Thermo Finnigan Trace 2000 series gas chromatograph (Braunschweig, Germany)
213 equipped with a Fison Instruments moving capillary stream switching system (MCSS)
214 (MainzꢀKastel, Germany) and linked to a Varian CP 3800 gas chromatograph and a
215 Varian Saturn 2000 ion trap mass spectrometer (Darmstadt, Germany).
216
The samples were injected by the coldꢀonꢀcolumn technique and after chromatogꢀ
217 raphy on the first column (DBꢀFFAP), the odorant and the respective internal standꢀ
218 ard were cut out by means of the MCSS and transferred to the second column via a
219 heated transfer line. The effluent was condensed in a cold trap before the separation
220 was continued on the second column (DBꢀ1701). The effluent was monitored by
221 means of the ion trap mass spectrometer in the CIꢀmode with methanol as reactant
222 gas. The heart cut time in the first dimension was determined using the respective
223 reference compound.38 Response factors (Rf) were determined by analyzing mixtures
224 containing known amounts of the unlabeled target compound and the respective isoꢀ
225 topically labeled internal standard in five different ratios (5:1, 3:1, 1:1, 1:3, and 1:5) by
226 either GCꢀMS or GC×GC/MS.
227
Quantitation of Acetaldehyde. Acetaldehyde was quantitated using an enzymatic
228 assay (Boehringer Mannheim/ RꢀBiopharm AG, Darmstadt, Germany) with photometꢀ
ACS Paragon Plus Environment

Page 11 of 51
Journal of Agricultural and Food Chemistry
11
229 ric monitoring. The workup procedure and photometric measurements were carried
230 out according to the instruction manual using a 0.5 mL solution. To get rid of color
231 and proteins, orange juice (16g) was clarified with CarrezꢀIꢀsolution (1 mL; potassium
232 ferrocyanide, 85 mM = 3.6 g K4[Fe(CN)6] x 3H2O / 100 mL) and CarrezꢀIIꢀsolution (1
233 mL; zinc sulfate, 250 mM = 7.20 g ZnSO4 x 7 H2O / 100 mL) and the samples were
234 filled up in a 20 mL volumetric flask. The assay was calibrated with freshly distilled
235 acetaldehyde.
236
Quantitation of Terpene Hydrocarbons. The quantitation of the major terpene
237 hydrocarbons limonene, myrcene and αꢀpinene was carried out by HRGCꢀFID using
238 transꢀpꢀmenthꢀ2ꢀene as the internal standard. For the quantitation of myrcene and αꢀ
239 pinene 50 g of juice were used. The concentration of transꢀpꢀmenthꢀ2ꢀene was choꢀ
240 sen between the expected concentrations of myrcene and of αꢀpinene. For limonene,
241 10 g juice were used. Extraction and SAFE distillation was done as described above.
242 The terpene hydrocarbons were enriched by silica column chromatography as folꢀ
243 lows: The concentrated distillate (1 mL) was placed on top of the glass column filled
244 with silica gel column in nꢀpentane, and the terpene hydrocarbons were eluted using
245 nꢀpentane. The solution was concentrated to 200 ꢁL and subjected to HRGCꢀFID.
246 Response factors were calculated from mixtures of known concentrations of transꢀpꢀ
247 menthꢀ2ꢀene, limonene, myrcene and αꢀpinene.
248
Sensory Evaluation. Aroma Profile Analysis. Sensory analyses were carried out
249 in a sensory room designed for this purpose with individual sections for each panelist.
250 The room temperature was adjusted to 20 – 25°C, and to avoid any influence of colꢀ
251 ored solutions, the evaluations were carried out in yellow tainted light. 20 panelists
252 were recruited from the German Research Center for Food Chemistry (Freising,
253 Germany), who were regularly trained in orthonasal odor perception.39 The panelists
254 were asked to evaluate the intensity of the following odor attributes: pungent (1ꢀ
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 12 of 51
12
255 pentenꢀ3ꢀone), green/grassy ((Z)ꢀ3ꢀhexenal), orange/green (octanal), vanillaꢀlike
256 (vanillin), citrusꢀlike/flowery (linalool), cloveꢀlike (2ꢀmethoxyꢀ4ꢀvinylphenol), fruity
257 (ethyl butanoate), caramelꢀlike (4ꢀhydroxyꢀ2,5ꢀdimethylꢀ3(2H)ꢀfuranone), sour (acetic
258 acid), cabbageꢀlike (dimethyl sulfide) and fresh (acetaldehyde) on a linear scale from
259 0 (not perceivable) to 3 (strongly perceivable) in steps of 0.5 units. The attributes
260 were defined in a first session with the panelists as most relevant to describe the
261 overall aroma of orange juice. For the subsequent aroma profile analysis reference
262 solutions of the aroma compounds given in parentheses in 20 – 40ꢀfold threshold
263
concentration (in water) were presented for each attribute. Samples (15 mL) were
264 presented in covered glass vessels (i.d. = 40 mm, total volume = 45 mL) at room
265 temperature, and the results obtained in two sessions were averaged and plotted in a
266 spider web diagram.
267
Triangle Test. In order to provide information on the differentiation of sensory
268 properties of samples, triangle tests were designed and interpreted according to the
269 method ISO 4120:2004.
270
Determination of Odor Thresholds. Odor thresholds were determined in tap water
271 as recently described.39 Because of its pH dependent molecular properties, the orꢀ
272 thonasal odor threshold of acetic acid was determined in an orange juice model maꢀ
273 trix, consisting of an aqueous solution of 4.5 % sucrose, 2% glucose, 2 % fructose,
274 1.1 % citric acid, 0.5 % ascorbic acid and 0.1 % sunflower oil (the pH was adjusted to
275 3.6).
276
Odor Activity Values. These were calculated as previously described.38
ACS Paragon Plus Environment

Page 13 of 51
Journal of Agricultural and Food Chemistry
13
277 Results and Discussion
278
Aroma Profile of unstored NFC juice from Hamlin oranges. First, the overall
279 aroma of the unstored juice was evaluated by an aroma profile analysis using 11
280 odor attributes agreed upon by the sensory panel in preliminary sessions. The evaluꢀ
281 ation of the intensity of each attribute was trained using solutions of single odorants
282 representing the respective odors in a concentration by factors of approximately 50
283 above their odor threshold. The major odor attributes described in the unstored Hamꢀ
284 lin orange juice were fresh with an intensity of 2.3 (Figure 1), followed by orꢀ
285 ange/green (1.8), pungent (1.7), fruity (1.6) and green/grassy (1.5). The other six atꢀ
286 tributes citrus/flowery, sour, caramelꢀlike, cloveꢀlike, vanillaꢀlike and cabbageꢀlike
287 were ranked with a lower intensity.
288
Identification of Aroma-Active Compounds in Hamlin orange juice. To eluciꢀ
289 date the compounds contributing to the overall aroma of the unstored Hamlin juice,
290 an aroma extract dilution analysis (AEDA) was carried out on the entire set of volaꢀ
291 tiles isolated by extraction and SAFE distillation. A total of 51 aromaꢀactive regions
292 could be located in the FD factor range of 8 to 8192 (Figure 2) among which two
293 compounds with a vanillaꢀlike (51) and a coconutꢀlike odor attribute (45) showed the
294 highest FD factors of 8192, followed by two compounds with a flowery (26) and a
295 smoky odor (47). Further compounds with a high FD factor were 3 (fir needle ꢀlike), 7
296 (green), 10 (citrusꢀlike), 21 (sour), 36 (deepꢀfried), 41 (metallic) and 42 (caramelꢀlike).
297 A comparison of the retention indices, mass spectra and perceived odors (quality and
298 intensity) with the institute’s database containing analytical and sensory data of apꢀ
299 proximately 1000 aroma compounds resulted in proposals for the chemical strucꢀ
300 tures. These were confirmed by the analysis of reference compounds.
301
The results of the identification experiments in combination with the FD factors
302 showed that vanillin (51; Figure 3), 3a,4,5,7aꢀtetrahydroꢀ3,6ꢀdimethylꢀ2(3H)ꢀ
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 14 of 51
14
303 benzofuranone (winelactone) (45) followed by (R)ꢀlinalool (26) caused the respective
304 odors of the odorꢀactive areas detected by AEDA (Figure 2). The enantiomeric purity
305 of linalool was higher than 95% (data not shown). Further odorants with high FD facꢀ
306 tors were αꢀpinene (3), hexanal (7), (R)ꢀlimonene (10), acetic acid (21), (E,E)ꢀ2,4ꢀ
307 decadienal (35), transꢀepoxyꢀ(E)ꢀ2ꢀdecenal (40) and 4ꢀhydroxyꢀ2,5ꢀdimethylꢀ
308 3(2H)furanone (41). The results of the further identification experiments (Table 1)
309 showed that 45 out of the 51 odorants detected by GC/Olfactometry could be identiꢀ
310 fied. In general, the results confirmed data of previous studies5,9,12,13,15,40ꢀ50 aimed at
311 identifying either volatile or odorꢀactive compounds in orange juices of different oriꢀ
312 gins and varieties. However, 2ꢀethylꢀ3ꢀ and 2ꢀethylꢀ5ꢀmethylpyrazine (16) and 3ꢀ
313 hydroxyꢀ4,5ꢀdimethylꢀ2(5H) furanone (46) are reported here as additional odorants in
314 orange juice, although both showed only low FD factors.
315
Influence of a chilled storage To elucidate changes occurring during chilled storꢀ
316 age of the Hamlin juice, its aroma profile was determined after 1, 2 and 10 months of
317 storage (Figure 4). A comparison with the profile of the unstored juice (black line in
318 Figure 4) indicated that in particular the freshness of the juice was decreased during
319 storage and also the pungent and green, grassy odor attributes were clearly lower in
320 the stored juices. This loss in odor intensity was already detectable after 1 month of
321 storage (dark green line in Figure 4).
322
To clarify the odorants responsible for these differences, the AEDA was applied on
323 a SAFE distillate obtained from the 10 months stored juice (10m; Table 2). A comparꢀ
324 ison with the results obtained for the unstored juice (Ust; Table 2) revealed that the
325 majority of the 51 odorants was qualitatively identical in both samples. But, 4 comꢀ
326 pounds were no longer detectable by GC/O in the stored juice, i.e., 1ꢀpentenꢀ3ꢀone
327 (4), (Z)ꢀ3ꢀhexenal (8), (E,Z)ꢀ2,6ꢀnonadienal (28) and 3ꢀhydroxyꢀ4,5ꢀdimethylꢀ2(5H)
328 furanone (46). Because differences of two or one dilution step lie within the error
ACS Paragon Plus Environment

Page 15 of 51
Journal of Agricultural and Food Chemistry
15
329 range of the AEDA, only the differences in FD factors determined for the following six
330 compounds may give additional hints, which odorant loss may contribute to the aroꢀ
331 ma differences detected in the aroma profiles: hexanal (7), octanal (14), nonanal
332 (18), (E,E)ꢀ2,4ꢀdecadienal (35), 4ꢀhydroxyꢀ2,5ꢀdimethylꢀ3(2H)ꢀfuranone (41) and vanꢀ
333 illin (51). On the other hand, the FD factor of only one compound, namely 2ꢀmethoxyꢀ
334 4ꢀvinylphenol (44) was higher in the stored juice compared to the unstored juice (Taꢀ
335 ble 2).
336
Quantitative Analysis. To get nearer to the reason for the aroma differences in
337 the juice, quantitative data were correlated with the odor thresholds of the odorants
338 by calculating odor activity values (ratio of concentration to odor threshold). Thus,
339 first a quantitative analysis of aroma compounds showing clear differences in their
340 FD factors as well as of compounds proposed in the previous literature as markers
341 for offꢀflavor in orange juices was performed mainly by means of stable isotope diluꢀ
342 tion assays.
343
In the unstored juice, (R)ꢀlimonene showed by far the highest concentration of
344 about 121 mg/kg; followed by acetaldehyde with 16.1 mg/kg, acetic acid with 5.5
345 mg/kg and myrcene with 1.5 mg/kg (Table 3). Moreover, αꢀpinene, linalool, hexanal,
346 ethyl butanoate, decanal, octanal and nonanal were found in concentrations between
347 99.6 and 757 ꢁg/kg. 1ꢀPentenꢀ3ꢀone, vanillin, (Z)ꢀ3ꢀhexenal and 2ꢀmethoxyꢀ4ꢀ
348 vinylphenol were measured in a lower concentration range between 0.74 and 35
349 ꢁg/kg.
350
Compared to the unstored juice, already after one month the concentrations of acꢀ
351 etaldehyde and of (Z)ꢀ3ꢀhexenal decreased significantly (Table 3), and 1ꢀpentenꢀ3ꢀ
352 one was decomposed to about one tenth of its initial concentration. It should, howevꢀ
353 er, be mentioned that (Z)ꢀ3ꢀhexenal is not stable during a freezing/thawing proꢀ
354 cess.2,51 So, its initial concentration in the unstored juice was probably much higher.
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 16 of 51
16
355 Additionally, hexanal, nonanal and decanal decreased in their concentrations. On the
356 other hand, acetic acid and 2ꢀmethoxyꢀ4ꢀvinylphenol showed an increase already
357 after 1 month (Table 3). Monitoring the changes in concentrations of the 15 aroma
358 compounds after ten months indicated a continuous loss of (R)ꢀlimonene, myrcene,
359 aꢀpinene, (Z)ꢀ3ꢀhexenal as well as of hexanal, octanal, nonanal and decanal. In parꢀ
360 ticular, 1ꢀpentenꢀ3ꢀone and acetaldehyde fell below their detection threshold already
361 after two months of chilled storage. While e.g., (R)ꢀlinalool and ethyl butanoate were
362 quite stable, 3 compounds increased after ten months, namely acetic acid, 2ꢀ
363 methoxyꢀ4ꢀvinylphenol and vanillin (Table 3).
364
To confirm the trend in the different stabilities of the juice odorants, the storage triꢀ
365 al was repeated using NFC juice from Hamlin oranges of the following harvest year.
366 In this trial, further 6 compounds were added to the set of 15 aroma compounds anaꢀ
367 lyzed in the first year: ethyl hexanoate, (E,E)ꢀ2,4ꢀdecadienal, βꢀionone, αꢀterpineol
368 and carvone The latter two odorants have previously been suggested as offꢀflavor
369 compounds in orange juice.7,9,14 The same trends in the stability of the odorants were
370 observed (Table 4). Besides ethyl butanoate, also further two esters, (S)ꢀethyl 2ꢀ
371 methylbutanoate and ethyl hexanoate, were quite stable during storage for six
372 months, and the same was true for (R)ꢀlinalool. However, 1ꢀpentenꢀ3ꢀone and acetꢀ
373 aldehyde rapidly disappeared during storage as already observed for the juice from
374 the year before. In addition, (Z)ꢀ3ꢀhexenal and (E,E) 2,4ꢀdecadienal as well as all
375 saturated aldehydes were clearly reduced during storage. But, the previously sugꢀ
376 gested offꢀflavor compounds in orange juice, carvone and vanillin, did not show an
377 increase compared to the unstored juice. But, the concentration of αꢀterpineol inꢀ
378 creased from 183 ꢁg/kg to nearly the tenꢀfold amount of 1100 ꢁg/kg.
379
As a next step, a calculation of odor activity values was done on the unstored
380 Hamlin juice as well as on the five stored samples (Table 4 and Table 5). In the unꢀ
ACS Paragon Plus Environment

Page 17 of 51
Journal of Agricultural and Food Chemistry
17
381 stored juice, (R)ꢀlimonene, (R)ꢀlinalool, myrcene and acetaldehyde followed by ethyl
382 butanoate showed the highest odor activity values. Neither αꢀterpineol, carvone and
383 vanillin nor 2ꢀmethoxyꢀ4ꢀvinylphenol were present above their odor threshold, while
384 acetic acid was already present with an OAV of 8.9 already in the unstored juice. A
385 comparison of the OAVs in the unstored juice and the 6 months stored juice clearly
386 indicated that in particular 1ꢀpentenꢀ3ꢀone and acetaldehyde fell below their odor
387 threshold after 6 months and, are, thus, lost as contributors to the overall aroma proꢀ
388 file. A reduced effect in the aroma contribution indicated by lower OAVs also ocꢀ
389 curred for hexanal, octanal, nonanal, decanal, (Z)ꢀ3ꢀhexanal and (E,E)ꢀ2,4ꢀ
390 decadienal.
391
392
Aroma-Active Compounds in NFC Valencia Orange Juice.
To indicate whether similar aroma losses also occur in juices from other varieties,
393 the same series of experiments was done on NFC juice from Valencia oranges. The
394 study started with the aroma profile analysis of the unstored juice and a juice stored
395 for 10 months (Figure 5). The most distinct aroma impression in the unstored juice
396 was orange/green” (2.0) and “fresh” (1.7), followed by the attributes “pungent” (1.4),
397 “fruity” (1.5) and “citrus/flowery” (1.4). The qualities “green/grassy” (1.3), “sour“ (1.1)
398 and “caramelꢀlike” (1.0) were rated less intense, while the odor impressions “cloveꢀ
399 like” (0.7), “vanillaꢀlike” (0.5) and “cabbageꢀlike” (0.2) showed a low impact on the
400 overall aroma of this juice. In the 10 months stored juice, the most obvious loss of
401 intensity was detected for the odor quality orange/green. The changes were similar
402 compared to those observed for the Hamlin juice (Figure 4).
403
An aroma extract dilution analysis applied on the unstored NFC juice from Valenꢀ
404 cia oranges followed by identification experiments revealed the same odorꢀactive
405 compounds in the Valencia NFC juice as in the Hamlin juice, although the FD factors
406 were different (see supplementary information) pointing to differences in the amounts
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 18 of 51
18
407 of several odorants between both unstored juices. Furthermore, an AEDA applied on
408 the stored Valencia juice also resulted in changes for the same aroma compounds
409 (data not shown). Therefore, the same set of aroma compounds as analyzed in the
410 Hamlin juices was also quantitated in the unstored and the stored Valencia juice (Taꢀ
411 ble 6). In the unstored juice, (R)ꢀlimonene, acetic acid and acetaldehyde followed by
412 linalool and myrcene showed the highest concentrations. Compared to the unstored
413 juice from the Hamlin oranges, in particular the concentrations of (R)ꢀlimonene, linalꢀ
414 ool as well as octanal, nonanal and decanal showed higher concentrations in the Vaꢀ
415 lencia juice (compare data in Tables 4 and 6). These differences might be the reason
416 for a more intense orangeꢀlike/green note in the overall aroma profile of the Valencia
417 juice (Figure 1 and Figure 5).
418
When the stability of the compounds during chilled storage was monitored, the
419 same trends as for Hamlin orange juice were observed (Table 6). For example, 1ꢀ
420 pentenꢀ3ꢀone and acetaldehyde showed a very rapid decrease and all unsaturated
421 and saturated aldehydes were significantly degraded during storage. On the other
422 hand, (R)ꢀlinalool and all esters were stable, and besides vanillin, in particular acetic
423 acid, αꢀterpineol as well as 2ꢀmethoxyꢀ4ꢀvinylphenol increased in concentration. The
424 calculation of odor activity values revealed (R)ꢀlinalool followed by (R)ꢀlimonene and
425 myrcene with the highest odor activities followed by acetaldehyde and ethyl butaꢀ
426 noate in the unstored NFC juice (Table 7). Except acetaldehyde, which was below its
427 odor threshold in the 10 months stored juice, the other 4 compounds still showed the
428 highest odor activity after a 10 months storage. However, significant losses in odor
429 activity were measured for (E,E)ꢀ2,4ꢀdecadienal, hexanal, octanal, nonanal and deꢀ
430 canal. But, also in the NFC Valencia juice neither 2ꢀmethoxyꢀ4ꢀvinylphenol, vanillin
431 and αꢀterpineol nor carvone were found as contributors to the changes in the overall
432 aroma, i.e. as offꢀflavor compounds. By application of the same concept on a Valenꢀ
ACS Paragon Plus Environment

Page 19 of 51
Journal of Agricultural and Food Chemistry
19
433 cia NFC juice from a second harvest year nearly identical results were obtained (data
434 not shown).
435
436
437
Sensory experiments on odorant contribution
During storage of the Hamlin juice, exclusively a degradation of odorants was deꢀ
438 tected, but no significant increase in any odor active compound was observed. Thus,
439 spiking a stored juice with the lost amounts of the odorants should bring back the
440 aroma profile of the unstored juice. Thus, the following sensory experiments were
441 carried out to study this assumption (Figure 6). In a six months stored juice the conꢀ
442 centrations of 12 aroma compounds were made up to their initial concentrations deꢀ
443 termined in the unstored juice (Table 8). While the six months stored juice and the
444 unstored juice were clearly distinguishable by their overall aroma profiles (data not
445 shown), the spiked juice could not be distinguished from the unstored juice (Table 8)
446 (α= 0.2). In a second set of experiments, aroma compounds were grouped according
447 to three aroma impressions, namely “green”, “fresh/ pungent”, and “citrus/ orangeꢀ
448 like” and subgroups correlating with this odor attribute were separately administered
449 to the six months stored juice. The stored juice spiked with the two “green” comꢀ
450 pounds (Z)ꢀ3ꢀhexenal and hexanal was clearly differentiated (α= 0.01) (Table 8) from
451 the unstored juice. However, after administering only the “fresh, pungent” compounds
452 1ꢀpentenꢀ3ꢀone and acetaldehyde, the panel could not differentiate between the
453 spiked and the unstored sample (α= 0.2). In addition, also spiking 4 aldehydes,
454 namely octanal, nonanal, decanal and (E,E)ꢀ2,4ꢀdecadienal showed a positive effect
455 on the aroma of the stored juice. Also this partial recombinate could no longer be disꢀ
456 tinguished from the unstored juice (α= 0.2). These results corroborate the importance
457 of 1ꢀpentenꢀ3ꢀone as well as acetaldehyde, octanal, nonanal, decanal and (E,E)ꢀ2,4ꢀ
458 decadienal for the overall aroma of orange juice.
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 20 of 51
20
459
With respect to fresh orange juice, these results confirm data of previous investigaꢀ
460 tions from our working group on fresh orange juices2,5,6,15 suggesting that besides
461 esters, such as ethyl butanoate, and terpenoids like (R)ꢀlimonene, several comꢀ
462 pounds known as degradation products of lipid hydroperoxides, such as 1ꢀoctenꢀ3ꢀ
463 one, (Z)ꢀ3ꢀhexenal and (E,E)ꢀ2,4ꢀdecadienal and saturated aldehydes are important
464 contributors to the aroma of unstored (fresh) orange juices. In addition, 1ꢀpentenꢀ3ꢀ
465 one and acetaldehyde were established as an important key odorant in NFC orange
466 juices. The formation of 1ꢀpentenꢀ3ꢀone can be assumed from the 15ꢀhydroperoxide
467 of linolenic acid by a radical mechanism (Figure 7) which is similar to a mechanism
468 proposed for the formation of 1ꢀoctenꢀ3ꢀone from linoleic acid in mushrooms.52 Howꢀ
469 ever, in particular the oxidation of the intermediate alcohol to the ketone has not yet
470 been proven in both cases. Thus, it might be assumed that 1ꢀpentenꢀ3ꢀone is enzyꢀ
471 matically formed in the orange fruit.
472
Furthermore, while previous investigations15,16 focused on the storage of juice
473 made from concentrate at room temperature or even under forced conditions, these
474 investigations focused for the first time on changes of the aroma compounds of NFC
475 orange juice during chilled storage, which is common in the juice industry. Our results
476 clearly show that despite these gentle storage conditions, changes in the overall
477 aroma profiles of juices from two orange varieties occurred leading to a decrease in
478 the intensity of several aroma attributes compared to the unstored juice. However, an
479 important role of previously suggested offꢀflavor compounds in orange juice, such as
480 carvone, 2ꢀmethoxyꢀ4ꢀvinylphenol or vanillin for changes in the overall juice aroma
481 could be ruled out for such chilled stored juices.
482
But, as confirmed by sensory experiments, the rapid decomposition of 1ꢀpentenꢀ3ꢀ
483 one and acetaldehyde are particularly involved in the loss of the typical aroma of the
484 unstored juice. In addition, the continuous degradation of (Z)ꢀ3ꢀhexenal, hexanal, ocꢀ
ACS Paragon Plus Environment

Page 21 of 51
Journal of Agricultural and Food Chemistry
21
485 tanal, nonanal, decanal, and (E,E)ꢀ2,4ꢀdecadienal for sure contribute to the overall
486 changes in juice aroma. However, neither the pathways for the decomposition of the
487 1ꢀpentenꢀ3ꢀone nor for the losses of the aldehydes are yet known. This challenge will
488 be addressed in further investigations.
489
Acknowledgments. We thank Tropicana Products Inc./ PepsiCo Global R & D for
490 funding this project.
491
SUPPORTING INFORMATION
492 Table S1. Odorants, isotopically labelled standards, response factors and mass tracꢀ
493
es used for the quantitation of aroma compounds in orange juices
494 Table S2. Most odorꢀactive volatiles (FD ≥ 16) in the unstored NFC juice from Valenꢀ
495 cia oranges
496
497
498
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 22 of 51
22
499 REFERENCES
500 (1)
501
Verband der deutschen Fruchtsaftindustrie e. V. (German Association of the
Fruit Juice Industry), Bonn, Germany;
https://www.fruchtsaft.de/presse/grafiken/ (last access 03/21/2018)
502
503 (2)
504
Seideneck, R.; Schieberle, P. Comparison of the key aroma compounds in
handꢀsqueezed and unpasteurised, commercial NFC juices prepared from
Brazilian Pera Rio oranges. Eur. Food Res. Technol. 2011, 232, 995ꢀ1005.
Hall, J. H.; Wilson, C. P. The volatile constituents of Valencia orange juice. J.
Am. Chem. Soc. 1925, 47, 2575ꢀ2584.
505
506 (3)
507
508 (4)
509
VCF Volatile compounds in food: database; version 16.4: orange juice. Nijsꢀ
sen, L. M., van IngenꢀVisscher, C. A., Donders, J. J.H., Eds.; Triskelion B.V.:
Zeist, the Netherlands, 1963 – 2017
510
511 (5)
512
Hinterholzer, A.; Schieberle, P. Identification of the most odourꢀactive volatiles
in fresh, handꢀextracted juice of Valencia Late orange juices by odour dilution
techniques. Flavour Frag. J. 1998, 13, 49ꢀ55.
513
514 (6)
515
Büttner; A. Important aroma compounds in freshly squeezed juices of different
orange varieties (Citrus sinensis (L.) Osbeck) and grapefruit (Citrus paradisi
Macf.) (in German). 1999, PhD thesis, Technical University of Munich, Germaꢀ
ny
516
517
518 (7)
519
PerezꢀCacho, P. R.; Rouseff, R. Processing and storage effects on orange
juice aroma: a review. J. Agric. Food Chem. 2008, 56, 9785ꢀ9796.
Shaw, P .E.; Wilson, C. W. Volatile sulfides in headspace gases of fresh and
processed citrus juices. J. Agric. Food Chem. 1982, 30, 685ꢀ688.
Kirchner, J. G.; Miller, J. M. Volatile waterꢀsoluble and oil constituents of Vaꢀ
lencia orange juice. J. Agric. Food Sci. 1957, 5, 283ꢀ291.
520 (8)
521
522 (9)
523
ACS Paragon Plus Environment

Page 23 of 51
Journal of Agricultural and Food Chemistry
23
524 (10) Nagy, S.; Rousseff, R. L.; Lee, H. S. Thermally degrade flavors in citrus juice
525
526
products. In Thermal generation of aroma; Parliament, T., McGorrin, R. J., Ho,
CꢀT., Eds., American Chemical Society, Washington, 1989, pp.331 – 334
527 (11) Rouseff, R. L.; Nagy, S.; Naim, M.; Zehavi, U. Offꢀflavor development in citrus
528
529
juice products. In Offꢀflavors in food and beverages, Charalambous, G., Ed;
Elsevier Science Publishers, Amsterdam, the Netherlands, 1992, pp. 211ꢀ227
530 (12) Schreier, P.; Drawert, F.; Junker, A.; Mick, W. II. Aroma compounds in orangꢀ
531
532
es and their changes during juice processing. (in German) Z. Lebensm. Unꢀ
ters. Forsch. 1977, 164, 188ꢀ193.
533 (13) Tatum, J. H.; Nagy, S.; Berry, R. E. Degradation products formed in canned
534 singleꢀstrength orange juice during storage. J. Food Sci. 1975, 40, 707ꢀ709.
535 (14) Moshonas, M. G.; Shaw, P. E. Changes in composition of volatile components
536
537
in aseptically packed orange juice during storage. J. Agric. Food Chem. 1989,
37, 157ꢀ161.
538 (15) Averbeck, M.; Schieberle P. Characterization of the key aroma compounds in
539
540
a freshly reconstituted orange juice from concentrate. Eur. Food Res. Technol.
2009, 229, 611ꢀ622.
541 (16) Averbeck, M.; Schieberle, P. Influence of different storage conditions on
542
543
changes in the key aroma compounds of orange juice reconstituted from conꢀ
centrate. Eur. Food Res. Technol. 2011, 232, 129ꢀ142.
544 (17) Schieberle, P.; Grosch, W. Potent odorants of the wheat bread crumb. J. Agric
545 Food Chem. 1991, 35, 252ꢀ257.
546 (18) Guth, H. Determination of the configuration of wine lactone. Helv. Chem. Acta.
547 1996, 79, 1559ꢀ1571.
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 24 of 51
24
548 (19) Moshonas, M. G.; Shaw, P. E. Isolation of trans,transꢀ2,4ꢀdecadienal and inꢀ
549
550
termedeol from coldꢀpressed citrus oils. J. Agric. Food Chem. 1979, 27, 210ꢀ
211.
551 (20) Blekas, G.; Guth, H. Evaluation and quantification of potent odorants of Greek
552 virgin olive oils. Dev. Food Sci. 1995, 37, 419ꢀ427.
553 (21) Fritsch, H. Impact of hops on important aroma compounds in Pilsner beer and
554
555
in intermediates of the brewing process (in German). 2001, PhD thesis, Techꢀ
nical University of Munich, Germany
556 (22) Guth, H.; Grosch, W. Deterioration of soybean oil: quantification of primary
557
558
flavor compounds using a stable isotope dilution assay. Lebensm. Wiss.
Techn. 1990, 23, 513–522.
559 (23) Guth, H.; Grosch, W. Quantification of potent odorants of virgin olive oil by
560 stable dilution assay. J. Am. Oil Chem. Soc. 1993, 70, 513ꢀ518.
561 (24) Guth, H.; Grosch, W. Odorants of extrusion products of oat meal ꢀ changes
562 during storage (in German). Z. Lebensm. Unters. Forsch. 1993, 196, 22ꢀ28.
563 (25) Guth, H. Quantification and sensory studies of character impact odorants of
564 different white wine varieties. J. Agric. Food Chem. 1997, 45, 3027ꢀ3032.
565 (26) Kerscher, R. Differences in the aromas of thermally treated meat from different
566
567
animal species (in German). 2000, PhD thesis, Technical University of Munich,
Germany
568 (27) Kirchhoff, E. K. Characterization of key aroma compounds in rye bread (in
569
German), 2000, PhD thesis, Technical University of Munich, Germany
570 (28) Kotseridis, Y.; Baumes, R.; Skouroumounis, G. Synthesis of labelled [2H4]ꢀ
571
572
573
damascenone, [2H2]ꢀ2ꢀmethoxyꢀ3ꢀisobutylpyrazine, [2H3]ꢀαꢀionone, and [2H3]ꢀβꢀ
ionone, for quantification in grapes, juices and wines. J. Chrom. 1998, 824, 71ꢀ
78.
ACS Paragon Plus Environment

Page 25 of 51
Journal of Agricultural and Food Chemistry
25
574 (29) Rigby, W. New reagent for the oxidation of acyloins to diketons. J. Chem. Soc.
575 1951, 793ꢀ795.
576 (30) Schieberle, P.; Hofmann, T. Evaluation of characteristic impact odorants in
577
fresh strawberry juice by quantitative measurements and sensory studies on
model mixtures. J. Agric. Food Chem. 1997, 45, 227ꢀ232.
Semmelroch, P. Unravelling the aroma differences between Arabia and Roꢀ
busta coffee (in German). 1995, PhD thesis, Technical University of Munich,
Germany
578
579 (31)
580
581
582 (32) Steinhaus, M. Important aroma compounds in hops (in German). 2000, PhD
583 thesis, Technical University of Munich, Germany
584 (33) Steinhaus, M.; Sinuco, D.; Polster, J.; Osorio, C.; Schieberle, P. Characterizaꢀ
585
586
587
tion of the key aroma compounds in pink guava (Psidium guajava L.) by
means of aroma reꢀengineering experiments and omission tests. J. Agric.
Food Chem. 2009, 57, 2882ꢀ2888.
588 (34) Börger, D. S. Correlation of odor character and chemical structure in homoloꢀ
589
590
gous series of unbranched and methyl branched allyl ketones and allyl alcoꢀ
hols (in German). 2008, PhD thesis, Technical University of Munich, Germany
591 (35) Engel, W.; Bahr, W.; Schieberle, P. Solvent assisted flavor evaporation. A new
592
593
594
and versatile technique for the careful and direct isolation of aroma comꢀ
pounds from complex food matrixes. Z. Lebensm. Unters. Forsch. A. 1999,
209, 237ꢀ241.
595 (36) Schieberle, P. Primary odorants in popcorn. J. Agric. Food Chem. 1991, 39,
596 1141ꢀ1144.
597 (37) Bemelmans, J. M. H. Review of isolation and concentration techniques. In
598
599
Progress in Flavour and Research; Land, G. G., Nursten H. E., Eds.; Proceedꢀ
ings, Appl. Sci.: London, UK, 1979; pp 79ꢀ98
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 26 of 51
26
600 (38) Schieberle, P. Recent developments in methods for analysis of flavor comꢀ
601
602
603
pounds and their precursors. In Characterization of Food: Emerging Methods;
Goankar, A. G., Ed.; Elsevier Science B.V.: Amsterdam, the Netherlands,
1995, pp. 403ꢀ431
604 (39) Czerny, M.; Christlbauer, Ma.; Christlbauer, Mo.; Fischer, A.; Granvogl, M.;
605
606
607
608
Hammer, M.; Hartl, C.; Moran Hernandez, N.; Schieberle, P. Re‐investigation
on odor thresholds of key food aroma compounds and development of an
aroma language based on odor qualities of defined aqueous odorant solutions.
Eur Food Res Technol. 2008, 228, 265–27.
609 (40) Coleman, R. L.; Shaw, P. E. Analysis of Valencia orange essence and aroma
610 oils. J. Agric. Food Chem. 1971, 19, 520ꢀ523.
611 (41) Pino, J.; Tapanes, R.; Roado, A.; Baluja, R. Analysis of volatile components in
612 Valencia orange juice from Cuba. Acta Aliment. 1986, 15, 291ꢀ297.
613 (42) Wolford, R. W.; Attaway, J. A. Analysis of recovered natural orange flavor enꢀ
614
615
hancement materials using gas chromatography. J. Agric. Food Chem. 1967,
15, 369ꢀ377.
616 (43) Wolford, R. W.; Attaway, J. A.; Alberdings, G. E.; Atkins, C. D. Analysis of the
617
618
flavor and aroma constituents of orange juices by gas chromatography. J.
Agric. Food Sci. 1963, 28, 320ꢀ328.
619 (44) Schultz, T. H.; Flath, R. A.; Mon, T. R. Analysis of orange volatiles with vapor
620 sampling. J. Agric. Food Chem. 1971, 19, 1060ꢀ1065.
621 (45) Lin, J. C. C.; Nagy, S.; Kim, M. Application of pattern recognition techniques to
622
623
sensory and gas chromatographic flavor profiles of natural orange aroma.
Food Chem. 1993, 47, 235ꢀ245.
ACS Paragon Plus Environment

Page 27 of 51
Journal of Agricultural and Food Chemistry
27
624 (46) Fischer, A. Characterisation of the odourꢀactive compounds in peel oils of Jeꢀ
625
626
ruk Pontinak orange (Citrus nobilis Lour. var. microcarpa Hassk.) and Brazilian
green mandarin (Citrus reticulata). 2007, Ph. D. Dissertation, TU München.
627 (47) Schultz, T. H.; Teranishi, R.; McFadden, W. H.; Kilpatrick, P. W.; Corse, J.
628
629
Volatiles from oranges II. Constituents of the juice identified by mass spectra.
J. Food Sci. 1964, 29, 790ꢀ795.
630 (48) Boelens, M. H.; Jimenez Sindreu, R. Essential oils from Seville bitter orange
631 (Citrus aurantium L. ssp. amara L.). Dev. Food Sci. 1988, 18, 551ꢀ65.
632 (49) Attaway, J. A.; Wolfod, R. W.; Alberding, G. E.; Edwards, G. J. Identification of
633
634
alcohols and volatile organic acids from natural orange essence. J. Agric.
Food Chem. 1964, 12, 118ꢀ121.
635 (50) Farnworth, E. R.; Lagacé, M.; Couture, R.; Yaylayan, V.; Stewart, B. Thermal
636
637
processing, storage conditions and the composition and physical properties of
orange juice. Food Res. Int. 2001, 34, 25ꢀ30.
638 (51) Kreissl, J.; Schieberle, P. Characterization of aromaꢀactive compounds in Italꢀ
639
640
ian tomatoes with emphasis on new odorants. J. Agric. Food Chem. 2017, 64,
5198ꢀ5208.
641 (52) Wurzenberger, M.; Grosch, W. The enzymatic oxidative breakdown of linoleic
642
643
acid in mushrooms (Psalliota bispora). Z. Lebensm. Unters. Forsch. 1982,
175, 186ꢀ190.
ACS Paragon Plus Environment

Journal of Agricultural and Food Chemistry
Page 28 of 51
28
644 FIGURE CAPTIONS
645 Figure 1. Aroma profile of the unstored NFC juice from Hamlin oranges.
646 Figure 2 Flavor dilution (FD) chromatogram obtained by application of the aroma exꢀ
647 tract dilution analysis on the volatiles isolated from unstored NFC juice from Hamlin
648 oranges.
649 Figure 3. Structures of the aroma active compounds identified with the highest flavor
650 dilution factors in NFC juice from Hamlin oranges. Numbering refers to Table 1.
651 Figure 4. Comparative aroma profile of the unstored NFC juice (black), the 1 month
652 stored juice (dark green), the 2 months stored juice (light green) and the 10 months
653 stored juice (orange) prepared from Hamlin oranges.
654 Figure 5. Aroma profile of the unstored NFC juice from Valencia oranges (green) and
655 the 10 months stored juice (blue)
656 Figure 6. Triangle test designed for the spiking experiments of the 6 months stored
657 Hamlin orange juice (6 m) in comparison to the unstored juice (Ust).
658 Figure 7. Hypothetical pathway leading to the formation of 1ꢀpentenꢀ3ꢀone by a degꢀ
659 radation
of
the
15ꢀhydroperoxide
of
linolenic
acid.
ACS Paragon Plus Environment

Page 29 of 51
Journal of Agricultural and Food Chemistry
29
Table 1. Most odorꢀactive volatiles (FD ≥ 8) in the unstored NFC juice from Hamlin oranges
RI ond)
FD
No.a) Odorantb)
Odor qualityc)
Fractiong)
NBF 3
NBF 3
NBF 1
NBF 3
NBF 2
NBF 2
NBF 3
NBF 3
NBF 1
NBF 1
NBF 3
NBF 2
NBF 3
NBF 3
NBF 3
FFAP
950
DBꢀ5
713
607
933
682
801
845
799
810
991
1028
1035
999
800
1003
976
factore) Literatureh)
40
1
2
3
4
5
6
7
8
9
ethyl propanoate
fruity, glueꢀlike
butterꢀlike
8
41
2,3ꢀbutanedione
αꢀpinene
972
32
42
resinꢀlike, fir needleꢀlike
pungent, train oilꢀlike
fruity
1011
1024
1029
1040
1074
1100
1160
1198
1209
1229
1273
1282
1291
512
40
1ꢀpentenꢀ3ꢀone
ethyl butanoate
(S)ꢀethyl 2ꢀmethylbutanoate
hexanal
32
43
256
44
fruity
8
9
green, grassy
512
45
(Z)ꢀ3ꢀhexenal
myrcene
green, grassy
16
9
geraniumꢀlike, carrotꢀlike
citrusꢀlike, carrotꢀlike
eucalyptusꢀlike
fruity, pineappleꢀlike
butterꢀlike, carrotꢀlike
citrusꢀlike, green
mushroomꢀlike
128
9
10 (R)ꢀlimonene
11 1,8ꢀcineol
512
46
64
47
12 ethyl hexanoate
13 2ꢀhydroxyꢀ3ꢀbutanone
14 octanal
64
12
64
9
256
15 1ꢀoctenꢀ3ꢀone f)
64
5
2ꢀethylꢀ3ꢀ and 2ꢀethylꢀ5ꢀ
ꢀ
16 methylpyrazine f)
earthy
NBF 3
1361
1000
64
ACS Paragon Plus Environment