Microchemical Journal 118 (2015) 124–130

Contents lists available at ScienceDirect

Microchemical Journal
journal homepage: www.elsevier.com/locate/microc

Determination of toxic substances, pesticides and mycotoxins, in ginkgo

biloba nutraceutical products by liquid chromatography
Orbitrap-mass spectrometry
Gerardo Martínez-Domínguez, Roberto Romero-González, Antonia Garrido Frenich ⁎
Research Group “Analytical Chemistry of Contaminants”, Department of Chemistry and Physics, Research Centre for Agricultural and Food Biotechnology (BITAL), University of Almeria,
Agrifood Campus of International Excellence, ceiA3, E-04120 Almeria, Spain

a r t i c l e i n f o a b s t r a c t

Article history: More than 250 toxic substances, including pesticides and mycotoxins, have been identified and quantified in
Received 17 July 2014 ginkgo biloba nutraceutical products using liquid chromatography (LC) coupled to high resolution mass
Received in revised form 3 September 2014 spectrometry (Exactive-Orbitrap). A “dilute and shoot” procedure, using acetonitrile with formic acid (1% v/v),
Accepted 4 September 2014
was applied for the extraction of the target compounds. In this case, after this procedure, a clean-up step was
Available online 16 September 2014
necessary using a mixture of sorbents (PSA, GBC, C18 and Z-Sep+). Parameters such as trueness, repeatability, in-
Keywords:
termediate precision, limits of detection and quantification were evaluated. Recoveries between 70 and 120%
Nutraceutical with relative standard deviation (RSD) values lower than 20% were obtained for 260 of the compounds. Limits
Ginkgo biloba of detection and quantification below 5 and 10 μg kg−1 were obtained, respectively. Nine samples of ginkgo
Pesticides biloba nutraceutical products were analyzed, finding pesticides in 8 out of 9 samples, such as hymexazol
Mycotoxins (10 μg kg−1) and tebufenozide (55–459 μg kg−1). In the case of mycotoxins, aflatoxin B1 (5–54 μg kg−1), aflatoxin
Ultra high performance liquid chromatography B2 (4–300 μg kg−1) and T-2 toxin (18–20 μg kg−1) were found in 6 samples.
High resolution mass spectrometry © 2014 Elsevier B.V. All rights reserved.

1. Introduction Among the plants that can be used for the preparation of
nutraceuticals, ginkgo biloba is widely used, obtaining the extract
Current lifestyle has provoked that society is looking for food alterna- from ginkgo leaves. Different medicinal properties have been given to
tives in order to obtain a good nutritional diet in a fast way [1]. Moreover, this nutraceutical, such as preventing and treating memory loss,
until 2010 nutrition related factors in developing nations are responsible Alzheimer sickness, vascular disease, vertigo and glaucoma [4], mainly
for 40% of mortality worldwide [2]. Dietary supplements and functional because the presence of flavonoids and terpene lactones found in
foods, also known as nutraceutical products, represent a perfect option ginkgo leaves [5].
to reduce this lack of a good nutrition, and therefore, the nutraceutical Keeping in mind that a plant extract is a concentrated form of this
market has been growing since the last decade, providing different alter- product and that the plant itself can be exposed to different toxic sub-
natives to prevent and treat modern diseases [1]. stances involved in its agricultural process, pesticides and/or myco-
The term nutraceutical, as defined by DeFelice, is a combination of the toxins could be detected in the final nutraceutical presentation. There
words “nutrition” and “pharmaceutical” referring to the foods intended to are some pesticide analysis performed in dietary supplements, such as
impart a physiological or medicinal effect after delivered ingestion in pre- Scutellaria baicalensis and Acacia catechu [6], ginseng and dandelion [7,
sentations that differ from conventional foods or beverages, such as 8]. For mycotoxins, soy isoflavones [9] green coffee beans [10], ginger
capsules, tablets, liquid extracts [1]. The National Nutraceutical Center in [11], herbs [12] wheat and oat [13] have been studied. In the case of gink-
the United States defines three types of nutraceuticals: dietary supple- go biloba, few works have been made to detect pesticides [14–16], not
ments (e.g. vitamins, minerals and plant extracts), functional foods reporting positives in any of them. Regarding mycotoxins, aflatoxins
(e. g. essentials oils) and medicinal foods (e.g. energy bars) [3]. As can [17,18] and ochratoxin A, zearalenone, deoxynivalenol, T-2 toxin and
be seen, a large variety of products can be considered as nutraceuticals fumonisins [18] were studied in ginkgo leaves, finding concentrations be-
and defining a global legislation that includes all these products is not a tween 9.1 (zearalenone) and 134.1 μg kg−1 (deoxynivalenol) [18] in the
straightforward procedure. analyzed samples. It is important to mention that these studies have
been carried out only in the raw material and not in nutraceutical
⁎ Corresponding author. Tel.: +34 950015985; fax: +34 95005008. products. As can be seen, pesticides and mycotoxins can be found in
E-mail address: [email protected] (A. Garrido Frenich). ginkgo biloba, so analytical methodologies that offer multi-class studies

http://dx.doi.org/10.1016/j.microc.2014.09.002
0026-265X/© 2014 Elsevier B.V. All rights reserved.
 G. Martínez-Domínguez et al. / Microchemical Journal 118 (2015) 124–130 125

are necessary in order to ensure food safety in this particular nutraceutical secondary amine (PSA) and graphitized black carbon (GBC) were ob-
product. tained from Scharlab (Barcelona, Spain). Bondesil-C18 was obtained
Regarding European legislation, there exist regulations that define from Agilent Technologies (Santa Clara, CA, USA). Anhydrous magne-
maximum residue limits (MRLs) of these contaminants. The Regulation sium sulfate, sodium acetate, zirconium oxide (Z-Sep+) and methanol
EC 396/2005 [19] defines MRLs for pesticides in every food and feed, LC–MS were obtained from Sigma-Aldrich.
while the Regulation EC 1881/2006 [20] establishes MRLs for myco-
toxins in foodstuffs. However, these MRLs are set in the raw material 2.2. Instrument and apparatus
and they are not applied to nutraceutical products. Therefore, method-
ologies that can determine and quantify these toxic substances in A Consul centrifuge from Orto Alresa (Madrid, Spain) was used for
ginkgo biloba nutraceutical products at low levels are necessary. centrifugation. To achieve end-over-end agitation, a Reax 2 overhead
Bearing in mind the lack of analytical methodologies to determine shaker from Heidolph (Schwabach, Germany) was used. In order to
these toxic substances in ginkgo biloba, the revised literature also in- homogenize the samples, a WX vortex from Velp Scientifica (Usmate,
cluded methods used for Chinese herbs and other extracts. Concerning Italy) was used.
the extraction procedure for pesticides, mainly the QuEChERS (quick, For chromatographic analysis, a Transcend 600 LC (Thermo Scientific
easy, cheap, effective, rugged and safe) [16,21,22], pressurized liquid Transcend™, Thermo Fisher Scientific, San Jose, CA, USA) coupled to a
extraction (PLE) [23,24], liquid–solid extraction (LSE) [25] and “dilute single stage Orbitrap mass spectrometer (Exactive™, Thermo Fisher
and shoot” [26] procedures were used. In the case of mycotoxins, specific Scientific, Bremen, Germany) was used. The system operated with a
separation with immunoaffinity columns [17,18,27] is mainly used, heated electrospray interface (HESI-II, Thermo Fisher Scientific, San Jose,
although there are other procedures that can be applied, such as the CA, USA) in positive (ESI+) and negative ionization mode (ESI−). A
micro solid phase extraction [28]. Hypersil GOLD aQ column (100 × 2.1 mm, 1.9 μm particle size) from Ther-
To achieve the chromatographic separation of pesticides, gas chro- mo Scientific (San Jose, CA, USA) was used for compound separation.
matography (GC) [21,23–26] and liquid chromatography (LC) [16,22] In order to process the data, the Xcalibur™ program version 2.2.1
are basically used. Regarding mycotoxins, LC [17,27,28] is commonly from Thermo Fisher Scientific (Les Ulis, France) with Qual and
utilized. These chromatographic techniques have been coupled to dif- Quanbrowser was used. In this case, Genesis peak detection was
ferent detectors, including low and high resolution mass spectrometers. applied. The ToxID™ program 2.1.1 and the LCQuan™ 2.6 software,
Single quadropole (Q) [21,26] and triple quadrupole (QqQ) [16,22,23] also from Thermo Scientific, were used for screening and quantification
have been the detectors most widely used for the detection of pesti- of the target compounds, respectively.
cides, whereas for mycotoxins, detectors such as QqQ [17,27] and time
of flight (QTOF) [28] have been used. Nowadays, high resolution mass 2.3. Samples
spectrometry is becoming more important because it allows the deter-
mination and quantification of a large quantity of compounds from dif- A ginkgo biloba nutraceutical product was obtained from a local
ferent classes and families, obtaining high resolving power, accurate store and was used for blank, fortified samples for recovery assays and
mass measurement and high full scan sensitivity [29]. Furthermore, a for matrix-matched standards during calibration purposes. For sample
large number of compounds can be determined at the same time, and study, three products were obtained from local supermarkets in
more than 300 compounds can simultaneously be analyzed by the Almeria, Spain, other three from pharmacies in Krakow, Poland and
Exactive-Orbitrap [30] and QTOF [31]. other three ordered from the United States. From these samples, most
For these reasons, the use of high resolution mass spectrometry is of them are gingko leaves extracts, except for two samples, one that is
needed in order to provide a comprehensive analysis of multi-class a mixture of ginkgo biloba and whitethorn (Crataegus oxyacantha)
toxic substances in ginkgo biloba nutraceutical products. Therefore, in (sample 6) and another one that also contains brown rice (sample 7).
this work a methodology has been developed to identify more than Sample preparation was carried out by homogenizing the contents
250 toxic substances, including pesticides and mycotoxins, in ginkgo of twenty capsules or tablets with a coffee grinder and stored with a
biloba nutraceuticals, studying different extraction procedures and desiccant at 5 °C in order to eliminate variations between them.
using ultra high performance liquid chromatography (UHPLC) coupled
to Exactive-Orbitrap for determination and quantification. 2.4. Extraction procedures

2. Materials and methods 2.4.1. Procedure I: acetate QuEChERS

First, acetate QuEChERS [32] was studied following the next steps:
2.1. Reagents and chemicals (1) 2 g of sample was weighed in a 50 mL centrifuge tube and fortified
with the appropriate volume of the multicompound standard solutions
Pesticide reference standards (purity higher than 99%) were pur- (pesticides and mycotoxins) to achieve a final concentration of
chased from Dr. Ehrenstofer (Augsburg, Germany) and Riedel-de- 100 μg kg−1; (2) 8 mL of water was added to the mixture and it was
Haën (Seelze-Hannover, Germany). Mycotoxins reference standards shaken vigorously for 30 s; (3) 10 mL of acetonitrile acidified with acetic
were purchased from LGC Standards (Wesel, Germany). Individual acid at 1% (v/v) was added and the tube was shaken vigorously for
stock standard solutions of 200 mg L− 1 (pesticides and mycotoxins) 1 min; (4) 4 g of magnesium sulfate and 1 g of sodium acetate were
were prepared by exact weighing of powder or liquid and dissolved in added to the mixture and shaken vigorously for 1 min, and (5) the
50 mL of HPLC-grade acetone (Sigma-Aldrich, Madrid, Spain) for pesti- tube was centrifuged at 3700 rpm (2755 g) for 10 min and 1 mL of the
cide solutions or 50 mL of acetonitrile LC–MS (Scharlab, Barcelona, organic phase was transfer to a vial for injection into the LC–Orbitrap-
Spain) for mycotoxins. Multicompound working standard solutions MS.
(344 compounds) for pesticides and mycotoxins (2 mg L−1 concentra-
tion of each compound) were prepared by appropriate dilutions of the 2.4.2. Procedure II: “dilute and shoot”
stock solutions with acetone or acetonitrile, respectively, and stored A “dilute and shoot” method tested in a previous work [30] was
under refrigeration at 4 °C. Water LC–MS (Scharlab, Barcelona, Spain) applied following the next steps: (1) 2.5 g of sample was weighed in a
was used for the preparation of buffers and other solutions. Formic 50 mL centrifuge tube and fortified with the appropriate volume of
acid (Optima LC–MS) was obtained from Fisher Scientific (Geel, the multicompound standard solutions (pesticides and mycotoxins) to
Belgium). Ammonium formate was obtained from Fluka (St. Gallen, achieve a final concentration of 100 μg kg− 1; (2) 2.5 mL of water
Switzerland). All solvents were pesticide residue grade. Primary was added to the mixture and it was shaken vigorously for 30 s;
126 G. Martínez-Domínguez et al. / Microchemical Journal 118 (2015) 124–130

(3) 7.5 mL of acetonitrile acidified with formic acid (1% v/v) was added samples in five different days at the same concentration levels for inter-
to the mixture and the tube was shaken end-over-end for 1 h, and mediate precision.
(5) the tube was centrifuged at 3700 rpm (2755 g) for 10 min Finally, to determine LODs and LOQs, five fortified samples were
and 1 mL of the extract was transferred to a vial for injection into the injected at lower concentration levels, being 0.5, 1.0, 2.0, 5.0 and
LC–Orbitrap-MS. 10.0 μg kg− 1. In this case, the LOD was defined as the minimum con-
centration where the molecular ion can be identified (mass error
2.4.2.1. Clean up step. A clean-up step was tested using a mixture of dif- value below 5 ppm) and the LOQ as the minimum concentration
ferent sorbents (PSA, GBC, C18, and Z-Sep+). After the centrifugation where it can be observed a linear response (mass error value below
step described previously, 1.5 mL of the extract was transferred to an 5 ppm).
Eppendorf microtube containing 200 mg of magnesium sulfate and
50 mg of each of the sorbents previously defined and the tube was shak- 3. Results and discussion
en vigorously for 1 min. Then, the tube was centrifuged at 3700 rpm
(1438 g) for 10 min and 1 mL of the organic phase was transferred to The Orbitrap-MS parameters were optimized in a previous work
a vial for injection into the LC–Orbitrap-MS. [33]. The exact mass database for all the studied compounds, including
elemental composition, retention time windows (RTWs), theoretical
2.5. UHPLC–Orbitrap-MS masses and accurate mass errors for the detected ions is presented in
the supplementary data (Table S-1).
A mixture of an aqueous solution of ammonium formate 4 mM and
formic acid 0.01% (v/v) (eluent A) and methanol solution with ammoni-
um formate 4 mM and formic acid 0.01% (v/v) (eluent B) was used as 3.1. Optimization of the extraction procedure
mobile phase at a constant flow rate of 0.3 mL min−1. The gradient for
the analysis of the compounds was applied as follows: the analysis First, the acetate QuEChERS [32] and the “dilute and shoot” [34]
started with 95% of eluent A. After 1 min, this percentage was linearly methodologies were compared following Procedures I and II, re-
decreased to 0% in 7 min. This composition was held during 4 min and spectively, fortifying blank samples at 100 μg kg− 1. These two pro-
increased again up to 95% in 0.5 min, followed by a re-equilibration cedures were chosen because they are quick, cheap and robust.
time of 1.5 min. The total running time was 14 min. Column tempera- Also, the solvent quantities used in the experiments are very low.
ture was set at 30 °C and the injection volume was 10 μL. In particular, the QuEChERS methodology has been used to deter-
The Orbitrap mass spectrometer parameters were as follow: spray mine pesticides in different matrices as mentioned before, whereas
voltage, 4 kV; skimmer voltage, 18 V (−18 V in ESI−); capillary voltage, the “dilute and shoot” procedure has been applied to extract pesti-
35 V (−35 V in ESI−); heater temperature, 305 °C; capillary tempera- cides and mycotoxins from different nutraceuticals, such as green
ture, 300 °C; sheath gas (N2, N95%), 35 (adimensional); auxiliary gas tea and royal jelly [33], as well as it has been successfully applied
(N2, N 95%), 10 (adimensional); and tube lens voltage, 95 V (−95 V in for the simultaneous extraction of hormones and veterinary drugs
ESI−). For full scan experiments, a mass range of m/z 100–1000 was de- from complex matrices [34]. The procedures were tested for 344
fined without lock mass. Mass accuracy was monitored as follows: the compounds included in the working standard solution. The results
use of multi-compound standards everyday to check performance; can be seen in Fig. 1a.
mass accuracy standards evaluation at least once a week and calibration It can be observed that similar results were obtained using both pro-
every two weeks. For the automatic gain control (AGC), a target value of cedures. The “dilute and shoot” methodology extracted 77% of the com-
1 × 106 was set. To obtain the mass spectra, four alternating acquisition pounds with recoveries between 70 and 120%, while the QuEChERS
functions were used: (1) full MS, ESI +, without fragmentation (the procedure extracted 64% of the compounds with suitable recoveries. Be-
higher collisional dissociation (HCD) collision cell was switched off), cause the amount of solvents and salts is used in the QuEChERS method,
mass resolving power = 25000 FWHM; scan time = 0.25 s; (2) full the “dilute and shoot” procedure represents a better option for analyti-
MS, ESI− using the same settings; (3) MS/MS, ESI+, with fragmenta- cal laboratories. Therefore, the “dilute and shoot” methodology was
tion (HCD on, collision energy = 30 eV), mass resolving power = selected for further experiments.
10000 FWHM; scan time = 0.10 s; and (4) MS/MS, ESI− using the set- Although good results were achieved at a fortified level of
tings explained before. An overall scan rate of 0.56 Hz was obtained con- 100 μg kg− 1, when lower concentrations were tested (50 μg kg− 1)
sidering the scan time of the four acquisition functions and the polarity only 27% of the compounds could be extracted with recoveries
switching (0.27 s). between 70 and 120%. Taking into account the complex matrix that
represents ginkgo biloba nutraceuticals, an additional clean up step
2.6. Validation procedure after the “dilute and shoot” procedure is recommended in order to
obtain better results and minimize matrix effect.
The method was validated in order to ensure that the results were In this case, a mixture of sorbents (PSA, GBC, C18, and Z-Sep+) was
reliable before its application in samples. The parameters studied were used following the steps presented in Section 2.4.2.1. The results can
linearity, trueness, intraday precision (repeatability), interday precision be observed in Fig. 1b. It can be observed that the use of a clean-up
(intermediate precision), limits of detection (LODs) and quantification step greatly improves the results, extracting 77% of the compounds
(LOQs). with suitable recoveries at 50 μg kg−1. Although good results were
For linearity, matrix matched standard calibration was used by achieved with this clean up step, 23% of the compounds were still unde-
analyzing spiked blank samples of ginkgo biloba nutraceuticals at five tected, probably because these compounds need a specific method for
concentration levels (2.5, 5.0, 12.5, 25.0 and 50.0 μg L−1). For trueness, their determination and quantification. Therefore, for the validation
recovery studies were evaluated by spiking blank samples with the procedure a clean-up step was included using a mixture of sorbents
corresponding volume of the multi-compound working standard (PSA, GBC, C18, and Z-Sep+) in order to improve the results at lower
solutions (pesticides and mycotoxins). Concentrations of 10, 50 concentrations.
and 100 μg kg− 1 were studied by spiking five blank samples at each
concentration. 3.2. Analytical method validation
Repeatability and intermediate precision, expressed as relative stan-
dard deviation (RSD), were studied from the injection of five spiked The optimized method was validated for 260 compounds using the
samples at 10, 50 and 100 μg kg−1 for repeatability, and three spiked best conditions defined previously.
 G. Martínez-Domínguez et al. / Microchemical Journal 118 (2015) 124–130 127

Fig. 1. Recoveries obtained for all the compounds studied using: (a) the QuEChERS (Procedure I) and “dilute and shoot” methodologies (Procedure II); and (b) the “dilute and shoot” meth-
odology without clean-up and with a clean-up step using a mixture of different sorbents (PSA, GBC, C18 and Z-Sep+). Samples were fortified at 100 μg kg−1 (a) and 50 μg kg−1 (b).

The matrix effect was studied for all the compounds analyzed, and it were also observed at this same concentration level for aflatoxin G2 and
was estimated using the following equation: demethon-S-methyl (128% for both), and hymexazol and methomyl
  (125% for both) (see Table S2). Finally, few compounds also obtained
Am low recoveries at 100 μg kg−1, as disulfoton sulfone and iprovalicarb
ME ¼ −1  100 ð1Þ
As (65%), and chlorbenside and isofenphos methyl (60%). Ethoxyquine
obtained a recovery of 125% at this level (see Table S2).
where As is the peak area for the solvent calibration point and Am the It has to be mentioned that 73 compounds could not be extracted at
peak area for the matrix calibration point. When the obtained value is 10 μg kg−1 (see Table S2). This can be explained because the Exactive-
between −20 and +20% it can be concluded that matrix effect can be Orbitrap does not have enough sensitivity to suitably detect these
considered negligible. From this study, only 3% of the compounds specific toxic substances at these low concentrations. Some mycotoxins
analyzed do not present matrix effect, whereas 94% of them got values could not be detected at the defined levels, as deoxynivalenol, fumonisin
below − 20%, presenting matrix suppression. Furthermore, matrix B1, fumonisin B2 and HT-2 toxin.
enhancement was observed for 3% of the compounds (values above In the case of repeatability, RSD values were below 20% for all the
20%). As mentioned before, the complexity of the matrix studied is the compounds studied, whereas for intermediate precision, RSD were
main reason why this matrix effect is still presented even after a below 25%. LODs were found between 0.5 and 5 μg kg− 1 while for
clean-up step was applied. LOQs between 1 and 10 μg kg−1 for most of the compounds studied,
Therefore, in order to avoid the mentioned matrix effects, matrix- except for those that could not be extracted at 10 μg kg−1. For these
matched calibration standards were used for all experiments. For linear- compounds, LODs and LOQs were 10 and 20 μg kg−1, respectively (see
ity evaluation, least-squares regression of peak area versus concentra- Table S2).
tion of the calibration standards in the whole range was applied. Good Comparing this study to other previous works in ginkgo biloba
linearity was observed, finding that determination coefficients (R2) leaves, it can be seen that the number of compounds analyzed is greatly
were higher than 0.98 in all the samples, with deviation ≤ 20% for increased. In the revised literature, between 74 [15] and 236 [16] com-
each individual level from the calibration curve. pounds were analyzed, while in this study 260 toxic substances were
Trueness, repeatability, intermediate precision, LODs and LOQs were
also evaluated. A summary of the results can be seen in Table 1, whereas Table 1
in Table S2, specific information for each compound is shown. Summary of the validation results for ginkgo biloba nutraceuticals.

In the case of trueness, recoveries between 70 and 120% were Concentration Number of compounds with suitable
obtained for most of the compounds at the three concentration levels recoveries (between 70 and 120%)
proposed before (10, 50 and 100 μg kg−1), except for diazinon, 10 μg kg−1 187
fenamiphos sulfoxide, fenpropimorph, fluazifop buthyl, flutolanil, 50 μg kg−1 260
iprovalicarb and thiazopyr, with recoveries between 60 and 64% at 100 μg kg−1 260
Repeatability b20%a
10 μg kg−1 (see Table S2). On the other hand, clodinafop propargyl, di-
Intermediate precision b25%a
sulfoton sulfone, hexazinone, mevinphos and oxydemeton methyl pro- LODsb 0.5–10.0 μg kg−1
vided recoveries above 120% (between 125 and 128%). At 50 μg kg−1 LOQsc 1.0–20.0 μg kg−1
there were also some compounds with low recoveries (between 60 and a
RSD values (n = 5).
65%), as acrinathrin, atrazine desisopropyl, diflufenican, dimethomorph, b
LOD: Limit of detection.
hexythiazox, methamidophos, phosmet and triasulfuron. High recoveries c
LOQ: Limit of quantification.
128 G. Martínez-Domínguez et al. / Microchemical Journal 118 (2015) 124–130

Table 2 the possible matrix effect. Overall, the use of high resolution mass spec-
Toxic substances, mycotoxins and pesticides detected and found in ginkgo biloba nutra- trometers gives advantages in terms of precision, robustness and com-
ceutical analyzed samples.
pounds quantity that can be analyzed at the same time. It has to be
Concentration (μg/kg) MRLa (μg/kg) mentioned that not only pesticides were analyzed, but also mycotoxins
Compounds S1 S2 S3 S4 S5 S6 S7 S8 as well using the same procedure. Therefore, for multicompound stud-
ies, the use of high resolution mass spectrometers is recommended,
Aflatoxin B1 – 5 – – – – 54 – NRb
Aflatoxin B2 – – – 4 – – – 300 NR such as the Exactive-Orbitrap used in this work.
Alachlor 207 41 – – – – – – 50
Atrazine desisopropyl – – – – – – 18 – 100
Azoxystrobin – – – – – 14 – – 50000 3.3. Application to real samples
Carbendazim – – – – – – – 25 100
Cyanofenphos – – – – 122 – – 46 NR
Etridiazole – – 22 – – – – – 50 As explained in Section 2.3, 9 ginkgo biloba nutraceuticals were an-
Fenbuconazole – – 75 – – – – – 50 alyzed with the validated method in order to determine pesticides and
Hymexazol – – 10 – – – – – 50 mycotoxins. The results can be seen in Table 2.
Propargite 22 – – – – – – – 20
From this study, pesticides were found in eight samples. In particu-
Prosulfocarb – – 23 – – – – – 2000
Pyrethrin I – – 114 – – – – – 500 lar, tebufenozide was found in three different samples (55, 67 and
T-2 toxin 18 – – – 20 – – – NR 459 μg kg− 1), while alachlor (41–207 μg kg− 1) and cyanofenphos
Tebufenozide – – 55 67 – 459 – – 100 (46–122 μg kg− 1) were found in two samples. Overall, the pesticide
Thiofanox sulfone – – – – – 42 – – NR concentrations found in these samples ranged from 10 μg kg− 1
Thiofanox sulfoxide – – 23 – – – – – NR
(hymexazol) to 459 μg kg− 1 (tebufenozide) (see Table 2). The MRLs
a
MRL: maximum residue limits for each toxic substances obtained from Regulation EC for these pesticides in ginkgo leaves defined by the European Union
396/2005 (pesticides) and Regulation EC 1881/2006 (mycotoxins).
b [19] are also presented in Table 2, finding that some concentrations
None reported.
were higher than these limits (alachlor, fenbuconazole, propargite and
tebufenozide). However, it has to be taken into account that these limits
determined and quantified. This can be explained because in other are only defined for the raw material and not for the nutraceutical prod-
works only low resolution mass spectrometers (QqQ) are used. All uct obtained from ginkgo biloba. Also, in one particular sample, seven
studies recommend the use of a clean-up step in order to minimize different pesticides were found, as can be seen in Table 2. As an

Fig. 2. Chromatograms and mass spectra obtained for: (a) tebufenozide (459 μg kg−1) and (b) aflatoxin B2 (300 μg kg−1) in ginkgo biloba nutraceuticals.
 G. Martínez-Domínguez et al. / Microchemical Journal 118 (2015) 124–130 129

example, the chromatogram and spectra for tebufenozide in sample 6 [7] K. Mastovska, P. Wylie, Evaluation of a new column backflushing set-up in the gas
chromatographic–tandem mass spectrometric analysis of pesticide residues in
(459 μg kg−1) can be observed in Fig. 2a. dietary supplements, J. Chromatogr. A 1265 (2012) 155–164.
Regarding mycotoxins, aflatoxin B1 (5–54 μg kg− 1), aflatoxin B2 [8] J. Kowalski, M. Misselwitz, J. Thomas, J. Cochran, Evaluation of QuECheRS, cartridge
(4–300 μg kg−1) and T-2 toxin (18–20 μg kg−1) were found in different SPE cleanup, and gas chromatography time-of-flight mass spectrometry for the
analysis of pesticides in dietary supplements, Agro. Food Ind. Hi Tech 22 (2011)
samples. It has to be mentioned that these levels are very high, especially 8–12.
the value obtained for aflatoxin B2 in sample 8 (300 μg kg−1), keeping in [9] J. Di Mavungu, S. Mobaliu, M. Scippo, G. Maghuin, Y. Schneider, I. Larondelle, A.
mind that some of the MRLs defined in the European legislation [20] for Callebaut, J. Robbens, C. Peteghem, S. De Saeger, LC–MS/MS multi-analyte method
for mycotoxin determination in food supplements, Food Addit. Contam. Part A 6
other matrices are below 2 μg kg−1. Also, this particular aflatoxin is con- (2009) 885–895.
sidered one of the most toxic mycotoxin [20], showing the chromatogram [10] L. Vaclavik, M. Vaclavikova, T. Begley, A. Krynitsky, J. Rader, Determination of multi-
and spectra for aflatoxin B2 in sample 8 in Fig. 2b. ple mycotoxins in dietary supplements containing green coffee bean extracts using
ultrahigh-performance liquid chromatography–tandem mass spectrometry
Bearing in mind the potential presence of mycotoxins, these com-
(UHPLC–MS/MS), J. Agric. Food Chem. 61 (2013) 4822–4830.
pounds should also be monitored in these nutraceutical products in [11] T. Whitaker, W. Trucksess, C. Weaver, A. Slate, Sampling and analytical variabil-
order to minimize potential health risks. It is important to mention ity associated with the determination of aflatoxins and ochratoxin A in bulk lots
that two samples contain different toxic substances, detecting at the of powdered ginger marketed in 1-lb bags, Anal. Bioanal. Chem. 395 (2009)
1291–1299.
same time pesticides and mycotoxins. Therefore, multi-class methods [12] L. Vaclavik, A. Krynitsky, J. Rader, Targeted analysis of multiple pharmaceuticals,
are necessary in order to identify simultaneously different toxic plant toxins and other secondary metabolites in herbal dietary supplements by
substances, like the one proposed in this article. ultra-high performance liquid chromatography-quadrupole-orbital ion trap mass
spectrometry, Anal. Chim. Acta. 810 (2014) 45–60.
[13] A. Vidal, S. Marín, A. Ramos, G. Cano, V. Sanchis, Determination of aflatoxins,
4. Conclusions deoxynivalenol, ochratoxin A and zearalenone in wheat and oat based bran supple-
ments sold in the Spanish market, Food Chem. Toxicol. 53 (2013) 133–138.
[14] L. Zhou, C. Duan, M. Wang, J. Wang, R. Zhang, Analysis of residues of 81 pesticides on
A multi-class method to determine and quantify more than 250 toxic ginkgo leaves using quechers sample preparation and gas chromatography/mass
substances in ginkgo biloba nutraceutical products, including pesticides spectrometry, J. AOAC Int. 1 (2011) 313–321.
and mycotoxins, was developed. A “dilute and shoot” procedure was ap- [15] Z. Jia, X. Mao, K. Chen, K. Wang, S. Ji, Comprehensive multiresidue method for
the simultaneous determination of 74 pesticides and metabolites in traditional
plied for extraction of the compounds, followed by a clean-up step using Chinese herbal medicines by accelerated solvent extraction with high-
a mixture of different sorbents (PSA, GBC, C18 and Z-Sep+). Trueness, in- performance liquid chromatography/tandem mass spectrometry, J. AOAC Int.
termediate precision and repeatability parameters were evaluated for 5 (2010) 1570–1588.
[16] Y. Chen, F. Al-Taher, R. Juskelis, J. Wong, K. Zhang, D. Hayward, J. Zweigenbaum, J.
validation, obtaining suitable results. Ginkgo biloba nutraceutical prod- Stevens, J. Cappozzo, Multiresidue pesticide analysis of dried botanical dietary
ucts were studied with the validated method, finding 17 different toxic supplements using an automated dispersive SPE cleanup for QuEChERS and high-
substances (14 pesticides and 3 mycotoxins) in most of them. It has to performance liquid chromatography–tandem mass spectrometry, J. Agric. Food
Chem. 60 (2012) 9991–9999.
be mentioned that, at the time of this investigation, there were few
[17] L. Liu, H. Jin, L. Sun, S. Ma, R. Lin, Determination of aflatoxins in medicinal herbs by
methods that investigate these compounds in ginkgo biloba and none high-performance liquid chromatography–tandem mass spectrometry, Phytochem.
in nutraceutical products, meaning that this is the first methodology Anal. 23 (2012) 469–476.
[18] L. Santos, S. Marín, V. Sanchis, A. Ramos, Screening of mycotoxin multicontamination
in this field, covering multi-class pesticides and mycotoxins. Taking
in medicinal and aromatic herbs sampled in Spain, J. Sci. Food Agric. 89 (2009)
into account the different classes of these toxic substances, methods 1802–1807.
that allow multi-class studies are necessary in order to ensure better [19] Council Regulation no 396/2005 of 23 February 2005 on maximum residue levels of
quality controls for these particular products and, therefore, health pesticides in or on food and feed of plant and animal origin and amending directive
no 91/414, Off. J. Eur. Communities L 70 (1) (3.16.2005) (Consolidate version up to
safety for the consumers. 02.02.2014; accessed 27.06.14).
[20] Council Regulation no 1881/2006 of 19 December 2006 setting maximum levels for
certain contaminants and foodstuffs, Off. J. Eur. Communities L 364 (5) (12.20.2006)
Acknowledgments
(Consolidate version up to 01.01.2014; accessed 27.06.14).
[21] R. Dai, X. Ren, X. He, Y. Huo, Convenient analytical method for quantitative determi-
The authors gratefully acknowledge the Spanish Ministry of Economy nation of 23 pesticide residues in herbs by gas chromatography–mass spectrometry,
and Competitiveness (MINECO) and FEDER (project ref. CTQ2012- Bull. Environ. Contam. Toxicol. 86 (2011) 559–564.
[22] E. Harris, S. Cao, B. Littlefield, J. Craycroft, R. Scholten, T. Kaptchuk, Y. Fu, W. Wang, Y.
34304). GMD acknowledges the Health Secretary from Veracruz, Liu, H. Chen, Z. Zhao, J. Clardy, A. Woolf, D. Eisenberg, Heavy metal and pesticide
Mexico and the Mexican Senate for the financial support. RRG is also content in commonly prescribed individual raw Chinese herbal medicines, Sci.
grateful for personal funding through the “University Research Plan” Total Environ. 409 (2011) 4297–4305.
[23] G. Du, Y. Xiao, H. Yang, L. Wang, Y. Song, Y. Wang, Rapid determination of pesticide
(Almería University) and Cajamar. residues in herbs using selective pressurized liquid extraction and fast gas chroma-
tography coupled with mass spectrometry, J. Sep. Sci. 35 (2012) 1922–1932.
Appendix A. Supplementary data [24] M. Rao, A. KumarMeena, Galib, Detection of toxic heavy metals and pesticide residue
in herbal plants which are commonly used in the herbal formulations, Environ.
Monit. Assess. 181 (2011) 267–271.
Supplementary data to this article can be found online at http://dx. [25] B. Łozowicka, M. Jankowska, E. Rutkowska, I. Hrynko, P. Kaczyński, J. Miciński, The
doi.org/10.1016/j.microc.2014.09.002. evaluation of a fast and simple pesticide multiresidue method in various herbs by
gas chromatography, J. Nat. Med. 68 (2014) 95–111.
[26] Y. Wang, H. Jin, S. Ma, J. Lu, R. Lin, Determination of 195 pesticide residues in Chinese
References herbs by gas chromatography–mass spectrometry using analyte protectants, J.
Chromatogr. A 1218 (2011) 334–342.
[1] D. Bagchi, Nutraceutical and Functional Food Regulations in the United States and [27] Y. Tan, Y. Kuang, R. Zhao, B. Chen, J. Wu, Determination of T-2 and HT-2 toxins in
Around the World, 2nd ed. Elsevier, USA, 2014. traditional chinese medicine marketed in China by LC-ELSD after sample clean-up
[2] Frost, Sullivan (FICCI), White Paper Global Nutraceutical Industry: Investing in by two solid-phase extractions, Chromatographia 73 (2011) 407–410.
Healthy Living, India, Available online in www.frost.com/prod/servlet/cio/ [28] L. Fang, A. Xiong, R. Wang, S. Ji, L. Yang, Z. Wang, A strategy for screening and
2361452722010 (Accessed July 2014). identifying mycotoxins in herbal medicine using ultra-performance liquid chroma-
[3] National Nutraceutical Center (NNC), http://www.clemson.edu/NNC/index.html tography with tandem quadrupole time-of-flight mass spectrometry, J. Sep. Sci. 36
(Accessed July 2014). (2013) 3115–3122.
[4] National Institute of Medicine (NIM), http://www.nlm.nih.gov/medlineplus/ [29] M. Gómez, C. Ferrer, O. Malato, A. Agüera, A. Fernández, Liquid chromatography-high-
spanish/druginfo/natural/333.html (Accessed July 2014). resolution mass spectrometry for pesticide residue analysis in fruit and vegetables:
[5] S. Myers, Adulteration Stifles the Ginkgo Biloba Market, Natural Products Insider, screening and quantitative studies, J. Chromatogr. A 1287 (2013) 24–37.
Available online in http://www.naturalproductsinsider.com/articles/2008/10/ [30] M.L. Gómez-Pérez, P. Plaza-Bolaños, R. Romero-González, J.L. Martínez-Vidal, A.
adulteration-stifles-the-ginkgo-biloba-market.aspx2008 (Accessed July 2014). Garrido-Frenich, Comprehensive qualitative and quantitative determination
[6] M. Jeohn, M. Zahn, T. Trinh, F. Brooke, W. Ma, Pesticide residue analysis of a dietary in- of pesticides and veterinary drugs in honey using liquid chromatography–
gredient by gas chromatography/selected-ion monitoring mass spectrometry using Orbitrap high resolution mass spectrometry, J. Chromatogr. A 12 (1248)
neutral alumina solid-phase extraction cleanup, J. AOAC Int. 91 (2008) 630–636. (2012) 130–138.
130 G. Martínez-Domínguez et al. / Microchemical Journal 118 (2015) 124–130

[31] M.E. Thurman, I. Ferrer, O. Malato, A.R. Fernández-Alba, Feasibility of LC/TOFMS and [33] G. Martínez-Domínguez, R. Romero-González, A. Garrido Frenich, Multi-class deter-
elemental database searching as a spectral library for pesticides in food, Food Addit. mination of contaminants in nutraceuticals by LC-HRMS, Food Chem. (2014) (under
Contam. 23 (2006) 1169–1178. review).
[32] S. Lehotay, M. Oneil, J. Tully, A. García, M. Contreras, H. Mol, V. Heinke, T. Anspach, G. [34] E. van der Heeft, Y.J.C. Bolck, B. Beumer, A.W.J.M. Nijrolder, A.A.M. Stolker, M.W.F.
Lach, R. Fussell, K. Mastovska, M. Pulsen, A. Brown, W. Hammack, J. Cook, L. Ander, K. Nielen, Full-scan accurate mass selectivity of ultra-performance liquid chroma-
Lindtner, M. Vila, M. Hopper, A. De Kok, M. Hiemstra, F. Schenck, A. Williams, A. Parker, tography combined with time-of-flight and Orbitrap mass spectrometry in
Determination of pesticide residues in foods by acetonitrile extraction and partitioning hormone and veterinary drug residue analysis, J. Am. Soc. Mass Spectrom. 20
with magnesium sulfate: collaborative study, J. AOAC Int. 90 (2007) 485–520. (2009) 451–463.