TRUCKSESS & TANG: JOURNAL OF AOAC INTERNATIONAL VOL. 82, NO.

5, 1999 1109
FOOD CHEMICAL CONTAMINANTS

Solid-Phase Extraction Method for Patulin in Apple Juice and

Unfiltered Apple Juice
TRUCKSESS & TANG: JOURNAL OF AOAC INTERNATIONAL VOL. 82, NO. 5, 1999
MARY W. TRUCKSESS
U.S. Food and Drug Administration, Center for Food Safety and Applied Nutrition, Washington, DC 20204
YIFENG TANG
World Health Organization/Joint Institute for Food Safety and Applied Nutrition, Washington, DC 20204, and Shanghai
Health and Quarantine, Shanghai, Peoples Republic of China

Patulin, a mold metabolite, is commonly found in

rotting apples. Some countries regulate patulin at
levels ranging from 30 to 50 mg/L. Most analytical
methods for patulin in apple juice include liquidliquid partitions. A solid-phase extraction
method has been developed for apple juice and unfiltered apple juice in the United States. A portion
of the test sample (5 mL) was passed through a
macroporous copolymer cartridge and was
washed with 1 mL 1% sodium bicarbonate and
then with 1 mL 1% acetic acid. Patulin was eluted
with 3 mL 2% acetonitrile in anhydrous ethyl ether
and was determined by reversed-phase liquid
chromatography with UV detection at 276 nm. Recoveries ranged from 93 to104% in test samples
spiked at 20100 mg/L.

atulin, 4-hydroxy-4H-furo[3,2c]pyran-2(6H)-one, is a
lactone-containing secondary metabolite of several species of Penicillium and Aspergillus. P. expansum is the
most common mold producing patulin in apples, pears, and
cherries. Patulin contamination is primarily associated with
areas of the rotten tissue. It can penetrate up to approximately
1 cm into the surrounding healthy tissue (1). The removal of
rotten spots and surrounding tissues from apples before processing has been reported to significantly reduce patulin levels
in juice products (2).
Patulin is a colorless crystal with a molecular weight of
154 daltons and a melting point of 111EC. It is soluble in water, ethanol, acetone, ethyl acetate, ethyl ether, and chloroform
but insoluble in benzene and petroleum ether, and it is stable to
heat processing at pH < 6. Patulin is gradually destroyed during storage in the presence of sulfites, sulfhydryl groups, and
ascorbic acid. Patulin is completely degraded in 15 s in aqueous solution by 10 wt % ozone (3). Fermentation of apple
juice to produce alcoholic beverages destroys patulin (4).
Patulin has a moderate degree of cellular toxicity and has
produced local irritation and acute intoxication in humans and

Received January 27, 1999. Accepted by AP March 16, 1999.

laboratory animals experimentally exposed to doses much

higher than would occur in dietary contamination (5). Evidence
of poisoning in animals in the field is indirect and inconclusive.
Results of laboratory tests, again using levels higher than those
in dietary contamination, for immunosuppression, carcinogenicity, and teratogenicity are mixed, and their interpretation is
controversial (6). The Joint Food and Agriculture Organization/World Health Organization Expert Committee on Food
Additives (JECFA) established a provisional maximum tolerable daily intake (PMTDI) for patulin of 0.4 g/kg body
weight/day. This limit is based on the calculated no observed
effect level (NOEL) and use of a 100-fold safety factor (7).
Many countries regulate patulin in juice at levels ranging from
20 to 50 g/L (8).
Methods of analysis for patulin in apple juice usually consist of multiple liquidliquid partition steps, concentration,
and separation and quantitation by gas chromatography (9),
thin-layer chromatography (10), or liquid chromatography
(LC; AOAC Official Method 995.10; 11). Recently a
multifunctional solid-phase column (12) has been used to replace the liquidliquid partition steps. We developed a
method that is applicable to both apple juice and unfiltered apple juice. Immediately after juice is processed from apples, it
undergoes a sequence of enzymatic changes to produce the
color and the aroma. The term cider in Britain and some European countries refers to fermented apple juice that is not
covered in this method. In the United States, the term apple
juice and apple cider are used synonymously for the same
product. The only difference between the 2 for most major retail brands is the label. The term apple cider can be used to describe the unfiltered shelf-stable apple juice. The raw juice is
nearly always turbid and brown. The raw juice can be pasteurized, filtered, or enzymatically hydrolyzed to a clear juice.
Apple juice is prepared commercially by flash heating or by
the addition of ascorbic acid, followed by filtration or
centrifugation and pasteurization and packaging. Other procedures are also being used.
The method described here is simple and rapid. It takes
7 min to extract, isolate, and purify the patulin from apple
juice and unfiltered apple juice. A commercial hydrophiliclipophilic macroporous copolymer sorbent cartridge is

1110 TRUCKSESS & TANG: JOURNAL OF AOAC INTERNATIONAL VOL. 82, NO. 5, 1999

used. The sorbent is a copolymer made from a balanced ratio

of 2 monomers, the lipophilic divinylbenzene and the hydrophilic N-vinylpyrrolidone. The test sample (juice) is applied to
the cartridge. Patulin is eluted with ethyl etheracetonitrile,
separated on a reversed-phase LC column, and detected with a
UV detector set at 276 nm.
METHOD
Safety notes: Take safety precautions. Wear protective
clothing, gloves, and eye protection. See Material Safety Data
Sheets or equivalent for each reagent. Dispose of waste solvents
according to applicable environmental rules and regulations.

Apparatus
(a) Centrifuge.High speed, and 10 mL centrifuge tubes.
(b) Solid-phase extraction cartridge.Oasis HLB extraction cartridge, 3 cc/60 mg (Cat. No. WAT094226; Waters,
Milford, MA).
(c) Polypropylene 15 mL solvent reservoir.Alltech,
Deerfield, NJ.
(d) Extraction cartridge manifold.12-position, 12 needle tips, with rack for 4 mL vials (Alltech).
(e) Vacuum pump.
(f) Glassware.4 mL vials, 5 mL volumetric pipet,
500 L syringe.
(g) Heating block.With 12 ports for 4 mL vials.
(h) LC system.Programmable solvent delivery system
capable of producing gradient mixtures of 2 solvents; 2 pumps
capable of delivering 0.110 mL/min; autosampler capable of
injecting 10200 L; variable-wavelength UV detector capable of monitoring at 190300 nm, set at 276 nm; computerized
data collection system and a printer.
(i) LC column.C18 reversed-phase, 4.6 250 mm, 5 m
(MetaSil AQP#0530; Metachem, Torrance, CA).

Reagents
(a) Water.Distilled, deionized water purified with a
Milli-Q purification system (Waters).
(b) Chemicals.HPLC-grade methanol, acetonitrile, and
ethyl acetate; reagent-grade ethyl ether; anhydrous,
trifluoroacetic acid (TFA); and absolute ethanol.
(c) High-purity compressed nitrogen.
(d) Patulin.P1639, Sigma Chemical Co., St. Louis, MO.
(e) 5-Hydroxymethylfurfural (HMF).Sigma.
(f) Sodium bicarbonate 1%.Dissolve 1 g sodium bicarbonate in 100 mL Milli-Q water.
(g) Acetic acid 1%.Add 1 mL acetic acid to 99 mL
Milli-Q water.
(h) Acetic acid solution.Adjust Milli-Q water to pH 4
with acetic acid.
(i) HMF solution.Dissolve 5 mg in 25 mL ethyl acetate.
(j) Patulin standard stock solution.Weigh ca 5 mg
patulin into a 25 mL volumetric flask, record weight, and dissolve patulin in ethyl acetate (ca 200 g patulin/mL). Pipet
500 L solution into a 10 mL volumetric flask and evaporate
to dryness in a 60EC water bath under stream of nitrogen. Immediately add absolute ethanol to dissolve residue and dilute

to volume (ca 10 g patulin/mL). Measure UV absorption of

patulin standard stock solution at 276 nm. Calculate the concentration as in AOAC Official Method 974.18C(d) (10) by
using patulin molar absorptivity of 14 600 and molecular
weight of 154. Store solutions in freezer.
(k) Patulin standard working solutions.Prepare 0.1,
0.2, 0.5, and 1.0 g patulin/mL acetic acid solution corresponding to 5, 10, 25, and 50 ng patulin/50 L injection.
Transfer 100 L patulin standard stock solution (200 g/mL)
into 10 mL volumetric flask. Evaporate just to dryness under
stream of nitrogen at room temperature. Immediately dilute to
volume with acetic acid solution and mix. Transfer 50, 100,
250, and 500 L portions to separate 1 mL volumetric flasks
and dilute to volume with acetic acid solution. Store patulin
working standards solutions in refrigerator at 2E5EC. Make
new patulin working standard solutions weekly.
(l) HMF-patulin solution.Transfer 100 L patulin standard stock solution and 100 L HMF solution to a 10 mL volumetric flask. Evaporate solvent under stream of nitrogen at
room temperature. Dissolve residue and dilute to volume with
acetic acid solution.
(m) LC
mobile
phases.Mobile
phase
A:
acetonitrile0.05% TFA in water (2 + 98); mobile phase B:
acetonitrilewater (1 + 1). Degas solvents by application of a
vacuum with an appropriately trapped water aspirator connected to a faucet.

Sample Preparation
No preparation is necessary for apple juice and clear unfiltered apple juice. Place cloudy unfiltered apple juice in a
15 mL polypropylene centrifuge tube and centrifuge at
7000 rpm for 10 min.

Solid-Phase Extraction
Place cartridge on manifold. Couple solvent reservoir to
cartridge. Pass 1 mL methanol through cartridge followed by
1 mL water, letting cartridge run dry. Pipet 5 mL test sample
into cartridge reservoir coupling to cartridge. Let test sample
flow through cartridge. Wash cartridge with 1 mL 1% sodium
bicarbonate solution and then with 1 mL 1% acetic acid solution. Apply vacuum to manifold. Let cartridge dry for a few
seconds. Place a 4 mL vial under cartridge. Pipet 3 mL
acetonitrileethyl ether (2 + 98) into cartridge reservoir. Apply positive pressure on top of reservoir until solvent starts to
flow. Let solvent flow through column. Evaporate solvent under a stream of nitrogen at room temperature (25EC). Dissolve
residue in 0.25 mL acetic acid solution and retain for LC determination. Store test solution in freezer at 20EC.

Liquid Chromatography
Set flow rate at 1 mL/min with mobile phase A. Condition
column for 20 min. Use the following step gradient elution:
022 min, 100% mobile phase A; 2230 min, 100% mobile
phase B; 3045 min, mobile phase A. Set UV detector wavelength to 276 nm and sensitivity to 0.02 absorbance unit full scale
(AUFS) or adjust detector and integrator system to obtain 50%
full scale deflection for 0.5 g/mL working standard solution.

TRUCKSESS & TANG: JOURNAL OF AOAC INTERNATIONAL VOL. 82, NO. 5, 1999 1111

Evaluate LC column performance. Inject 50 L

HMFpatulin solution onto LC column. HMF and patulin
should elute as 2 separate peaks with baseline separation in ca
13 and 15 min, respectively. If HMF and patulin are not completely separated, modify the mobile phases. Analysis cannot
be performed unless HMF and patulin are separated.
Inject 50 L mobile phase A and each patulin working
standard solution. Prepare standard curve by plotting peak
area vs concentration of patulin working standard solutions.
Inject 50 L test solution. Patulin concentration in test solution can be read directly from plotted graph or calculated from
peak area of patulin working standard solution. Dilute test solution and rerun LC analysis if peak area of test solution is outside range of standard curve.

Calculation
Calculate concentration of patulin in test solution (Ct,
g/mL) as follows:

Ct = [(Cs Ht)/Hs] F
where Cs = concentration of patulin in working standard solution, Ht = response for injected test solution, Hs = response for
injected working standard solution, and F = dilution factor.
Calculate concentration of patulin in apple juice or apple
cider as follows:
Patulin in apple juice, g/L = (Ct 1000)/20
where 20 = volume ratio of apple juice test solution and 5 mL
apple juice is represented by 0.25 mL test solution.
Results and Discussion
Patulin is unstable in a basic environment and as a dry film.
It is therefore necessary to wash the cartridge with 1% acetic
acid immediately after washing with 1% sodium bicarbonate.
After evaporation of the eluate, the residue is dissolved in ace-

Figure 1. Liquid chromatograms for apple juice separated with (A) mobile phase A, CH3CN0.05% TFA (4 + 96), and
(B) mobile phase A, CH3CN0.05% TFA (2 + 98).

1112 TRUCKSESS & TANG: JOURNAL OF AOAC INTERNATIONAL VOL. 82, NO. 5, 1999
Table 1. Method performance for determination of patulin
Juice

Patulin added, ng/mL

Recovery, %a

SD, %b

RSDr, %c

Apple

20

95

7.8

8.2

Unfiltered apple juice (clear)

Unfiltered apple juice (cloudy)

b
c

50

93

7.6

8.2

100

94

3.5

3.7

20

99

13.3

13.4

50

93

1.0

1.1

100

95

0.9

0.9

20

104

4.3

4.1

50

93

4.1

4.4

100

94

3.1

3.3

Number of analyses at each added patulin level: 8, 4, and 4 for apple juice, clear unfiltered apple juice, and cloudy unfiltered apple juice,
respectively.
Standard deviation.
Within-laboratory coefficient of variation.

tic acid solution to avoid low recoveries. The test solution is

stored in a freezer at 20EC if LC analysis is delayed.
Various concentrations of acetonitrile (310%) in water
were used to wash the cartridge in an attempt to obtain a clean
final extract. Recoveries of patulin added at 100 ng/mL were
only 50% even with a cartridge prewashed with 3%
acetonitrile before washing with 1% sodium bicarbonate.
Therefore this step was eliminated.
Ethyl acetate, acetonitrile, methanol, or various combinations of water with acetonitrile and methanol were used to
elute patulin from the cartridge. Liquid chromatograms of
the eluates showed baseline elevation and interfering peaks
at or near the patulin peak. No baseline separation of the
peaks was observed. Chromatograms with the least interfering peaks were obtained by using either 3 mL 2% acetonitrile
in anhydrous ethyl ether or 5 mL ethyl ether to elute patulin.
When ethyl ether containing ethanol as preservative was
used, no clear separation of patulin from interfering peaks
was achieved.
It is important to dry the column with suction before
eluting patulin from the cartridge. Prolonged heating of the
eluate during evaporation could result in decomposition of
patulin. The time required for evaporation is about 1 min. Add
0.2 mL acetonitrile to the residual solvent if evaporation is not
complete within a short time.
The amount of acetonitrile in the mobile phase is extremely
crucial to separation of patulin from HMF and unexpected
late-eluting peak in some of the apple juice. HMF was well
separated from patulin in apple juice, clear apple cider, and
cloudy apple cider when 2% acetonitrile was used as mobile
phase A. When mobile phase A of higher acetonitrile concentration (4% acetonitrile) was used to speed analysis time, we
noticed a shift of the interfering peak from a retention time
longer than that of patulin to one overlapping with that of
patulin (Figure 1). The amount of acetonitrile in the mobile
phase is extremely crucial to separation of patulin from unexpected interfering peak in some of the apple juice. When using

an old column or other types of reversed-phase columns, the

performance of the columns must be properly evaluated.
A step gradient with mobile phase B containing 50%
acetonitrile was used to elute compounds with retention times
much longer than that of patulin. Otherwise the late-eluting
peaks sometimes would overlap with the patulin peak in the
subsequent injection. The disadvantage is that it takes at least
15 min to recondition the column.
Method detection limit was about 5 g/L. Our detector
could detect 5 ng patulin with a signal-to-noise ratio of 3:1.
Table 1 shows the performance of this method. Apple juice
and unfiltered apple juice containing patulin at <5 g/L were
used. Recoveries of patulin added at 20100 g/L to clear apple juice, clear unfiltered apple juice, and cloudy unfiltered
apple juice ranged from 93 to 104%; standard deviations and
relative standard deviations ranged from 0.9 to 13.3% and
from 1.1 to 13.4%, respectively. This method was used to analyze for patulin in 10 apple juice and unfiltered apple juice samples purchased from local grocery stores. The amount of patulin
found ranged from <5 g/L (8 samples) to 110 g/L (1 sample,
80 g/L and 1 sample, 110 g/L). A published gas chromatography/mass spectrometry method (13) was used to confirm the
identity of patulin at a level of >50 g/L in apple juice.
References
(1) Taniwaki, M.H., Hoenderboom, C.J.M., deAlmeida Vitali,
A., & Uboldi Eiroa, M.N. (1992) J. Food Prot. 55, 902904
(2) Sydenham, E.W., Vismer, H.F., Marasas, W.F.O., Brown, N.,
Schlechter, M., van der Westhuizen, L., & Rheeder, J.P.
(1995) Food Control 6, 195200
(3) McKenzie, K.S., Sarr, A.B., Mayura, K., Bailey, R.H.,
Miller, D.R., Rogers, T.D., Norred, W.P., Voss, K.A.,
Plattner, R.D., Kubena, L.F., & Phillips, T.D. (1997) Food
Chem. Toxicol. 35, 807820
(4) Ministry of Agriculture, Fisheries, and Food (1998) Food
Safety Inf. Bull. 96, 13

TRUCKSESS & TANG: JOURNAL OF AOAC INTERNATIONAL VOL. 82, NO. 5, 1999 1113
(5) Hopkins, J. (1993) BIBRA Bulletin 32, 34
(6) World Health Organization, International Agency for Research on Cancer (1985) IARC Monographs on the
Evaluation of the Carcinogenic Risk of Chemicals to Humans
40, 9293
(7) World Health Organization (1995) Forty-Fourth Report of
the Joint FAO/WHO Expert Committee on Food Additives,
Technical Report Series 859, Geneva, Switzerland, pp 3638
(8) Food and Agriculture Organization of the United Nations
(1996) FAO Food and Nutrition Paper, No. 64

(9) Tarter, E.J., & Scott, P.M. (1991) J. Chromatogr. 538,

441448
(10) Official Methods of Analysis (1995) 16th Ed., AOAC INTERNATIONAL, Arlington, VA, Method 974.18
(11) Brause, A.R., Trucksess, M.W., Thomas, F.S., & Page, S.W.
(1996) J. AOAC Int. 79, 451455
(12) Malone, B.R., Humphrey, C.W., Fleetwood, K.D., & Romer,
T. (1998) AOAC INTERNATIONAL Annual Meeting,
Montral, Canada, Abstract No. A113
(13) Roach, J.A.G. (1998) in Spectral Methods in Food Analysis:
Instrumentation and Analysis, M.M. Mossoba (Ed.), Marcel
Dekker, New York, NY, pp 159250

1114 JOURNAL OF AOAC INTERNATIONAL VOL. 82, NO. 5, 1999

JOURNAL OF AOAC INTERNATIONAL VOL. 82, NO. 5, 1999