Detection of Extracts from Pineapple Peels and Cores in Juice Pure Juice Project WP 8
Polyphenolic Screening & Isotopic Testing
by
Dr. David A. Hammond
Authenticity Services
What problems are we looking to stop?
Problem I• Pineapple juice is often produced as a by product in
factories that manufacture canned products (i.e. rings, chucks, crushed).
• In these factories there is an excess of skins and cores. These can be processed on the “juice” line extending the product with sub-standard material.
Problem II• “Cloud” stability in pineapple juice is sometimes bad, due
to poor fruit quality &/or processing conditions. Under these conditions pectin or another hydrocolloid may be added to enhance the product’s stability and not declared.
Typical Processing of Pineapples
Juice material (35%)
Juice Presses
NFC Juice (32%)
SS PJ (34%)
Concentrate production?
Drained juice (2%)
Whole Cylinders (27%)
Trimming table
Table end material
Crushed preparation
Crushed (9%)
Rings and/or chunks (16%)
Decrowned pineapples (topped and tailed)
Peeling(Ginace machines)
In: Fruit juice processing technology, Edited by Nagy, Chen and Shaw. Pub. By AGScience
Rejected fruit/other materials (36.5%)
By product juice presses
Wet pulp
Mill Juice (26.0%)
Ion exchange/conc
Bran (3.5%)Liquid Sugar (2%)
Drying
Existing Methods
Problem I• Distortion of the glucose to fructose ratio (> 1.2) • Inclusion of excessive levels of skins and cores• Elevation of some of the minerals (potassium, calcium and
phosphate)• Inclusion of excessive amounts of skin• Elevated levels of nitrate ions • Inclusion of excessive amounts of core• S. Wallrauch. Flussiges Obst (59) 92/1 20 - 26
Problem II• Detection of elevated pectin levels in juice using IFU 26
(ethanolic ppt and carbazole method)• Addition of pectin (apple or citrus) as a stabiliser
Project Objectives
• RSSL• To determine the usefulness of a polyphenol profiling
method to detect the presence of added peel or core extracts in pineapple juice.
• Assess the usefulness of carbon isotope ratio analysis to detect added pectin as a stabiliser in pineapple juices.
• CheLab• Assess conventional analytical methods to detect the
addition of peel and core extracts to pineapple juice.• Assess the usefulness of polysaccharides as a means to
detect peel and core extracts to pineapple juices.
HPLC methods
• A number of isocratic and gradient HPLC conditions were examined to optimise the separation of the polyphenols
• Finally modified the one referenced at the bottom of the page by using acetic acid as solvent on a C18column and detection at 280 nm or DAD
• Replaced phosphate buffer (0.05M pH 3.0) with acetic acid as it is easier on piston seals. However, can not work below ca 240 nm with DAD and if required have to use phosphate buffer. In our hands both methods gave the same results.
Kirksey et al. In: Methods to detect adulteration of fruit juice beverages,Vol. 1, page 145 –166. Published by AG Science Inc, Auburndale Florida.
HPLC methodCentrifuged juice at 40k g for 10mins to clarifyDetection UV: 280 nm or DADColumn: : ODS 2, 5 um, 250 mm x 4 mmSolvent A: 0.5 % acetic acid Solvent B: 70 % acetonitrileFlow rate: 1.2 ml/minTime %A %B0 100 03 100 038 56 4440 0 10050 0 10050.1 100 070 100 0
Batch peak R.T. variability fixed with three standards
0
50
100
0
100
200m
Vo
lts
0 10 20 30 40
Minutes
Arbutin
Chlorogenic acid Phloridzin
Why use standards?
•The peaks of interest are unknowns so proper standards can not be used
•Chromatograms are very busy so little or no room for internal standard/s (best option)
•External standards help fix retention times within and between runs
•Allows detector response to be monitored with time
HPLC Profile for Pineapple Concentrate%
Mo
bile
Ph
ase
0
50
100
0
10
mV
olts
10 20 30
Minutes
Peaks of interest
HPLC Profile of Commercial Skin Extract
0
50
100
% M
ob
ile P
hase
0
10
20
30
40
mV
olts
10 20 30
Minutes
c:\gilson\rssl\system_1\polyphen.006\polyphen.gdt : UV channel 1 : skin summerpride box1: Inj. Number: 7
Aqueous
Organic
Peaks of interest
HPLC Traces for juice and skin extracts
0
50
100
% M
ob
ile P
ha
se
-20
0
20
40
mV
olts
10 20 30
Minutes
c:\gilson\rssl\system_1\polyphen.007\polyphen.gdt : UV channel 1 : skin philippine F200plant: Inj. Number: 21
c:\gilson\rssl\system_1\polyphen.007\polyphen.gdt : UV channel 1 : skin philippine f200first: Inj. Number: 23
c:\gilson\rssl\system_1\polyphen.004\polyphen.gdt : UV channel 1 : Philipine f200 first: Inj. Number: 6
c:\gilson\rssl\system_1\polyphen.004\polyphen.gdt : UV channel 1 : philippine f200 plant: Inj. Number: 8
Skin related peaks
Serial addition of skin extract
0
50
100
% M
ob
ile P
ha
se
0
10
mV
olts
10 20 30
Minutes
c:\gilson\rssl\system_1\polyphen.017\polyphen.gdt : UV channel 1 : 10535 : Inj. Number: 7
c:\gilson\rssl\system_1\polyphen.017\polyphen.gdt : UV channel 1 : 10535 5%mexico: Inj. Number: 8
c:\gilson\rssl\system_1\polyphen.017\polyphen.gdt : UV channel 1 : 10535 10%mexico: Inj. Number: 9
c:\gilson\rssl\system_1\polyphen.017\polyphen.gdt : UV channel 1 : 10535 20%mexico: Inj. Number: 10
Skin related peaks
Samples examined
2003 2004 2005 Total Total of totalsSGFNFC 8 16 4 28 63Concs 10 7 7 24 80Grade B 10 10 12Mill Juice 3 1 4 5Fruit 13 13 45
RSSLNFC 15 20 35Concs 14 20 22 56Grade B 2 2Mill Juice 1 1Fruit 2 20 10 32Totals 50 81 74 205
Typical Peak Areas for Marker Peaks in Juice, Core & Skin Extracts
Approx retention time
7 min 12 min 28 min 30 min
JuiceN = 188
13500 CV = 15%
15400 CV = 65%
21800CV = 55%
12150CV = 52%
Core ExtN = 45
8100CV = 15%
36150CV = 25%
23250CV = 17%
131200CV = 17%
Skin Ext
N = 62 158500
CV = 45%551000
CV = 30%99500
CV = 55%95000
CV = 76%
Validation exercise
• Provided a copy of the SOP to a person who had not used the method before
• Provided two training samples to familiarise them with the procedure
• a) pure juice • b) a 50/50 mix of juice plus skin extract • Examined 10 samples as blind duplicates and
asked to assess the samples on the basis of pure juice and 50% spike sample provided (samples a and b were not within the assessment set)
Validation Exercise Results
SampleBlind
duplicate 1Blind
duplicate 2Found Found
1 Pure Juice (A) pure pure2 20% skin added 22% skin 24% skin3 20% skin added 18% skin 22% skin4 30% skin added 33% skin 35% skin5 Pure Juice (B) pure pure6 30% skin added 35% skin 33% skin7 10% skin added pure pure8 10% skin added pure pure9 Pure Juice (C) pure pure
10 Pure Juice (D) pure pure
Final validation samples
• Two sets of samples received for assessment
• Set one AN 11 –15
• Showed variable profiles typical of juice samples we have examined
• Set two AN 6 –10
• All showed very similar profiles with little or no variation in components present (unusual)
HPLC traces for AN 11 - 15
min5 10 15 20253035
Norm.
01020
30
40
50
60
70
80
90
DAD1 A, Sig=280,4 Ref=360,100 (POL20115\11B.D) DAD1 A, Sig=280,4 Ref=360,100 (POL20115\12B.D) DAD1 A, Sig=280,4 Ref=360,100 (POL20115\13B.D) DAD1 A, Sig=280,4 Ref=360,100 (POL20115\14B.D) DAD1 A, Sig=280,4 Ref=360,100 (POL20115\15A.D)
1.8
18
2.0
13
2.1
32
2.3
29
2.4
39
2.6
72
3.0
66
3.3
12
3.4
01
3.7
46
4.2
93
4.9
23
8.1
48
8.8
10
9.0
48
10
.78
6
11
.85
8
12
.80
1
13
.53
7 1
3.8
86
14
.37
1
14
.95
0
15
.59
1
16
.24
6 1
6.4
35
16
.63
7 1
6.8
58
17
.10
5
17
.62
6
18
.09
9
18
.52
5 1
9.1
04
19
.33
3 1
9.5
64
19
.78
2 2
0.8
39
21
.08
2 2
1.2
71
22
.45
4 2
2.8
40
23
.45
1 2
5.8
51
26
.63
8 2
7.5
56
31
.35
1 3
2.3
19
32
.62
4 3
3.4
22
34
.41
9 4
1.7
88
42
.50
0 4
7.8
71
AN 11 –15 Marker peak II
min10 11 12 13141516
mAU
05
10
15
20
25
DAD1 A, Sig=280,4 Ref=360,100 (POL20115\11B.D)
41
.78
7 4
2.5
00
47
.87
4
DAD1 A, Sig=280,4 Ref=360,100 (POL20115\12B.D)
10
.76
5
11
.30
0
11
.92
3
12
.80
3
13
.53
6 1
3.8
74
14
.36
4 1
4.9
51
15
.38
4 1
5.5
94
16
.26
9 1
6.6
34
16
.85
5 17
.10
5 1
7.6
17
18
.09
7 1
8.5
10
19
.10
8 1
9.5
66
20
.84
1 2
1.2
69
22
.47
3 2
2.8
59
23
.46
0 2
4.8
16
25
.83
7 2
6.6
69
27
.27
0 2
7.5
25
29
.42
9 3
1.2
71
32
.27
7 3
2.6
19
33
.39
1 3
4.4
23
35
.94
8 4
1.7
91
42
.00
7 4
2.5
01
47
.87
4
DAD1 A, Sig=280,4 Ref=360,100 (POL20115\13B.D)
10
.07
6
10
.27
1
10
.72
6
11
.29
2
11
.94
6
12
.80
1
13
.53
8 1
3.8
76
14
.36
6 1
4.9
51
15
.41
9 1
5.5
84
16
.25
5 1
6.4
43
16
.63
1 1
6.8
58
17
.10
3 1
7.6
25
18
.09
4 1
8.5
23
19
.06
5 1
9.3
34
19
.56
2 1
9.8
22
20
.84
0 2
1.2
17 23
.45
0 2
5.8
27
26
.66
5 3
1.3
11
32
.24
9 3
2.6
32
41
.78
9 4
2.5
01
47
.87
5
DAD1 A, Sig=280,4 Ref=360,100 (POL20115\14B.D)
11
.80
0
12
.79
2
13
.52
8 1
3.8
69
14
.36
3 1
4.9
40
15
.57
8 1
6.2
34
16
.43
1 1
6.6
24
16
.84
6 1
7.0
98
17
.61
4 1
8.0
90
18
.51
7 1
9.0
34
19
.32
7 1
9.5
52
19
.77
4 2
0.8
32
21
.08
0 2
1.2
62
22
.42
6 2
2.8
13
23
.44
4 2
5.7
92
26
.52
9 2
7.5
00
31
.50
4 3
2.2
37
32
.63
1 3
3.3
59
33
.81
5 3
4.3
88
42
.50
1 4
7.8
76
DAD1 A, Sig=280,4 Ref=360,100 (POL20115\15A.D)
10
.78
6
11
.85
8
12
.80
1
13
.53
7 1
3.8
86
14
.37
1 1
4.9
50
15
.59
1 1
6.2
46
16
.43
5 1
6.6
37
16
.85
8 1
7.1
05
17
.62
6 1
8.0
99
18
.52
5 1
9.1
04
19
.33
3 1
9.5
64
19
.78
2 2
0.8
39
21
.08
2 2
1.2
71
22
.45
4 2
2.8
40
23
.45
1 2
5.8
51
26
.63
8 2
7.5
56
31
.35
1 3
2.3
19
32
.62
4 3
3.4
22
34
.41
9 4
1.7
88
42
.50
0 4
7.8
71
Marker Peak II
South African samples AN 6 -10
min0 5 10 15 20253035
mAU
05
10
15
20
25
DAD1 A, Sig=280,4 Ref=360,100 (POL22115\10A.D) DAD1 A, Sig=280,4 Ref=360,100 (POL22115\6A.D) DAD1 A, Sig=280,4 Ref=360,100 (POL22115\7A.D) DAD1 A, Sig=280,4 Ref=360,100 (POL22115\8A.D) DAD1 A, Sig=280,4 Ref=360,100 (POL22115\9A.D)
2.0
22 2
.139
2.3
35 2
.696
2.9
51
3.3
72 3
.760
4.3
80 4
.615
5.3
81
6.0
63 6
.367
6.8
43
7.4
28
8.2
58
9.0
44 9
.326
10.
132
10.
309
10.
611
11.
185
11.
496
11.
710
11.
889
12.
820
13.
574
13.
879
14.
346
14.
951
15.
349
15.
567
16.
197
16.
449
16.
857
17.
119
17.
390
17.
631
18.
104
18.
557
18.
838
19.
339
19.
561
20.
351
20.
859
21.
234
22.
404
23.
510
24.
794
25.
212
25.
709
26.
246
27.
150
27.
474
28.
022
29.
366
29.
772
32.
084
32.
689
33.
165
35.
089
41.
667
42.
177 42.
499
47.
822
Marker peak II region AN 6 -10
min10 10.5 11 11.5 1212.51313.5
mAU
02.5
5
7.5
10
12.5
15
17.5
20
DAD1 A, Sig=280,4 Ref=360,100 (POL22115\10A.D)
41
.663
42
.497
47
.822
DAD1 A, Sig=280,4 Ref=360,100 (POL22115\6A.D)
10
.108
10
.281
10
.630
11
.234
11
.475
11
.769
12
.805
13
.568
13
.868
14
.315
14
.919
15
.532
16
.145
16
.420
16
.828
17
.091
17
.367
17
.603 1
8.0
76 1
8.5
42 1
8.9
53 1
9.3
20 1
9.5
44 2
0.8
44 2
1.2
14 2
2.3
79 2
3.4
77 2
5.4
18 2
5.6
53 2
6.3
83 2
7.0
99 2
7.4
83 2
9.3
73 2
9.8
29 3
2.1
35 3
2.6
82 3
3.2
21 3
5.1
55 4
1.6
70 4
2.1
81 4
2.5
02 4
7.8
13
DAD1 A, Sig=280,4 Ref=360,100 (POL22115\7A.D)
10
.130
10
.303
10
.637
11
.233
11
.492
11
.764
12
.820
13
.575
13
.874
14
.330
14
.931
15
.360
15
.551
16
.173
16
.442
16
.625
16
.848
17
.110
17
.385
17
.622
18
.093
18
.550 1
8.9
31 1
9.3
26 1
9.5
49 2
0.8
50 2
1.2
23 2
2.3
91 2
3.4
82 2
4.1
07 2
4.7
75 2
5.3
46 2
5.6
79 2
6.3
57 2
7.1
21 2
7.4
82 2
8.0
24 2
9.3
76 2
9.8
12 3
1.1
34 3
2.1
20 3
2.6
87 3
3.2
00 3
5.1
30 4
1.6
69 4
2.1
79 4
2.5
00 4
7.8
15
DAD1 A, Sig=280,4 Ref=360,100 (POL22115\8A.D)
10
.295
10
.638
11
.223
11
.488
11
.754
12
.818
13
.572
13
.874
14
.336
14
.934
15
.554
16
.175
16
.443
16
.847
17
.110
17
.386
17
.617 1
8.0
86 1
8.5
42 18
.919
19
.327
19
.551
20
.853
21
.228
22
.389
23
.490
25
.356
25
.679
26
.343
27
.119
27
.470
29
.371
29
.798
32
.104
32
.678
33
.184
35
.117
41
.668
42
.180
42
.501
47
.820
DAD1 A, Sig=280,4 Ref=360,100 (POL22115\9A.D)
10
.132
10
.309
10
.611
11
.185
11
.496
11
.710
11
.889
12
.820
13
.574
13
.879
14
.346
14
.951
15
.349
15
.567
16
.197
16
.449
16
.857
17
.119
17
.390
17
.631
18
.104
18
.557
18
.838 1
9.3
39 1
9.5
61 2
0.3
51 2
0.8
59 2
1.2
34 2
2.4
04 2
3.5
10 2
4.7
94 2
5.2
12 2
5.7
09 2
6.2
46 2
7.1
50 2
7.4
74 2
8.0
22 2
9.3
66 2
9.7
72 3
2.0
84 3
2.6
89 3
3.1
65 3
5.0
89 4
1.6
67 4
2.1
77 4
2.4
99 4
7.8
22
Marker peak II
Marker peak I AN 6 - 10
0 5 10
mAU
05
10
15
20
25
DAD1 A, Sig=280,4 Ref=360,100 (POL22115\10A.D)DAD1 A, Sig=280,4 Ref=360,100 (POL22115\6A.D)DAD1 A, Sig=280,4 Ref=360,100 (POL22115\7A.D)DAD1 A, Sig=280,4 Ref=360,100 (POL22115\8A.D)DAD1 A, Sig=280,4 Ref=360,100 (POL22115\9A.D)
Marker Peak I
RSSL’s conclusions for final assessment samples AN 6 –15
CodePeak I 7 min
Peak II 12 min
Peak III 20 min
Peak IV 28 min
RSSL's Assessment Level
AN6 ++ + + + Skin added 5%AN7 +++ + + + Skin added 10%AN8 + + + + Juice 0%AN9 ++++ +++ ++ +++ Skin added 15%AN10 ++ + + + Skin added 10%AN11 + + absent absent Skin added 10%AN12 +++ +++ absent absent Skin added 20%AN13 +++ +++ absent absent Skin added 20%AN14 absent absent absent absent Juice 0%AN15 absent absent absent absent Juice 0%
Final Assessment samples
Code Origin GradeRSSL's
Assessment LevelAN6 South Africa 50 % B Skin added 5%AN7 South Africa 100 % B Skin added 10%AN8 South Africa 100 % A Juice 0%AN9 South Africa 20 % B Skin added 15%AN10 South Africa 50 % B Skin added 10%AN11 South Africa/Thailand 100 % A Skin added* 10%AN12 Philippines 100 % A Skin added* 20%AN13 South Africa 100 % B Skin added 20%AN14 Indonesia/Kenya 100 % A Juice 0%AN15 Indonesia/Kenya 100 % A Juice 0%
* Two false positive samples
Conclusions
•Levels of marker peaks in juices is variable, and some skin extracts had quite low values in third year samples
•Detection limit for detection of skin extracts can be in the region of 15 to 20% but very dependant on skin extract
•Core extracts can not be reliably seen in juice by polyphenol profiling
Masking Skin Extract Addition
• Cloud stability of pineapple juices is generally poor and stabilisers are sometimes added to limit cloud loss
• Addition of skin extracts, which sometimes have little cloud, can add to these problems
• Both stability problems can be overcome by the addition of a thickening agent, such as pectin
• Most commercial pectins are either derived from citrus fruits, typically lemon, or apple as they are by-products of juice processing
Carbon isotope ratios in whole juice
isolated pectins and stabilisers
Isolated Citrus Apple Whole Juice Pectin pectin pectin
Number 50 50 10 5
Mean (permil) -12.55 -12.8 -25.8 -25.7
Std Div -0.97 -1.15 -1.56 -1.48
CV (%) 7.7 9.0 6.0 5.8
Addition of C3 Derived Pectin
Whole Juice Un-spiked Spiked Pectin level Spike levelPectin Pectin
-11.80 -12 210 0-11.75 -12.7 5-11.78 -13.4 10-11.90 -14.8 20
-12.30 -12.3 295 0-12.40 -13.0 5-12.38 -13.7 10-12.35 -15.0 20
Pectin Addition in Commercial Juices?
Origin Commerical Whole juice Pectin ratio Difference ConclusionsJuice type (permil) (permil) (permil)
Thialand conc -11.8 -12.3 -0.5 OKIndonesia conc -12.3 -12.9 -0.6 OKThialand conc -11.9 -11.8 0.1 OK
Malayasia conc -11.2 -11.5 -0.3 OKKenya conc -12.4 -12.4 0 OK
UK conc -11.2 -11.6 -0.4 OKCosta rica conc -11.6 -12 -0.4 OK
Spain Finished prod -11.5 -11.4 0.1 OKSpain Finished prod -12.3 -12.8 -0.5 OK
Azores NFC -13.5 -13.6 -0.1 OKCosta rica NFC -11.5 -11.8 -0.3 OKCosta rica NFC -12.7 -12.4 0.3 OKCosta rica NFC -12.9 -13.5 -0.6 OK
Spain Finished prod -11.6 -13 -1.4 Added pect.Spain Finished prod -11.4 -13.3 -1.9 Added pect.Poland Finished prod -15.6 -11.8 3.8 Added C3 sug
Conclusions
• Carbon isotope ratio shows addition of C3
stabilisers to pineapple juices at 10 to 20% level
• Can detection sugar addition if added from C3
source by shift in whole juice vs. pectin values
• Note previous work by RSSL as part of FSA funded project found a weaker correlation between sugars and pectin than between individual sugars and citric acid.
Thanks
My co-workers
Joaquim Magalhaes
Jonathan Nodjiadjim
EU Commission for funding this project