You are on page 1of 15

678 WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO.

3, 2006
FOOD CHEMICAL CONTAMINANTS

Liquid Chromatographic Analysis of Aflatoxin Using


1
Post-Column Photochemical Derivatization: Collaborative Study
ARTHUR E. WALTKING
Waltking Associates, 482 Rock Rd, Glen Rock, NJ 07452
DAVID WILSON
University of Georgia, Coastal Plain Experiment Station, Tifton, GA 31793
Collaborators: D. Chan; E. Dunn; J. Humphries; H. Kandler; E. Sizoo; D. Wilson

Aflatoxin analysis, with post-column derivatization


using a photochemical reactor for enhanced
detection (PHRED) system for derivatization, has
been compared to the officially recognized iodine
and Kobra cell derivatization systems. This
photochemical system has been extensively used
for screening peanuts by some U.S. Department of
Agriculture laboratories for many years. From their
periodic method checks, using standard spiked
samples, an 80 sample series with each of the
3 derivatization methods was statistically analyzed.
Paired comparisons, using the same sample
extract, were also made between the PHRED and
one of the other 2 methods, among laboratories in
4 different countries, on a variety of naturally
contaminated commodity products. The
differences between the techniques were not
significant for peanuts, but for corn the
photochemical system consistently gave slightly
higher values for aflatoxins B1 and B2 than the
Kobra cell method. However, a comparison of all
sample results showed no significant differences
between methods. The Pearson correlation
coefficients for aflatoxin B1 in 102 test samples and
aflatoxin B2 in 94 test samples were 0.9994 and
0.9874, respectively. The probability factor was
P < 0.0001, and the t-tests were not significantly
different except for the corn. These indicated that
the PHRED system is equivalent to the iodine and
Kobra cell methods for peanuts relative to the
current official procedures, but the PHRED system
has a slightly high bias for corn compared to the
iodine and Kobra cell systems.
Submitted for publication March 2006.
The recommendation was approved by the Methods Committee on
Natural Toxins and Allergens as First Action. See Official Methods
Program Actions, (2006) Inside Laboratory Management,
January/February issue.
Corresponding authors e-mail: arthur.waltking@verizon.net
1
The study is a change of method proposal for AOAC Methods 991.31
and 999.07 and the modification (from TLC to LC) for AOAC Method
970.45.

hen aflatoxins were found to adversely affect the


health of various animal species in the 1960s,
attempts were made to provide fast, inexpensive,
and comprehensive detection methods. Two methods
developed by researchers of the U.S. Food and Drug
Administration, requiring up to 5 h for 4 samples (1, 2), were
the first significant improvements from the almost 24 h
required with the initial method (3). The development of the
Best Foods (BF) method permitted the analysis of 4 samples
to be completed within 1 h (4, 5). These methods, in the 1960s,
provided the basic tools for isolating aflatoxin contaminants
from foods and commodities.
The principal emphasis in the 1970s was on improving
sampling techniques and statistical evaluations of the
data (68). Studies also continued on many other aspects of
the problem and improvements were made in the
identification of the biological conversion to additional
metabolites, the M1 (milk toxins; 9) and in studies of the
development and degradation of the aflatoxins during storage
of the host commodity or during commercial food
preparations (1012). A major advance was the development
of a confirmation of identity method by Przybylski (13),
which readily proved the presence of the aflatoxins B1 and G1
by forming derivatives that have different transport positions
on the thin-layer plates. Also, the analyses became more
objective in form as the subjective evaluation of the thin-layer
plates was automated by densitometers or supplanted by
liquid chromatography (LC; 1416).
The advances in the use of LC and various enhancement
techniques were well documented in a review article in
1992 (17). Many of the currently used LC methods use
derivative formation (1823) and therefore result from the
original chemical confirmation test (13). In 1993, aflatoxin
analysis using a photochemical derivatization technique was
introduced by Joshua (24). His paper critiques the merits of
the 3 chemical approaches, using trifluoroacetic acid (TFA),
iodine, or bromine, relative to the photochemical technique.
Because the photochemical reactor for enhanced detection
(PHRED) system provides the derivative formation using
only UV light, the ability to obtain the same results with a

WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006 679

simple sample preparation and less chemical waste is a


distinct advantage in the present day laboratory.
An important improvement in LC detection was reported
recently by Joshua (25), who demonstrated near baseline
separation of all 6 aflatoxins (B1, B2, G1, G2, M1, and M2), as well
as zearalenone and ochratoxin A on a single chromatogram when
using post-column photochemical derivatization and
fluorescence detection. The possibility of providing a single
detector condition, which could offer a universal quantitative
procedure for sample extracts for all 8 of these contaminants
independent of the extraction method used, prompted this study.
Collaborative Study
Purpose
The purpose of the study is to determine the equivalency of
the performance characteristics for an LC photochemical
derivatization technique (PHRED) to that of other previously
AOAC-approved post-column derivatization systems,
iodine (20), and Kobra cell (26, 27), thereby recognizing the
reduced cost and simplicity advantages.
As a first step, a new technique must accurately measure
aflatoxin with both corn and peanut crops, which are among
the most highly regulated in the United States and Europe.
However, this study was considered to be unnecessarily
complex if it were to involve multiple commodities, using
multiple aflatoxins with a number of official methods.
Although for many collaborative studies, a precollaborative
study would be performed to determine the efficacy or
ruggedness of the method, in this case such studies have
already been done in part by at least 2 laboratories (28, 29). In
these cases, the data were reported to provide equivalent
results for the photochemical method with the iodine and the
Kobra cell methods. The data from the first of these
references (28), which compared the PHRED to the iodine
method (20), are included with the authors permission in this
data set. The iodine method was also used as one of the
reference methods in the AOAC collaborative study, which
compared 9 aflatoxin methods (30). The other reference (29)
showed equivalency of the Kobra cell to the PHRED.
Performance
In ref. 29, where the relative standard deviation (RSD)
values of the validation testing ranged from 0.3 to 1.8% for the
PHRED and 0.9 to 2.0% for the Kobra cell, the equivalency of
these systems is indicated.
Principle
The PHRED is a compact unit placed between the LC
column and the detector in the same manner as the Kobra
cell (31) to perform online post-column derivatization of
aflatoxins to increase detectability and/or selectivity of
response for the detector. Unlike the iodine and Kobra cell, it
performs the derivatization photochemically without
additional chemicals added to the mobile phase. The method
for the PHRED is therefore an adjunct method. It is used only
for the enhancement of aflatoxins B1 and G1 in the extract of

any of the official preparation methods using LC separation


and post-column derivatization. Instead of a full collaborative
study, it was proposed and accepted that a study of the
PHRED consisting of a direct comparison of the same extracts
with the Kobra cell or iodine systems would be appropriate for
statistical analysis. This study used the procedure described in
the Method section with the PHRED system.
Quality AssuranceControl
Three European laboratories actively involved with
aflatoxin analysis were asked to use the PHRED equipment on
a minimum of 4 sample extracts of a naturally contaminated
material which was prepared and analyzed by their current
method. They were provided a mixed aflatoxin standard
prepared from a single supply, so that differences in the
standard used would be negated. Their submission of the
comparative results and photocopies of the chromatograms of
the samples and standards were used to certify their capability
prior to delivery of a collaborative sample extract series.
Subsequently, with the change from a standard collaborative
to an equivalency review, these data have been included with
the extensive data available from the United States.
Participating Laboratories
Six laboratories participated in the study: RIVM (Bilthoven,
The Netherlands); Kantonales Labor (Zurich, Switzerland);
Central Science Laboratory (York, United Kingdom);
University of Georgia (Tifton, GA); U.S. Department of
Agriculture (USDA; Blakely, GA); USDA (Madill, OK).
Study Design
The USDA laboratory in Blakely, GA, has used the
PHRED photochemical system as well as the iodine method
for many years for screening peanuts while the laboratory in
Madill, OK, used the Kobra cell. From their periodic method
checks using peanut samples spiked with aflatoxin standards,
the last 80 samples, with each of the 3 derivatization methods,
were tabulated and analyzed as shown in Tables 14. All of
their extracts were prepared using the extractions in AOAC
Method 970.45. Segments of sample slurries which proved to
be free of contamination were spiked with a mixed aflatoxin
standard and carried through as confirmation of proper
analysis.
Earlier USDA results of 19 naturally contaminated peanuts
from ref. 27, also using extracts from AOAC Method 970.45,
are shown in Table 5. The extracts in this study compared the
PHRED and the iodine system. An independent series of
naturally contaminated corn and peanuts, 37 and 30 samples,
respectively, were analyzed by PHRED and the Kobra cell
system at the University of Georgia. These extracts, as shown
in Table 6, were prepared using AOAC Method 991.31.
Adding to these the 16 samples from the European checks
of the PHRED equipment operation (Table 7), provided a total
of 102 samples of naturally contaminated extracts which were
statistically analyzed, independently and combined. The
European laboratories reported using the following
referenced methods for sample extract preparation: United

680 WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006
Table 1. Peanuts and pistachio nuts spiked with 12.0 ng/g aflatoxin standard using AOAC Official Method 970.45 for
extraction and iodine system for determinations performed by USDA (Blakely, GA)a
Aflatoxin, ng/g
Sample No.

B1

B2

G1

G2

Total

4.0

1.2

5.4

1.0

11.6

6.9

0.9

2.8

0.3

10.9

5.8

1.6

3.9

0.6

11.9

4.0

1.2

4.2

1.3

10.7

5.2

1.5

4.3

0.9

11.9

4.9

1.3

4.2

0.4

10.8

4.7

1.2

4.0

1.1

11.0

3.9

1.3

6.3

0.8

12.3

4.6

1.6

4.7

1.6

12.5

10

5.9

1.3

4.8

0.3

12.3

11

4.5

1.4

5.0

1.2

12.1

12

3.7

1.1

4.0

1.0

9.8

13

3.5

1.4

4.1

0.8

9.8

14

4.0

1.3

4.0

0.4

9.7

15

3.9

1.2

3.8

1.2

10.1

16

4.0

1.1

5.0

0.4

10.5

17

4.2

1.4

4.6

1.5

11.7

18

4.0

1.0

4.4

0.6

10.0

19

3.2

1.0

4.7

0.9

9.8

20

4.4

1.0

4.4

0.5

10.3

21

6.0

0.9

5.6

0.0

12.5

22

4.3

1.3

5.6

1.5

12.7

23

4.2

1.2

5.3

1.2

11.9

24

3.7

1.3

5.4

1.1

11.5

25

4.4

1.3

3.2

1.1

10.0

26

4.6

1.6

4.3

0.6

11.1

27

4.5

1.5

3.7

1.1

10.8

28

5.2

1.3

3.5

1.2

11.2

29

3.9

1.2

3.8

0.9

9.8

30

3.9

1.2

3.9

0.9

9.9

31

4.0

1.3

4.0

0.9

10.2

32

4.1

1.3

3.7

0.9

10.0

33

4.1

1.3

4.0

0.9

10.3

34

3.9

1.3

4.1

0.8

10.1

35

4.1

1.3

4.0

0.8

10.2

36

4.2

1.3

3.9

0.8

10.2

37

4.2

1.3

3.9

0.9

10.3

38

4.3

1.4

4.3

0.9

10.9

39

4.5

1.5

4.3

1.0

11.3

40

2.3

1.0

5.8

1.8

10.9

41

3.9

1.1

5.9

1.0

11.9

42

3.8

1.0

5.2

0.8

10.8

43

4.5

1.5

4.1

1.4

11.5

44

4.4

1.6

3.2

0.9

10.1

WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006 681
Table 1. (continued)
Aflatoxin, ng/g
Sample No.

B1

B2

G1

G2

Total

45

4.7

1.5

4.2

0.6

11.0

46

4.7

1.7

3.9

0.6

10.9

47

5.0

1.6

5.2

1.4

13.2

48

3.7

1.5

4.0

1.2

10.4

49

4.9

1.4

3.7

0.6

10.6

50

4.4

1.3

4.6

0.5

10.8

51

4.2

1.3

4.6

0.9

11.0

52

4.0

1.2

4.6

0.8

10.6

53

3.7

1.0

4.4

1.0

10.1

54

4.0

1.1

4.8

0.9

10.8

55

4.4

1.2

5.4

1.3

12.3

56

4.0

1.1

4.8

0.8

10.7

57

3.8

1.1

4.6

1.2

10.7

58

4.5

1.2

5.4

1.2

12.3

59

4.6

1.3

5.5

1.2

12.6

60

3.9

1.1

4.9

1.0

10.9

61

3.9

1.2

5.2

1.0

11.3

62

3.9

1.0

4.7

1.1

10.7

63

3.9

1.2

5.0

1.0

11.1

64

3.7

0.9

4.3

1.1

10.0

65

3.8

1.0

4.6

1.1

10.5

66

4.4

1.3

5.5

1.3

12.5

67

3.6

0.9

4.4

1.0

9.9

68

5.1

1.3

5.5

1.2

13.1

69

4.1

1.0

5.5

0.9

11.5

70

3.6

1.1

5.0

1.0

10.7

71

3.4

1.0

4.5

1.2

10.1

72

3.8

0.9

4.6

0.8

10.1

73

4.1

1.1

5.2

0.9

11.3

74

4.2

1.1

5.4

0.9

11.6

75

4.4

1.1

5.5

1.1

12.1

76

3.6

0.9

4.7

0.8

10.0

77

3.9

1.1

5.5

1.1

11.6

78

3.9

1.0

5.2

0.9

11.0

79

4.4

1.1

5.9

0.9

12.3

80

4.0

1.0

5.4

0.9

11.3

All samples are peanuts except Nos. 20, 32, 44, 48, and 69 which were pistachio. Standard concentrations (ng/g) used for spiking all
samples: B1 at 4.6; B2 at 1.4; G1 at 4.6; G2 at 1.4.

682 WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006
Table 2. Peanuts spiked with 12.0 ng/g aflatoxin standard using AOAC Official Method 970.45 for extraction and
PHRED system for determinations at USDA (Blakely, GA)a
Aflatoxin, ng/g
Sample No.

B1

B2

G1

G2

Total

4.5

1.4

5.0

1.3

12.2

4.1

1.6

3.8

0.8

10.3

3.2

2.0

5.0

0.6

10.8

4.4

1.1

5.5

0.9

11.9

4.1

1.1

4.7

0.8

10.7

3.2

1.2

5.0

0.6

10.0

4.3

1.6

5.0

0.9

11.8

4.5

1.4

4.7

0.7

11.3

5.2

1.2

4.6

0.9

11.9

10

4.7

1.2

4.4

0.4

10.7

11

5.3

1.0

6.1

0.5

12.9

12

3.6

1.0

4.4

1.2

10.2

13

7.8

1.2

4.0

0.2

13.2

14

5.1

0.9

5.5

0.9

12.4

15

6.1

1.1

4.7

0.9

12.8

16

3.4

1.2

4.4

0.8

9.8

17

5.3

1.2

5.6

1.1

13.2

18

4.3

1.2

5.3

1.0

11.8

19

4.1

1.3

6.0

0.9

12.3

20

4.1

1.3

5.2

1.6

12.2

21

4.1

1.0

4.4

0.8

10.3

22

5.2

1.4

5.6

0.8

13.0

23

4.0

1.2

5.4

0.8

11.4

24

5.2

1.0

5.3

0.8

12.3

25

4.9

1.3

5.8

1.2

13.2

26

3.6

1.1

4.7

0.9

10.3

27

4.0

1.2

4.4

0.7

10.3

28

4.9

1.3

5.0

1.6

12.8

29

4.2

1.2

3.8

0.7

9.9

30

3.7

0.8

4.6

1.0

10.1

31

4.2

1.1

4.7

1.0

11.0

32

4.2

1.0

5.3

0.9

11.4

33

4.1

1.2

5.2

0.8

11.3

34

3.7

1.2

4.4

0.9

10.2

35

4.1

1.1

3.9

0.7

9.8

36

4.7

1.2

4.3

0.9

11.1

37

4.9

1.9

4.4

0.9

12.1

38

4.4

1.2

4.5

1.2

11.3

39

4.1

1.2

4.2

0.8

10.3

40

4.5

1.3

4.4

1.0

11.2

41

4.5

1.3

4.5

1.4

11.7

42

4.4

1.3

4.6

1.2

11.5

43

4.8

1.5

5.3

1.4

13.0

44

3.7

1.1

4.2

0.8

9.8

WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006 683
Table 2. (continued)
Aflatoxin, ng/g
Sample No.

B1

B2

G1

G2

45

3.7

1.1

4.2

0.8

9.8

46

4.6

1.1

4.5

0.8

11.0

47

3.7

1.3

4.4

0.7

10.1

48

3.8

1.4

4.5

1.3

11.0

49

5.1

1.2

5.2

1.2

12.7

50

5.1

0.7

4.8

1.2

11.8

51

5.4

0.6

4.4

0.5

10.9

52

3.9

1.3

4.5

1.0

10.7

53

4.0

1.2

3.8

1.1

10.1

54

3.8

1.2

4.4

0.8

10.2

55

4.1

1.3

4.9

0.7

11.0

56

4.3

1.5

5.4

1.3

12.5

57

3.7

1.1

4.7

1.0

10.5

58

4.3

1.5

5.4

1.3

12.5

59

4.6

1.5

4.5

0.9

11.5

60

4.6

1.1

5.1

0.9

11.7

61

4.4

1.2

5.1

1.4

12.1

62

4.1

1.1

4.3

0.9

10.4

63

4.1

1.1

4.5

1.4

11.1

64

4.0

1.1

5.3

1.5

11.9

65

5.1

1.3

5.7

0.9

13.0

66

3.6

0.9

4.5

0.6

9.6

67

3.7

1.0

4.3

0.7

68

4.5

1.2

5.3

1.0

12

69

4.5

1.4

5.2

1.5

12.6

70

4.4

1.0

4.5

0.6

10.5

71

3.7

1.0

5.8

1.6

12.1

72

3.6

1.0

4.6

1.0

10.2

73

3.7

1.1

4.9

0.6

10.3

74

3.5

1.0

4.5

0.8

9.8

75

3.8

1.0

4.5

0.9

10.2

76

3.3

0.9

4.7

0.9

9.8

77

4.5

1.3

6.0

0.4

12.2

78

4.3

0.9

5.0

0.7

10.9

79

5.0

1.1

5.3

0.4

11.8

80

5.0

1.2

4.6

0.9

11.7

Standard concentrations (ng/g) used for spiking all samples: B1 at 4.6; B2 at 1.4; G1 at 4.6; G2 at 1.4.

Total

9.7

684 WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006
Table 3. Peanuts spiked with 12.0 ng/g aflatoxin standard using AOAC Official Method 970.45 for extraction and
Kobra cell system for determinations at USDA (Madill, OK)a
Aflatoxin, ng/g
Sample No.

B1

B2

G1

G2

Total

4.2

1.3

4.2

0.9

10.6

3.8

1.2

3.9

0.9

9.8

4.1

1.3

4.4

1.0

10.8

3.8

1.1

4.3

0.8

10

3.7

1.1

4.4

1.0

10.2

4.0

1.2

4.6

0.8

10.6

4.1

1.3

4.8

0.7

10.9

5.1

1.5

5.5

0.8

12.9

4.4

1.4

5.0

1.1

11.9

10

3.8

1.2

4.3

0.9

10.2

11

4.3

1.3

4.8

0.9

11.3

12

4.6

1.5

5.2

1.1

12.4

13

4.2

1.3

4.7

0.9

11.1

14

3.9

1.2

4.3

1.0

10.4

15

4.6

1.4

5.0

1.3

12.3

16

4.1

1.3

4.6

1.2

11.2

17

4.1

1.3

4.5

1.0

10.9

18

4.4

1.4

4.9

0.9

11.6

19

4.7

1.3

4.6

0.9

11.5

20

4.0

1.2

4.4

0.8

10.4

21

4.2

1.2

4.7

0.7

10.8

22

4.3

1.3

4.7

0.8

11.1

23

4.8

1.5

5.3

1.0

12.6

24

4.2

1.3

4.6

0.8

10.9

25

3.8

1.2

4.1

0.8

9.9

26

3.9

1.2

4.2

0.8

10.1

27

4.7

1.4

4.8

0.6

11.5

28

5.5

1.2

4.1

0.5

11.3

29

4.5

1.4

4.7

0.7

11.3

30

4.0

1.3

4.4

1.0

10.7

31

3.9

1.2

4.3

0.9

10.3

32

4.2

1.3

4.5

0.9

10.9

33

4.5

1.4

4.9

0.9

11.7

34

4.9

1.4

4.3

0.9

11.5

35

4.1

1.3

4.4

0.8

10.6

36

4.3

1.4

4.8

0.8

11.3

37

4.5

2.2

4.5

0.7

11.9

38

3.9

1.2

4.1

0.6

9.8

39

4.7

1.5

5.1

0.8

12.1

40

4.1

1.3

4.6

0.8

10.8

41

4.7

1.5

5.2

0.7

12.1

42

4.3

1.4

5.0

1.0

11.7

43

3.8

1.2

4.5

0.9

10.4

44

4.2

1.4

4.8

0.9

11.3

WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006 685
Table 3. (continued)
Aflatoxin, ng/g
Sample No.

B1

B2

G1

G2

Total

45

4.1

1.3

4.5

0.8

10.7

46

4.3

1.4

4.9

0.9

11.5

47

4.6

1.5

5.1

1.0

12.2

48

5.4

1.6

5.3

0.6

12.9

49

5.1

1.6

5.3

0.7

12.7

50

5.3

1.5

5.0

1.1

12.9

51

4.7

1.5

5.0

1.0

12.2

52

4.6

1.4

4.8

0.9

11.7

53

4.5

1.4

4.7

0.8

11.4

54

4.4

1.3

4.7

0.8

11.2

55

4.2

1.4

4.6

0.7

10.9

56

4.5

1.4

4.9

0.6

11.4

57

4.6

1.4

5.0

0.8

11.8

58

4.2

1.4

4.7

1.1

11.4

59

4.3

1.3

4.5

0.8

10.9

60

4.3

1.3

4.5

0.7

10.8

61

5.1

1.4

4.6

0.7

11.8

62

4.2

1.3

4.6

0.7

10.8

63

4.6

1.5

5.1

1.3

12.5

64

4.2

1.3

4.6

1.0

11.1

65

4.9

1.4

4.5

0.9

11.7

66

3.7

1.2

4.4

0.8

10.1

67

3.9

1.2

4.6

0.8

10.5

68

3.9

1.2

4.5

0.7

10.3

69

4.4

1.4

4.9

0.6

11.3

70

5.0

1.5

5.0

0.7

12.2

71

4.5

1.5

5.2

0.8

12.0

72

4.7

1.1

4.1

0.4

10.3

73

3.6

1.3

5.2

0.8

10.9

74

4.7

1.5

5.3

0.7

12.2

75

4.5

1.2

4.4

0.6

10.7

76

4.1

1.2

4.4

0.8

10.5

77

4.1

1.2

4.4

0.7

10.4

78

4.4

1.3

4.7

0.9

11.3

79

4.2

1.3

4.7

0.8

11.0

80

4.2

1.3

4.8

0.8

11.1

Standard concentrations (ng/g) used for spiking all samples: B1 at 4.6; B2 at 1.4; G1 at 4.6; G2 at 1.4.

686 WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006
Table 4. Summary comparison of the 3 systems (iodine, PHRED, and Kobra cell) using AOAC Method 970.45
extractiona
Technique

Mean, ng/g

SD, ng/g

RSDr, %

PRSDR, %

HorRat

Aflatoxin B1
Iodine

0.64

15.13

36.06

0.42

4.23

Kobra

4.35

0.41

9.43

35.91

0.26

PHRED

4.34

0.70

16.13

35.93

0.45

Aflatoxin B2
Iodine

1.22

0.22

16.39

43.46

0.38

Kobra

1.34

0.15

11.19

42.85

0.26

PHRED

1.19

0.22

18.49

43.62

0.42

Aflatoxin G1
Iodine

4.62

0.72

15.58

35.59

0.44

Kobra

4.68

0.34

7.26

35.52

0.20

PHRED

4.81

0.54

11.23

35.38

0.32

Aflatoxin G2
Iodine

0.95

0.31

32.63

45.12

0.72

Kobra

0.84

0.16

19.05

45.96

0.41

PHRED

0.93

0.30

32.26

45.26

0.71

Total aflatoxin
Iodine

11.02

0.89

8.08

31.24

0.26

Kobra

11.21

0.78

6.96

31.16

0.22

PHRED

11.27

1.05

9.32

31.13

0.30

Tables 13 samples spiked with 12.0 ng/g aflatoxin standard (n = 80). For each set in Tables 13, the following assessments of the data
were obtained: mean; standard deviation (SD); percent relative standard deviation for repeatability (RSDr, %); predicted relative
reproducibility as calculated from the Horwitz equation (PRSDR, %) and the HorRat based on intralaboratory repeatability (RSDr).

Kingdom (32); Switzerland (26); The Netherlands: Set 1 (26),


Set 2 (33).
Statistical analyses of all the natural contamination series
were performed only for the B1 and B2 aflatoxins because of
absence of G aflatoxins in so many of the products. A
univariate statistical analysis of the spiked samples showed
similarity of performance of the aflatoxin G1 series to the
aflatoxin B1 results.
AOAC Official Method 2005.08
Analysis of Aflatoxin
Liquid Chromatography with
Post-Column Photochemical Derivatization
First Action 2005

[The photochemical reactor for enhanced detection


(PHRED) is applicable to determination of aflatoxins in test
extracts of corn and peanuts when using AOAC Methods
991.31, 999.07, or 970.45 with LC. Although no significant
difference exists for the PHRED in comparison to other
post-column methods with peanuts, a slightly high bias is
obtained with corn when compared with the iodine or Kobra
cell possibly due to higher recovery.]

Caution: Mycotoxins are toxic substances. Perform


manipulations in a hood wherever possible, taking particular
precautions, such as using a glove box, when toxins are in dry
form because of their electrostatic nature and tendency to
disperse. Swab any accidental spills and all glassware and
waste materials with 5% NaOCl bleach. Use UV glasses if
there is exposure to any direct or reflected UV light from the
light source.
Aflatoxins must be handled with extreme caution because
they are known to be carcinogens. Use hypochlorite bleach for
cleaning glassware and when disposing of waste materials.
A. Principle
Post-column derivatization of aflatoxins can increase
detectability and/or selectivity of responses for the HPLC
detector. By performing the derivatization photochemically,
the derivative structures B2a and G2a are apparently obtained,
providing the enhanced signals for the B1 and G1 aflatoxins
without effect on the B2 and G2 aflatoxins.
B. Apparatus
Note: Evaluate any leakage of UV light from source
equipment and if detected, provide shielding or protective

WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006 687
Table 5. Comparison of PHRED and iodine systems with naturally contaminated peanuts using AOAC Official
Method 970.45 extraction
PHRED

Iodine
Aflatoxin, ng/g

B2

G1

G2

Total

B1

B2

G1

G2

Total

2.44

0.57

NDa

ND

3.01

2.36

0.60

ND

ND

2.96

31.53

4.91

0.42

0.21

37.07

30.62

5.06

ND

ND

35.68

42.22

6.52

ND

0.10

48.84

38.90

6.97

ND

ND

45.87

20.68

3.12

ND

0.15

23.96

19.36

3.32

ND

ND

22.68

37.22

5.79

ND

0.10

43.12

35.63

6.33

ND

ND

41.96

B1

42.53

6.14

ND

0.13

48.81

50.69

ND

ND

ND

50.69

42.69

6.57

ND

0.12

49.38

45.62

6.79

ND

ND

52.41

33.13

4.96

ND

ND

38.09

32.88

5.05

ND

ND

37.93

25.40

2.99

ND

0.13

25.52

22.74

3.33

ND

ND

26.07

16.38

2.47

ND

ND

18.86

15.03

2.58

ND

ND

17.61

2.74

0.42

ND

ND

3.16

2.60

0.43

ND

ND

3.03

33.53

4.91

ND

ND

38.44

35.77

5.86

ND

ND

41.63

0.67

0.04

ND

ND

0.72

1.03

0.21

ND

ND

1.24

2.44

0.51

ND

ND

2.95

2.69

0.66

ND

ND

3.35

2.80

1.85

0.06

0.11

4.82

2.49

2.11

ND

ND

4.60

2.33

1.27

0.52

0.19

4.31

2.57

0.98

0.62

ND

4.17

2.72

1.97

1.91

1.34

7.94

3.09

1.60

2.22

0.87

7.78

4.41

2.43

1.22

0.43

8.49

4.82

2.21

1.42

0.39

8.84

3.41

1.23

0.69

ND

5.33

3.13

1.32

0.76

0.26

5.48

ND = None detected.

glasses during use. To prevent leakage of the knitted reactor


coil, do not overtighten the connection. If leakage occurs,
disconnect power to the photochemical reactor before
inspecting the unit.
Equipment noted is not restrictive; equivalent systems can
be substituted.
(a) LC system.SP 8700 XR pump, SP 4200 computing
integrator, SP 8780 autosampler, and SP WINNER software
(Spectra-Physics Analytical, San Jose, CA) with Kratos FS
970 LC fluorometer set to provide 365 nm excitation and
435 nm emission.
(b) Column.Beckman Ultrasphere C18, 150 4.6 mm
with 5 mm particle size (No. 235330; Alltech Associates, Inc.,
Deerfield, IL; www.alltechweb.com/US/Home.asp).
(c) PHRED photochemical reactor.With low-pressure
mercury lamp and knitted reactor coils, preferably KRC 25-25
with a 25 m 0.25 mm id coil (AURA Industries, New York, NY;
www.aura-inc.com).
(d) Silanized vials.Four mL, amber with Teflon-lined
screw caps (No. 72680; Alltech).
(e) Pipet.Class A, volumetric, 2 mL.
(f) Replacement plastic grooved ferrules.No. ZGF1PK-10
(Valco
Instrument
Co.
Inc.,
Houston,
TX;
www.vici.com/profiles/prof_val.php).

C. Reagents
(a) Degassed mobile phase.40% Methanol in 60%
water (v/v), or a suitable mixture of methanol, acetonitrile,
and water that results in baseline separation of the aflatoxins.
Note: Unless otherwise specified, use only analytical grade
reagents. One of the solvent systems indicated for the
photochemical system permits the elimination of acetonitrile.
This can be advantageous for some laboratories. Any
combination of methanol, water, and acetonitrile can be used
so long as the mobile phase provides baseline separation of the
aflatoxins.
(b) Injection solvent.Same as that used in mobile phase.
(c) Aflatoxin standards.Supelco Inc. (Bellefonte, PA;
www.sigma-aldrich.com), or other suppliers.
D. Fluorescence Detector Conditions
Use detector parameters which are applicable for the
available equipment in accordance with the manufacturers
recommendations. (Conditions found optimum in one
laboratory for fluorescence detector: Excitation, 365 nm;
emission, 435 nm; filter mode, resistor-capacitor circuit (RC)
response setting of slow; digital filter, 3 or 5 s; gain, 10 or
above; attenuation, 1 or as needed.)

688 WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006
Table 6. Comparison of PHRED and Kobra cell analysis of naturally contaminated peanuts and corn using AOAC
Official Method 991.31 for extraction
Peanuts

Corn

PHRED

Kobra cell

PHRED

Kobra cell

Aflatoxin, ng/g
B2

B1
a

ND

3.68

B1

B2

B1

B2

B1

B2

ND

ND

ND

0.50

ND

0.50

ND

1.79

3.88

2.01

1.94

ND

1.67

ND

25.80

4.00

24.64

3.30

ND

ND

ND

ND

507.09

40.00

528.23

40.0

0.50

ND

0.44

ND

16.25

2.18

15.42

1.95

ND

ND

ND

ND

64.13

9.03

56.93

8.09

23.87

2.00

23.65

1.90

18.75

2.59

16.59

2.41

5.92

0.25

5.52

0.25

3.25

ND

3.85

ND

14.46

0.75

14.13

0.72

3.52

0.50

3.95

0.50

4.72

1.00

4.26

0.99

10.37

1.50

9.70

1.89

ND

ND

ND

ND

5.83

2.76

5.31

3.00

30.19

1.54

29.28

1.30

1.61

0.25

1.51

0.25

1.68

0.25

1.94

0.25

0.55

0.25

0.92

0.25

ND

ND

ND

ND

1.84

ND

1.73

ND

ND

ND

ND

ND

6.58

0.50

5.49

0.25

4.65

0.25

4.44

0.25

9.04

0.79

8.25

1.53

19.54

0.74

18.23

0.67

14.71

2.97

12.90

2.48

10.98

0.76

10.56

0.68

19.91

3.95

17.80

3.36

28.33

1.29

27.60

1.23

1.47

ND

1.04

0.25

14.71

0.53

14.98

0.49

2.06

0.50

2.47

0.50

12.94

0.48

11.56

0.43

8.87

1.50

8.73

1.83

26.97

2.28

26.41

2.00

0.70

ND

0.45

ND

22.80

0.84

22.49

0.81

23.85

2.95

21.43

2.26

2.51

0.25

2.50

0.25

14.81

0.68

13.86

1.63

61.33

2.58

60.99

2.52

16.34

0.77

14.23

1.40

33.27

3.20

32.22

2.90

1.23

ND

1.89

ND

16.03

1.00

15.52

0.96

0.43

ND

0.55

ND

ND

ND

ND

ND

105.66

11.28

98.76

9.98

1.68

ND

2.16

ND

13.12

2.59

12.27

3.00

ND

ND

ND

ND

13.19

2.72

11.08

2.30

ND

ND

ND

ND

ND = None detected.

2.71

ND

2.54

ND

ND

ND

ND

ND

1.00

ND

ND

ND

0.50

ND

0.50

ND

0.30

ND

0.50

ND

12.23

0.25

11.98

0.25

22.32

1.25

21.51

1.19

WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006 689
Table 7. Analyses performed by European laboratories while checking PHRED equipment using a variety of
extraction methods
Photochemical analysis using extract from Reference Method

Standard analysis using Reference Method

Aflatoxin, ng/g
Sample

B1

B2

G1

G2

Total

B1

B2

G1

G2

Total

Switzerland (Method A; ref. 26)


1 Corn

NDa

ND

ND

ND

ND

ND

ND

ND

ND

ND

2 Peanuts

4.6

0.57

ND

ND

5.2

4.49

0.52

ND

ND

5.0

3 Corn

7.23

0.86

1.23

0.2

9.5

7.23

0.82

0.97

0.13

9.2

4 Peanut butter

1.59

0.31

0.41

0.21

2.5

1.59

0.46

0.68

0.14

2.9

United Kingdom (Method B; ref. 32)


1 Figs

26.0

2.5

12.6

1.2

42.3

28.1

2.3

14.0

1.2

45.6

2 Figs

23.3

2.2

12.8

1.0

39.3

25.5

2.1

16.1

1.1

44.6

3 Figs

31.1

2.6

14.5

1.0

49.2

33.4

2.5

15.2

1.0

52.1

4 Figs

26.5

2.4

15.5

1.1

45.5

28.8

2.3

17.3

1.0

49.4

Photochemical

Reference method
Aflatoxin B1, ng/g

Sample

The Netherlands (Method A; ref. 26)


1 Senna herb

10.42

10.05

2 Senna herb

10.8

10.35

3 Peanut butter

2.24

2.22

4 Animal feed

6.50

6.49
The Netherlands (Method C; ref. 33)

1 Senna herb

15.99

16.42

2 Senna herb

13.81

14.14

3 Peanut butter

1.52

1.55

4 Animal feed

6.56

6.46

ND = None detected.

E. Standards
To establish a standard curve, prepare 5 concentrations of a
mixed aflatoxin standard containing from 0.1 to 1.0 ng toxin
per 20 mL injection solvent. Evaporate these individual
extracts to dryness under nitrogen in silanized vials.
Subsequently, reconstitute each of the extracts with 2.0 mL
injection solvent and stir with a Vortex mixer for at least 2 min.
F. Analysis
Prepare test extracts by any of the officially recognized
AOAC aflatoxin extraction methods using post-column
derivatization such as 991.31 and 999.07, or the non-LC
Method 970.45. Dry to a film in silanized vials and store frozen
at 20C until ready for evaluation.
(1) Reconstitute test extracts with 2.0 mL injection solvent
and stir with a Vortex mixer for at least 2 min.
(2) Using a 1 mL/min flow rate of the mobile phase and
20 mL injections of standards and extracts, begin with the use

of a fresh preparation of aflatoxin standards and confirm that


the equipment operation is providing the expected standard
values for aflatoxins B1, B2, G1, and G2. Adjust flow rate if
necessary to effect best separation.
(3) Inject 20 mL of each test extract, recording both peak
height and area data. If concentration is found to be below
5 ng/g, re-inject 100 mL which will increase sensitivity by a
factor of 5.
(4) Calculate concentrations using the equations in
991.31, 999.07, or derived from
Aflatoxin, ng/g = A (T/I) (1/W)
where A = ng of aflatoxin as eluate injected, T = final test
solution eluate volume (mL), I = volume eluate injected into
LC (mL), W = mass (g) of commodity represented by final
extract.
The use of at least a 3-point calibration curve is preferable but
a single-point calibration can be used if the response has

690 WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006
Table 8. Satistical evaluation: PHRED and iodine systems with naturally contaminated peanuts
Regression equations from Table 5 (19 samples); P (PHRED), I (iodine)
2

PB1 = 0.95364 IB1 + 0.7135 (r = 0.9812)


2
PB2 = 0.74882 IB2 + 0.9052 (r = 0.6384)

t-Test of the mean of the differences between P and I


Difference

Degrees of freedom

t-Value

Pr > absolute value of t

Significance

PB1 IB1

18

0.26

0.7965

Not significant

PB2 IB2

18

0.51

0.6157

Not significant

Pearson correlation coefficients


B1

0.99058

<0.0001

B2

0.79903

<0.0001

previously been shown to be linear and injections of the standard


have been made throughout the run to demonstrate the detector
response is within acceptable limits. For example, the 0.5 ng/20
mL may be used for the calibration after the curve is verified. The
useful life of the UV bulb is approximately 3000 h, or until a
consistent decrease of response (i.e., 10%) is noticed by a change
of the B1/B2 peak area ratio for the aflatoxin standard.
Reference: J. AOAC Int. 89, 678(2006).
Results
The data of the spiked samples are shown in Tables 13
(iodine, PHRED, and Kobra cell series, respectively).
Discussion
Two separate evaluations were made in this study. The first
was the tabulation of toxin-free peanut samples which had
been spiked with an aflatoxin standard as a check on their
aflatoxin analyses at 2 USDA laboratories. The summary
results in Table 4 show that essentially identical means were
obtained by the 3 different post-column derivatization

systems. The data suggest that a slightly better precision may


be obtained with the Kobra cell, while the iodine and PHRED
are equivalent. All 3 methods are acceptable, with precision
and recovery excellent for this level, and acceptable limits for
the HorRat are 0.31.3, according to Horwitz (W. Horwitz,
personal communication, 2005).
The second evaluation combined the data from paired
analyses (using the same sample or even the same extract)
with both the photochemical system and either the iodine or
Kobra cell system on naturally contaminated products.
Table 5 data are from one of the USDA laboratories analyzing
peanuts by both the iodine and photochemical systems.
Table 6 data are from the University of Georgia comparing
both corn and peanuts with the Kobra cell and the
photochemical system. Table 7 provided results from 3
European laboratories which used a PHRED photochemical
system with samples they had previously determined were
positive using a Kobra cell (Switzerland and The Netherlands)
or a similar system using pyridium bromide perbromide
(United Kingdom).

Table 9. Satistical evaluation: PHRED and Kobra cell systems with naturally contaminated peanuts
Regression equations from Table 6 (30 samples); P (PHRED), K (Kobra cell)
2

PB1 = 0.99623 KB1 + 1.4975 (r = 0.9992)


2

PB2 = 1.01096 KB2 + 0.0917 (r = 0.9959)


t-Test of the mean of the differences between P and K
Difference

DF

t-Value

Pr > absolute value of t

Significance

PB1 KB1

29

0.44

0.6622

Not significant

PB2 KB2

29

0.61

0.5440

Not significant

Pearson correlation coefficients


B1

0.99056

<0.0001

B2

0.99793

<0.0001

WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006 691
Table 10. Satistical evaluation: PHRED and Kobra cell systems with naturally contaminated corn
Regression equations from Table 6 (37 samples); P (PHRED), K (Kobra cell)
2

PB1 = 1.01578 KB1 + 0.0801 (r = 0.9992)


2
PB2 = 1.08753 KB2 + 0.0067 (r = 0.9965)

t-Test of the mean of the differences between P and K


Difference

DF

t-Value

Pr > absolute value of t

Significance

PB1 KB1

36

3.96

0.0003

Difference 99%+

PB2 KB2

36

2.99

0.0050

Difference 99%+

Pearson correlation coefficients


B1

0.99960

<0.0001

B2

0.99825

<0.0001

As stated in the design, statistical analysis was performed


only for the B1 and B2 aflatoxins because of the absence of the
G aflatoxins in so many of the products.
For all the naturally contaminated products, we ran the
SAS programs (34) for first-order regression and the t-test
using the paired samples. We also obtained the Pearson
correlation coefficients. The results of these analyses are
shown in Tables 812.
Conclusions
Only the corn data results in Table 6 were significantly
different between the PHRED and Kobra cell values,
apparently because the PHRED provided on average
0.29 ng/g more aflatoxin B1 and 0.04 ng/g more aflatoxin B2
in these samples. The PHRED values are almost exclusively
higher with only 4 out of the 74 sets providing a slightly
higher value by Kobra cell.
In paired comparisons between the PHRED and the other
methods, the Pearson correlation coefficients (r values) were

highly correlated. In all the subsets studied, the values are


above 0.98 except for one aflatoxin B2 peanut series.
However, because the aflatoxin B2 is not derivatized, it is not
affected by the PHRED. With the combined aflatoxin B1 data,
r = 0.9994; for the aflatoxin B2 data, r = 0.9874. Most of the
t-test results indicated a similar nondifference between the
various derivatization systems. The probability factor for the
correlation was P < 0.0001, indicating that the PHRED system
is equivalent to the iodine or Kobra cell portions of current
AOAC official procedures for peanuts but with a slightly high
bias for corn.
Recommendations
It is recommended that the PHRED derivatization
technique be incorporated into AOAC Methods 991.31 and
999.07 as an alternative to the use of the iodine or Kobra cell
post-column derivatization systems for derivatizing the
extracts and to be added to Method 970.45 with LC analysis,
as it is currently being used by the USDA.

Table 11. Satistical evaluation: PHRED vs mix of systems with mix of European products
Regression equations from Table 7 (16 samples); P (PHRED), O (official)a
PB1 = 0.91474 OB1 + 0.5093 (r2 = 0.9982)
2
PB2 = 1.07357 OB2 + 0.0462 (r = 0.9961)

t-Test of the mean of the differences between P and O


Difference

DF

t-Value

PB1 OB1

15

2.10

0.531

Not significant

PB2 OB2

1.53

0.1700

Not significant

Pearson correlation coefficients


B1

0.99371

<0.0001

B2

0.99803

<0.0001

Official = A combination of official methods is involved.

Pr > absolute value of t

Significance

692 WALTKING & WILSON: JOURNAL OF AOAC INTERNATIONAL VOL. 89, NO. 3, 2006
Table 12. Satistical evaluation: PHRED vs both iodine and Kobra (corn, peanuts, and foreign) combined
Regression equations from Tables 57 (102 B1 and 94 B2 evaluations); P (PHRED), O (official)a
2

PB1 = 0.96425 OB1 + 0.7353 (r = 0.9987)


2
PB2 = 0.99805 OB2 + 0.7671 (r = 0.9750)

t-Test of the mean of the differences between P and O


Difference

DF

t-Value

Pr > absolute value of t

Significance

PB1 OB1

101

0.37

0.7132

Not significant

PB2 OB2

93

1.0

0.3199

Not significant

Pearson correlation coefficients


B1

0.99935

<0.0001

B2

0.98741

<0.0001

Official = A combination of official methods is involved.

Acknowledgments
We thank Mary Trucksess (U.S. Food and Drug Administration,
College Park, MD) for helpful critique of the manuscript and
William Horwitz (FDA, retired) and Kaine Bondari (University of
Georgia, Tifton, GA) for statistical analysis.
We also thank the following collaborators for their
cooperation:
Eric Sizoo, RIVM, Bilthoven, The Netherlands
Helmut Kandler, Kantonales Labor, Zurich, Switzerland
Danny Chan, Central Science Laboratory, York, United
Kingdom
Jana Humphries, USDA, Blakely, GA
Eric Dunn, USDA, Madill, OK
David Wilson, University of Georgia, Tifton, GA
References
(1) Changes in Methods: Nuts and Nut Products (1966)
J. Assoc. Off. Anal. Chem. 49, 229231
(2) Changes in Methods: Nuts and Nut Products (1968)
J. Assoc. Off. Anal. Chem. 51, 485489
(3) Coomes, T.J., & Sanders, J.C. (1963) Analyst 88, 209213
(4) Waltking, A.E., Bleffert, G., & Kiernan, M. (1968) J. Am. Oil
Chem. Soc. 45, 880884
(5) Waltking, A.E. (1970) J. Assoc. Off. Anal. Chem. 53, 104113
(6) Whitaker, T.B., Dickens, J.W., & Monroe, R.J. (1974) J. Am.
Oil Chem. Soc. 51, 214218
(7) Whitaker, T.B., & Dickens, J.W. (1976) J. Am. Oil Chem.
Soc. 53, 545547
(8) Waltking, A.E. (1980) J. Assoc. Off. Anal. Chem. 63, 103106
(9) Sieber, R., & Blanc, B. (1978) Mitt. Geb. Lebensmittelunters.
Hyg. 69, 477491
(10) Lee, L.S., Cucullo, A.F., Franz, A.O., Jr, & Pons, W.A., Jr
(1969) J. Agric. Food Chem. 17, 451453
(11) Tiemstra, P.J. (1969) J. Am. Oil Chem. Soc. 46, 667672
(12) Waltking, A.E. (1971) J. Assoc. Off. Anal. Chem. 54, 533539
(13) Przybylski, W.J. (1975) J. Assoc. Off. Anal. Chem. 58, 163

(14) Method Bulletin (November 1982) Pert Laboratories,


Edenton, NC
(15) Tarter, E.J., Hauchay, J.P., & Scott, P.M. (1984) J. Assoc. Off.
Anal. Chem. 67, 597600
(16) Tosch, D., Waltking, A.E., & Schlesier, J.F. (1984) J. Assoc.
Off. Anal. Chem. 67, 337339
(17) Holcomb, M., Wilson, D.M., Trucksess, M.W., & Thompson,
H.C., Jr (1992) J. Chromatogr. 624, 341352
(18) Takahashi, D.M. (1977) J. Assoc. Off. Anal. Chem. 60, 799804
(19) Jaimez, J., Fente, C.A., Vasquez, B.I., Franco, C.M.,
Cepeda, A., Muhazier, G., & Progon, P. (2000)
J. Chromatogr. 882, 110
(20) Shepherd, M.J., & Gilbert, J. (1984) Food Addit. Contam. 1,
325335
(21) Kok, W.T., van Neer, T.C.H., Traag, W.A., & Tuinstra, L.G.
(1986) J. Chromatogr. 367, 231236
(22) Jansen, H., Jansen, R., Brinkman, U.A.Th., & Frei, R.W.
(1987) Chromatographia 24, 555559
(23) Trucksess, M.W., Stack, M.E., Nesheim, S., Albert, R.H., &
Romer, T.R. (1994) J. AOAC Int. 77, 15121521
(24) Joshua, H. (1993) J. Chromatogr. 654, 247254
(25) Joshua, H. (1999) Inform 10, S96 (AOCS newsletter)
(26) Stroka, J., Anklam, E., Jorissen, U., & Gilbert, J. (2000)
J. AOAC Int. 83, 320339
(27) Official Methods of Analysis (2005) 18th Ed., Ch. 49, W. Horwitz
(Ed.), AOAC INTERNATIONAL, Gaithersburg, MD
(28) Falk, J. (1999) Mycotoxin Committee Presentation, AOCS
Annual Meeting, Orlando, FL
(29) Papadopoulou-Bouraoui, A., Stroka, J., & Anklam E. (2002)
J. AOAC Int. 85, 411416
(30) Koeltzow, D., & Tanner, S. (1990) J. Assoc. Off. Anal. Chem.
73, 584589
(31) Kobra cell, manual on the instructions for use, Rhone
Diagnostics Technologies, Ltd., Glasgow, UK
(32) Sharman, M., & Gilbert, J. (1991) J. Chromatogr. 543, 220
(33) Official Method of the European Communities (1992) Off. J.
Eur. Commun. L327, 5462
(34) SAS/STAT (2003) User's Guide Version 9.1, SAS Institute
Inc., Cary, NC