Pure & Appl. Chern., Vol. 60, No. 6, pp. 871-876, 1988.

Printed in Great Britain.

@ 1988 IUPAC

INTERNATIONAL UNION OF PURE

AND APPLIED CHEMISTRY
APPLIED CHEMISTRY DIVISION
COMMISSION ON FOOD CHEMISTRY*

A Collaborative Study of

HPLC METHODS FOR THE

DETERMINATION OF PATULIN IN
APPLE JUICE
Prepared for publication by
STANISLAW J. KUBACKI and HALINA GOSZCZ
Department of Analysis, Institute of the Fermentation Industry,
Rakowiecka 36,02-532, Warszawa 12, Poland

* Membership of the Commission during the preparation of the report (1985-87) was as
follows:
Chairman: A. E. Pohland (USA); Vice-Chairman: P. S. Steyn (Republic of South Africa);
Secretary: D. L. Park (USA); Titular Members: R. Battaglia (Switzerland); P. Krogh
(Denmark); H. A. M. G. Vaessen (Netherlands); Associate Members: M. J.-J. Castegnaro
(France); H. B. S. Conacher (Canada); R. L. Ellis (USA); J. Fremy (France); J. Jacob
(FRG); M. Jemmali (France); S. J. Kubacki (Poland); R. Livingston (USA); P. G. Thiel
(Republic of South Africa); Y. Ueno (Japan); T. Yasumoto (Japan); National Representatives:
M. M. Abdel Kader (Arab Republic of Egypt); A. Calvelo (Argentina); P. B. Czedik-
Eysenberg (Austria); J. Davidek (Czechoslovakia); E . Luck (FRG); R. LBsztity (Hungary);
Z. Berk (Israel); S. S. A. Marina Miraglia (Italy); T. Kato (Japan); C. L. Lim (Malaysia); A.
Rutkowski (Poland); L. E. Coles (UK).

Republication of this report is permitted without the need for formal IUPAC permission on condition that an
acknowledgement, with full reference together with IUPAC copyright symbol (0 1988 IUPAC), is printed.
Publication of a translation into another language is subject to the additional condition of prior approval from the
relevant IUPAC National Adhering Organization.
 A collaborative study of HPLC (high performance
liquid chromatography) methods for the
determination of patulin in apple juice
Abstract - Two HPLC methods f o r the determination of patulin i n apple
juice were collaboratively t e s t e d i n 12 laboratories from 10 coun-
t r i e s . The collaborators themselves spiked the previously pasteurized
apple juice samples with patulin standard solution before analysis.
Two samples at three l e v e l s of f o r t i f i c a t i o n and one blank sample were
analymd f o r each tested method. Although both of the methods were
based on reversed phase HPLC they employed d i f f e r e n t clew-up proce-
dures: partitioning extraction and column chromatography. Mean reco-
v e r i e s of patulin ranged from 77 t o 85 % and mean coefficients of
variation were 7.3 7 and 15 k f o r the two methods respectively.

INTRODUCTION
Oswald e t al. ( r e f . 1) found no tumorigenicitywhen patulin was administered o r a l l y t o
mice and r a t s ; nevertheless there is considerable i n t e r e s t i n t h i s myootoxin f o r two
reasons:
1 P a t u l i n i s toxic and produces tumors i n r a t s a t the place of i n j e c t i o n
when injected subcutaneously (ref. 2 ) .
2 Patulin i s a good q u a l i t y indicator of f r u i t s used i n the processing of
apple j u i c e , since the p a t u l i n concentration i s reduced by only about
20 % during processing ( r e f . 3 ) .

The absence of carcinogenicity of patulin resulted i n a lack of regulatory a c t i o n in

most co t r i e s , although Sweden, Norway and Switzerland have a c t i o n l e v e l s of
50,ug.L -ynof juice ( r e f . 4).

The most widely used quantitative t o o l s f o r patulin determination have been TLC and,
more recently, HPLC ( r e f . 5). Scott ( r e f . 6 ) organized and described r e s u l t s of a col-
laborative study of a TLC method f o r the determination of p a t u l i n in apple juice. The
method has been adopted by the A.O.A.C. a s an o f f i c i a l first a c t i o n method f o r the semi-
quantitative analysis of patulin in apple juice. Although TLC methods predominated i n
the e a r l y seventies they l a t e r gave way t o those based on HPLC. The following four rea-
sons were responsible f o r this:
1 TLC i s tedious and time consuming.
2 Confirmation is needed because of poor resolution from coextractants,
especially from 5-hydroxymethylfurfural (HMF).
3 The methods provide semiquantitative r e s u l t s .
4 The methods a r e not s u f f i c i e n t l y s e n s i t i v e (20,143.L'~, ref. 7).
For these reasons and because of recent advances i n H-SlrC technology, HPLC soon became
not only an a t t r a c t i v e a l t e r n a t i v e t o conventional TLC or GC methods but i s at present
the method of choice f o r the determination of patulin i n food products. Therefore it
has been decided by the NPAC Commission on Food Chemistry t o e s t a b l i s h an internatio-
n a l l y recommended method of analysis f o r p a t u l i n based on HPLC. Two a n a l y t i c a l methods
were selected f o r the collaborative study. Although both of them a r e based on reversed
phase HPLC they employ e n t i r e l y d i f f e r e n t clean-up procedures.

O R G A N I Z A T I O N OF THE STUDY
Due t o the decomposition of patulin i n apple juice over the several weeks time period
of a collaborative study ( r e f . a), and the formation of an i n t e r f e r i n g substance (HMF)
in apple juice a f t e r the container was opened ( r e f , 7), it was decided t h a t the colla-
borators themselves would spike the previously pasteurized apple juice samples j u s t
before analysis.
The participants were each provided with two (one f o r each method) 125 ml hypo-vials of
pasteurised apple juice concentrate, twelve (six f o r each method) 6 ml hypo-vials con-
taining aqueous s p i k i n g solutions of p a t u l i n in acetate buffer a t pH = 4, two (one f o r
each t e s t e d method) 6 ml v i a l s containing acetate buffer solution and one 6 ml v i a l
containing standard solution of p a t d i n in acetate buffer (1.25 &a). The collabora-
t o r s were asked t o s t o r e a l l the v i a l s in the r e f r i g e r a t o r until needed.
872
 Determination of patulin in apple juice by HPLC methods a73

Each laboratory was asked f i r s t t o determine the concentration of the standard solution
by means of reversed phase HPLC. Then, t o c a m out analyses of six samples of apple ju-
ice f o r each method followed by one a n a l y s i s of sample of acetate buffer. In order t o
get apple juice ready t o be analyzed by the method of Tanner and Zanier 10 g of the con-
centrate (sample S1) was diluted t o 50 m l w i t h d i s t i l l e d water and seven 5 ml portions
were transferred e i t h e r t o centrifuge tubes or t o separatory funnels f o r f u r t h e r analy-
sis. The e n t i r e contents of v i a l s A, B, C , D, E, F (containing spiking solutions) were
added t o six of the above mentioned portions of the apple juice. To reduce the losses
of the spiking substance each hypo-vial was rinsed twice with 2.5 ml of ethyl acetate.
In a l l cases bath r i n s i n g s were a l s o added t o the samples followed by partitionong
extraction. One sample of diluted apple juice was spiked with the a c e t a t e buffer (sample
C). For the method of Stray the collaborators were instructed t o d i l u t e 100 g of the
apple juice conoentrate (sample S2) t o 500 ml with d i s t i l l e d water. Seven 50 ml portions
of the diluted juice were transferred t o i n d i v i d u a l m u l s e p a r a t o r y funnels f o r further
analysis. Six of them were f o r t i f i e d with the spiked solutions i n V i a l s H, I, K, L, M
and N. The acetate buffer sample was added t o the last portion of the diluted juioe.
All the v i a l s containing spiking solutions and buffer acetate were rinsed with two 5 m l
portions of e t h y l a c e t a t e taken from the volume ( 5 0 ml) dedicated f o r the first p e r t i -
tioning eXtZ%CtiOR. The p a r t i c i p a n t s were asked t o complete the analyses f o r each me-
thod in one working day.

METHODS
The following two HPLC methods were c o l l a b c r a t i v e l r studied on the b a s i s of the previo-
usly prepared l i t e r a t u r e survey(ref. 13):
1 method developed by Stray ( r e f . 31,
2 method by Tanner and Z a n i e r ( r e f . 9). T h i s method had been published by
the same authors e a r l i e r ( r e f . 10) before it was adopted by the Federal
Commission f o r the Swiss Food Manual as o f f i c i a l ( 1984).
Fig. 1 and Fig. 2 show flow diagrams of the method of Tanner and Zanier and the method
of Stray respectively.
The s t a t i s t i c a l analyses of the r e s u l t s were carried out following guidelines recommen-
ded by the International Standards Organization ( r e f . 11). Additionally, analyses of
variances according t o Steiner ( r e f . 12)were a l s o performed.

Sample ( 5 ml)
1 TAIILZ 1. The collaborative r e s u l t s f o r the
Extraction with e t h y l a c e t a t e
t determination of patulin i n the standard
Partitioning extraction by aqueous solution
sodium carbonate solution 7

! discarded Lab. Standard solution

Conoe t r a t i o n (1.25r@;/ml)
1 1.75
HPLC ( ODS oolumo) 2 1.62
3 1* 5 0
Big. 7 . Schematic representation of Tanner 4 1.28
and Zanier procedure 5 1.14
6 1.38
7 1.22
Sample (50 m l ) 8 1.19
9 1.25
4
Extraction with e t h y l acetate
10 1.80
11 1.60
1
Ccncentrat ion
12 1.10
i
Adsorption chromatography of patu- Mean 1.35
lin on s i l i c a gel, solvent; Standard
toluene-ethyl acetate ( 7 t 3 ) deviation 0.22
t
Concentration Rel. Std.
t Dev. ( % ) 16.8
HPLC (ODs column 1
Fig. 2. Schematic representation of s t r a y
procedure
874 COMMISSION ON FOOD CHEMISTRY

RESULTS

The collaborators were provided with the samples in July, 1 9 8 6 and the r e s u l t s were
returned by October 15, 1986. O f the 21 l a b o r a t o r i e s invited t o p a r t i c i p a t e in the stu-
dy, 13 agreed t o take part. Finally 1 2 laboratories from 10 countries submitted results
( s e e Acknowledgments).
From the collaborators r e s u l t s f o r the concentration of the standard solution sample (ST),
determined by reversed phase (ODs column) HPLC, the 1.25,ug/ml s o l u t i o n i n a c e t a t e buf-
f e r ( pH = 4 ) appears t o be s t a b l e (Table 1). Some l a b o r a t o r i e s (nos. 2 , 3, 10 and 11)
f a i l e d t o reproduce the t r u e concentration with s u f f i c i e n t accuracy [< 10 $6).
The r e s u l t s obtained by the p a r t i c i p a n t s f o r the concentration of patulin in the spiked
apple juice (samples A through F and samples H through N) as well as i n the blank apple
juice (samples 0 and 0 ) a r e tabulated i n Table 2 f o r the method of Tanner and Zanier
and i n Table 3 f o r the method of Stray. The samples were spiked in duplicate at three
known ooncentrations; 5, 50 and 250,ug.L-l. The f i r s t spiking concentration was e i t h e r
at the limit of detection (method o Tanner and Zanier) o r very close t o it (method of
Q
Stray, l i m i t of detection = 1 pg.L- 1. Most of the findings obtained f o r these samples
indicate that the limits of detection reported by the authors of the methods a r e beyond
the reach of most laboratories. Because of the presence of i n t e r f e r i n g substances i n
e x t r a c t s as analyeed by HPLC the r e a l limits of detection a r e between 10 and 20 ug.L'l.
Occassionally f a l s e positives estimated t o be a t the l e v e l of 10,ug.L" were recorded
f o r the blank samples G and 0.
It i s e a s i l y seen from a cursory examination of the data i n Table 2 and Table 3 t h a t
some laboratories a r e outliers. The r e s u l t s provided by the l a b o r a t o r i e s nos. 5, 8, 10
and 11 f o r the method of Tanner and Zanier and the laboratories nos. 3, 4 and 1 1 f o r
the method of Stray deviate so much from comparable e n t r i e s from other l a b o r a t o r i e s that
they may be considered as irreconcilable w i t h the other data without a p p l y i n g Dixon's
o u t l i e r t e s t . 1% was apparent that these l a b o r a t o r i e s d i d not have the methods under
control. Additional enquiry sent t o the participants revealed that a l l the l a b o r a t o r i e s
except one i d e n t i f i e d as o u t l i e r s had never used the methods before. I n most oases poor
HPLC resolution of p a t u l i n from concurrent i n t e r f e r i n g substances was responsible f o r
the erroneous r e s u l t s . An apple juice concentrate r e l a t i v e l y r i c h i n i n t e r f e r i n g substan-
ces was deliberately selected f o r t h i s study. The rejected o u t l i e r s i n Table 2 and in
Table 3 a r e put i n parenthesis.
Average recoveries obtained by $he collabmrators f o r the concentration of patulin i n the
spiked apple juice were 85 $% (samples B and E) and 83 % (samples C and F ) f o r the method
of Tanner and Zanier and 80 $% (samples I and $4) and 77 $ (samples K and I?) f o r the me-
thod of Stray f o r the spiking l e v e l s 50,ug.L' an8 2 5 0 p . L W 1 respectively. The d i f f e -
-
rences between the mean recoveries found f o r both the spiking l e v e l s and the methods
were not s t a t i s t i c a l l y distinguishable ( t t e s t ) .

pared with TLC (mf. 6 ) .

The r e s u l t s showed the low v a r i a t i o n anticipated f o r a method that i s based on HPLC com-
Coefficients o f v a r i a t i o n were 7.5 (samples B and E ) and

TABLE 2. R e s u l t s reported by participants (method by Tanner and Zanier)

Lab. 5 M.L'l 50 ,ug.L" 250 ,ug.L" 0

~

A D B E C P 0

1 2 4 40 37 2 03 201 Trace
~~

2 R e s u l t s n o t r e p o r t e d
-
3 5.8 3.9 42.4 36.6 177.3 202.3 2.0
4 26.7 n.d. 40.0 46.7 186.2 179.3 n.d.
5 n. d. n.d. (n.d.) (n.d.) (25) (360) n.d.
6 20 20 49.2 47.2 209.2 246.2 20
7 8.9 n.d. 45.6 50.8 218.4 217.1 nod.
8 10 20 (10) (interf .) (interf .) ( 208) 10
9 10 5 40 35 220 250 5
10 91 580 ( 235 (139) (471) 266) 115
11 Destroyed - (1500) (1400) (-) (1900) 1300

i 1 rejected outliers
 Determination ofpatulin in apple juice by HPLC methods a75

7.2 [samples C and F) f o r the method of Tanner and Zanier and 12 % ( samples I and M )
and 18 ii ( s m l e s K and N) f o r the method of S t r a y f o r the spiking l e v e l s of 50,ug.L-’
and 25Opg.L-7 respectively. The mean c o e f f i c i e n t s of v a r i a t i o n were 7.3 % f o r the me-
thod of Tanner and Zanier and 15 % f o r the method of Stray, Calculation of within and
between-laboratory components o f t o t a l variances (ref.12) f o r the two sample s e t s ( s p i -
k i n g l e v e l s ) f o r each method revealed the l a r g e random e r r o r c o n t r i b u t i o n t o t h e t o t a l
variance ( Table 4 and Table 5). F-ratios Sd2/Sr2 gave no evidence f o r t h e presence of
s i g n i f i c a n t systematic e r r o r s among t h e l a b o r a t o r i e s . The values of r ( r e p e a t a b i l i t y )
and R [ r e p r o d u c i b i l i t y ) computed according t o t h e IS0 g u i d e l i n e s a r e shown i n Table 6.
They mean t h a t t h e d i f f e r e n c e between two s i n g l e determinations found e i t h e r i n one la-
boratory o r i n t w o d i f f e r e n t l a b o r a t o r i e s on i d e n t i c a l t e s t m a t e r i a l w i l l exceed the
r e p e a t a b i l i t y ( r 1 o r r e p r o d u c i b i l i t y ( R 1 r e s p e c t i v e l y not more than in 5 % of t h e cases
(95 k p r o b a b i l i t y ) .

TABLE 3. Results reported by participants (method by Stray)

Lab. 5pgX7 50 250 ,ug.L-’ 0

~

H L I M X B 0

1 n.a. n.d. 31 Trace 162 133 12

2 6.0 n.m 40.8 31.2 211.4 145.4 0.4
3 10 10 [ 33.9) (-1 (124.3) (76.7) 10
4 n.d. n.d. ( 16.7) ( 8.3) (86.4) (65.0) n.do
5 n.d. n.d. 23 30 163 150 n. de
6 10 14.8 36.9 47.6 199.2 270.6 10

7 R e s u l t s n o t r e p o r t e d

8 10 10 47.2 45.7 220.6 138.0 10

9 5 15 50 45 250 250 5
10 49 54 n.a. 48 203 160 16
11 Destroyad 1770 (1410) (1710 ) (1 670) 890) 1480
12 0 0 38.1 40.9 220.6 216.3 0
~~

( ) rejected outliers

TABLE 4. Within-and between-laboratory variances of patulin analyses method of Tanner and Zanier

Samples
(spiking Average
patulin Average
recovery Total W ithin-lab. Be tween-lab,
level) found s; COY. ( % ) sr2 C.V. (7%) s; C.V. (%)
($1
LUe;.L-l,
B, E
(50 ,W.L-~) 42.2 80.0 24.2 11.6 10.2 7.56 14.0 8.06
c, F 107.0 82.8 512.4 10.9 223.9 7.23 288.5 8.20
(250 pg.L”)

TABLE 5. Within-and between-laboratory variances of patulin analyses {method of Stray)

( sSamples
piking Average
patulin Average
recovery Total W ithin-lab. Be tween-lab
level) found C.V.(%l
(%)
(/Ug.L-l/

1, M
(50 ,ugeL-’) 39.7 79.4 71.6 21.3 24.2 12.4 47.4 17.3

x, N 193.3 77.3 1967.4 22.9 1197 17.9 770.4 14.3

( 250 ,ug.Lml)
876 COMMISSION ON FOOD CHEMISTRY

TABLE 6. Computed values o f e.d., r and R

Samples Sed. r s.d. R

(spiking level) ( sr 'b)

Method by Tanner and Zanier

B,
5Om.L-l) 3.19 8.9 3.74 10.5
c, F
( 2 5 0 pg.L-l) 14.9 41.9 16.9 47.5
Pnethod by S t r a y
I, M
( 50,w.L-l) 4.91 13.7 6.88 19.3

K, N
( 2 5 0 pg.L-') 34.6 96.8 27.7 77.7

s.d. standard d e v i a t i o n : S r within-laboratory s.d.: s b between-laboratory s.d.

RECOMMENDATIONS
Both the t e s t e d m e t h o d s f o r the d e t e r m i n a t i o n o f p a t u l i n i n a p p l e j u i c e b y means of H i 5 C
a h o u l d be o f f i c i a l l y recommended by IUPAC.

ACKNOWLEDGEMENTS
Tho a u t h o r s a r e g r e a t l y indebted t o D r . A.E. Pohland, US Food and D r u g Administration,
Center f o r Food S a f e t y and Applied N u t r i t i o n , Division of Contaminants Chemistry, Natu-
ral Products snd I n s t r u m e n t a t i o n Branch, Washington, D.C. 20204 f o r providing t h e patu-
l i n standard and kind a s s i s t a n c e i n preparing t h i s a r t i c l e . Many thanks a r e due t o t h e
followiw collaborators for t h e i r participation:
I.B. Agater, Laboratory of t h e Government Chemist, London, England.
J. Anderson, Council f o r S c i e n t i f i c and I n d u s t r i a l Research, Natural
Food Research I n s t i t u t e , P r e t o r i a , Republic of South Africa.
R. B a t t a g l i a , Xantonales Laboratorium Zurich, Zurich, Switzerland.
J.R. Bealing, Food I n s p e c t i o n S e r v i c e , Rotterdam, The Netherlands.
J.M. Fremy, L a b o r a t o i r e C e n t r a l d 'Hygiene Alimantaire, D i r e c t i o n de l a
Q u a l i t e S e r v i c e s V e t e r i n a r i e s , P a r i s , France.
H. GOSZCZ, I n s t i t u t e of the Fermentation I n d u s t r y , Warszawa, Poland,
W. iKronert, Max von P e t t e n k o f e r - I n s t i t u t des Bundesgesundheitsamtes,
B e r l i n , FRG.
C. L a i r i e , D i r e c t i o n du L a b o r a t o i r e du Controle de l a Q u a l i t e , Rennes,
France.
C.W. Thorpe, Food and Drug Administration, Washington, Dc, USA.
P. T h i e l , National Research I n s t i t u t e f o r N u t r i t i o n a l Diseases, SA Ide-
d i c a l Researoh Council, Tygerberg, Republic of South Africa.
A. Vidnes, National Control Authority f o r Processed F r u i t s and Vegeta-
b l e s , Oslo, Norway.
A. V i s c o n t i , I n s t i t u t o Tossine e l i c o t o s s i n e CNR, Bari, I t a l y .

REFERENCES
1. H. Oswald, H.K. Frank, D. Komitowski and H. iVinter, Fd. Cosmet. Toxicol., 16,243-
247 (1978).
2. F. Dickenb and H.C.H. Jones, B r i t . J. Cancer, 3, 85-89 (1961).
3. H. S t r a y , Journal of t h e A.O.A.C., 6 l , 1359-1362 [ 1978j.
4. P.M. S c o t t , H.L. Trenholm and I.D. Sutton, Mycotoxins: A Canadian PersI)ective,
p.29, P u b l i c a t i o n s NHCC/CNRC, Ottawa (1985).
5. N.T. Crosby, Food Additives and Cantaminants, 1 , 39-44 (1984).
6.
7.
P.M. S c o t t , Journal of t h e A.O.A.C.,
P.M. S c o t t and B.P.C.
a,621-6?5 (1974).
Kennedy, Journal of t h e A.O.A.C., s,
813-816 (1973).
8. P.M. S c o t t and E. Somers, Journal A R r . Food Chem., 9, 483-485 (1968).
9. H. Tanner and C. Zanier, M i t t . Gebiete lebsm. Hyg., a,
506-513 (1984).
10.
11.
12.
IS0 I n t e r n a t i o n a l Standard no. 5725 (1981) .
H. Tanner and C. Zanier, Schweie. 2. Obst und Weinbau, 112, 656-662 (1976).

W.J. Youden and E.H. S t e i n e r , S t a t i s t i c a l Manual of t h e A.O.A.C. p.p. 66-83,

A.O.A.C.,Washington D.C. (1975).
13. P.S. Steyn and R. Vleggaar, Mscotoxins and Phvcotoxins, pip. 293-304, E l s e v i e r
Science Publisher B.V., Amsterdam, New York, London, (1986).