Determinations for Pesticides on Black, Green, Oolong, and White

Page 1 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Journal of Agricultural and Food Chemistry

Determinations for Pesticides on Black, Green, Oolong and White Teas by Gas Chromatography Triple-Quadrupole Mass Spectrometry Douglas G. Hayward1, Jon W. Wong1, Hoon Y. Park2 1

U.S. Food and Drug Administration, Center for Food Safety and Applied Nutrition, 5100 Paint Branch Parkway, HFS-706, College Park, MD 20740-3835

2

Joint Institute for Food Safety and Applied Nutrition, University of Maryland 1122 Patapsco Building, College Park, MD 20742-6730

Correspondence addressed to: Douglas G. Hayward, Tel: (240)-402-1654; FAX: (240) 402-2332; email: [email protected] , Jon W. Wong, Tel: (240)-402-2172; FAX: (301)-4362332; email: [email protected]

Keywords: multi-residue pesticide analysis, tea, GCB/PSA SPE, GC-MS/MS Running title: Determinations for Pesticides on Black, Green, white and Oolong Teas measured by GC-MS/MS

1 ACS Paragon Plus Environment


ABSTRACT

22 23 24

Black, green, white, and Oolong teas, all derived from leaves of Camellia sinensis, are

25

widely consumed throughout the world and represent a significant part of the beverages

26

consumed by Americans. A gas chromatography-triple quadrupole based method, previously

27

validated for pesticides on dried botanical dietary supplements, including green tea, was used

28

to measure pesticides fortified into black and green teas at 10, 25, 100 and 500 μg/kg. Teas

29

from 18 vendors of tea products were then surveyed for pesticides. Of 62 black, green, white,

30

and Oolong tea products, 31 (50%) had residues of pesticides having no United States

31

Environmental Protection Agency tolerance established for tea. The following pesticides were

32

identified on tea leaves, with concentrations between 1-3200 μg/kg: anthraquinone,

33

azoxystrobin, bifenthrin, buprofesin, , chlorpyrifos, cyhalothrin, cypermethrin, DDE-p,p’, DDT-

34

o,p, DDT-p,p’, deltamethrin, endosulfan, fenvalerate, heptachlor, hexachlorocyclohexanes

35

(α,β,γ,δ), phenylphenol, pyridaben, tebuconazole, tebufenpyrad and triazophos. DDT-p,p’ was

36

found at much higher concentrations than DDE-p,p’ or DDT-o,p’ in 9 of 10 teas with DDTs. A

37

comparison between three commercially-available solid-phase extraction (SPE) column brands

38

of the same type revealed that two brands of SPE columns could be interchanged without

39

modification of the tea method.

40 41 42 43

INTRODUCTION


Page 2 of 30

Page 3 of 30


44 45

Beverages made from the extraction of tea leaves (Camellia sinensis) are the most widely

46

consumed drinks in the world besides water1. Tea beverages are consumed for their aromatic

47

flavors as well as for their presumed medicinal properties. Notably, green teas have been

48

associated with improving cardiovascular health and reduced inflammation1. Tea consumed in

49

the United States is imported from countries such as India, China, Kenya, Sri Lanka and Turkey,

50

the major producers in the world. Teas are often marketed for their benefits to health although

51

they have no direct nutritional value in the diet as do other foods and beverages.

52

The United States Food and Drug Administration (U.S. FDA) has the responsibility for

53

ensuring that foods do not contain pesticide residues in excess of established tolerances or

54

pesticides that have not been registered for use on tea by U.S. Environmental Protection

55

Agency (US EPA). Table 1 lists the 18 pesticides registered by U.S. EPA for use on tea, along with

56

their tolerances as of December, 2013. In response to a report that certain green, black and

57

Oolong teas contained residues of one or more pesticides not registered for use on tea by the

58

U.S. EPA, the FDA conducted a cross-sectional survey of teas in 20132.

59

Previous studies reported pesticides on teas, such as the following with their registered

60

tolerances in the US, acetamprid4 (50 ppm), bifenthrin3,5,6(30 ppm), buprofesin3,5 (20 ppm), and

61

dicofol8 (50 ppm) (Table 1). Another frequently reported pesticide, cypermethrin, is registered

62

in the US for tea plant oil with a tolerance of 0.4 ppm (Table 1). Pesticides with no US registered

63

tolerance have been reported on tea such as cyhalothrin3,6, DDTs6,7, endosulfans5,6,

64

fenvalerate3, hexachlorocyclohexanes (HCHs)6and imidicloprid4. Persistent organic pollutants



65

(POPs), including polychlorinated dibenzo-p-dioxins/dibenzofurans (PCDD/Fs), have also been

66

reported on teas7.

67

Dried botanical products, such as tea, can present severe challenges to measuring trace

68

pesticide residues at 10 μg/kg3,4,6. Pang et al. (2013) developed and collaboratively tested a

69

method for teas using acetonitrile for extraction of the dry tea9 followed by clean up with SPE

70

column containing graphitized carbon black (GCB) and primary and secondary amine silica gel

71

(PSA)10. Both GC-MS and LC-MS/MS were used for the measuring 650 pesticides on black,

72

green and Oolong teas9,10. Pang et al.10 also compared 12 SPE column combinations of C18,

73

PSA, GCB, and NH2 in his method and determined that they performed in a comparable

74

manner, but stated that two vendor’s dual phase GCB/PSA were preferred10. The QuEChERS

75

(Quick, Easy, Cheap, Effective Rugged and Safe)11 extraction procedure has also been adapted

76

for use on tea leaves after adding water to the dry tea3,4,6,12 then using GC-MS/MS3,6,12 and LC-

77

MS/MS4,12 and/or LC-QTOF4 for measurement. Other methods for extraction have also been

78

employed, for example pressurized solvent extraction with ethyl acetate:hexane13 or subcritical

79

water14.

80

Acetonitrile extracts using QuEChERS from botanicals like tea have too much co-

81

extracted matrix to be injected directly into GC-MS3,6,12. Cleanup methods have included using

82

dispersive SPE (as used in QuEChERS with the addition of GCB4 or with CaCl2 instead of

83

MgSO46,12), using dispersive multi-walled carbon nanotubes15, using liquid:liquid extraction

84

(LLE)5, or using an SPE column containing PSA and GCB3,10. Recently, comparisons of three9,16 or

85

four6 different extraction and clean up procedures have been published, although some with

86

mixed conclusions. One suggested that hydration was essential for good recoveries6, while


Page 4 of 30

Page 5 of 30


87

another concluded the opposite9. A third study concluded that QuEChERS, or a commonly used

88

ethyl acetate extraction, was more effective than a “mini-Luke” procedure based on the older

89

Luke method16. In 2008, Lee et al.,17 also compared three extraction and clean-up procedures

90

for 49 pesticides and concluded that QuEChERS11 gave superior recovery for some pesticides

91

when compared to LLE after hydration or pressurized liquid extraction with acetone for tobacco

92

leaves.

93

In a prior study in our lab with 310 pesticides3, a dual phase GCB/PSA SPE cleanup was

94

shown to provide good results with green tea. In this study, we determined whether two

95

alternative brands of SPE dual phase cleanup columns (carbon/PSA) could be used in our

96

method without modification3. We then employed the method3 with an alternative SPE brand

97

for measuring 170 pesticides in 62 teas collected from 18 different vendors.

98 99 100

MATERIALS AND METHODS

101 102

Materials and Standards Preparation. Pesticide standard mixes (custom-made

103

solutions in acetone at 100 mg/L) were obtained from AccuStandard (New Haven CT), and are

104

prepared for the US FDA pesticide monitoring laboratories for GC-MS analysis. The following 12

105

custom mixes were used: S-10954-R2, S-10236-R4, S-10238-R4, S-10237A-R3, S-10693B-R1, S-

106

10237B-R2, S-10696B-R1, S-10956-R1, S-10953, S-10952-R1, S-10694-R6, S-10692A-R3. These

107

custom mixes contained a combined total of 227 pesticides, synergists, isomers, metabolites,

108

flame retardants and plastizers. The internal standards were hexachlorocyclohexane-D6 (α)

109

methoxychlor-p,p’-D6, and tri-butyl phosphate-D27 (C/D/N Isotopes Inc. Pointe-Claire, Quebec,



110

Canada) and the mass spectrometry quality standards were acenapthalene-D10, phenanthrene-

111

D10, and chrysene-D12 (Sigma-Aldrich Milwaukee, WI). The internal standard, tris-(1,3-

112

dichloroisopropyl) phosphate, was purchased from TCI America (Portland, OR). Pesticide-grade

113

acetonitrile and toluene, HPLC-grade water, and certified-grade anhydrous sodium sulfate and

114

sodium chloride were purchased from Thermo Fisher Scientific (Pittsburgh, PA). The SPE

115

columns, consisted of a dual phase 250 mg GCB (top layer) and 500 mg PSA (bottom layer) with

116

a Teflon frit (p/n ECPSACB256) (“ECPSA” in the Tables), purchased from United Chemical

117

Technologies (UCT) (Bristol, PA). Two alternative SPE columns were: Agela “Cleanert” (p/n

118

TPT0006) (“TPT” in the Tables) dual phase consisting of 500 mg PSA (top layer) and 500 mg GCB

119

(bottom layer), purchased from Bonna-Agela Technologies Inc. (Wilmington DE), and a dual

120

phase QuECHERS SPE tubes (p/n 1605349139) from United Sciences (Center City, MN )

121

consisting of columns with 500 mg carbon X Plus, carbon coated on alumina (COA in tables)

122

with 6% synthetic carbon coated to it (top layer) and 500 mg PSA (bottom layer).

123

The following types of black, green, Oolong, or white teas (Camellia sinensis) were

124

collected: “black”, “Ceylon orange pekoe”, “decaffeinated green”, “decaffeinated Earl Grey”,

125

“decaffeinated Lady Grey” “Darjeeling”, “English Breakfast”, “Earl Grey”, “green”, “Irish

126

Breakfast”, “Jasmine green”, “Lady Grey”, “Lapsang Souchong”, “Oolong”, “Organic Green

127

Mint”, “Organic Jasmine”, “Rose Oolong”, “Dao Ren”, “Pu-erh”, “Rooibos”, and “Rooibos red”.

128

These tea products were purchased or obtained from retail commercial outlets, including tea

129

marketed as “organic” retail sources believed to be free of pesticide residues. All products

130

were accepted as being the product indicated by the label and our selection of products was

131

not intended to be a representative sampling any particular tea product sold in the US, nor


Page 6 of 30

Page 7 of 30


132

represent all tea products sold in the US. The black and green teas used for fortification were

133

teas marketed as “organic” and therefore represented to be free of pesticides. Analysis of the

134

unfortified teas did not detect the pesticides monitored in this study.

135

Twelve AccuStandard custom mixes were combined in a 10 mL volumetric flask by

136

transferring 0.5 mL from each of the twelve mixes. The remaining volume was made up with

137

toluene to produce a stock at 5 mg/L for calibration and fortification standards. The working

138

standards used for quantification were prepared by dilution of this 5 mg/L mixture to give 50

139

mL of each in a volumetric flask for 0.5, 0.2, 0.1 0.05 mg/L in toluene. The resulting standards

140

were then used to prepare lower concentration calibration standards at 0.020, 0.010, 0.005,

141

0.002 and 0.001 mg/L of the pesticides in toluene. The internal and quality control standards

142

were

143

dichloroisopropyl)phosphate and tributyl phosphate-D27 to a 0.060 mg/L working solutions in

144

acetonitrile, and the deuterated polycyclic hydrocarbons to a 0.50 mg/L working solution in

145

toluene.

prepared

by

dissolving

hexachlorocyclohexane-D6,

methoxychlor-D6,

tris-(1,3-

146

Tea Leaf Extraction. Tea leaves were ground and homogenized to a fine powder that

147

would passed through an 18 mesh sieve. Dried homogenized tea powder (1.00 ± 0.02 g), 10 mL

148

water,

149

hexachlorocyclohexane-D6 (α), methoxychlor-D6, tris-(1,3-dichloroisopropyl) phosphate, and

150

tributyl phosphate-D27 in acetonitrile) were added to a disposable 50 mL centrifuge tube. To

151

ensure complete wetting of the botanical samples, tubes were shaken and allowed to stand for

152

15 minutes after the water and the extraction solvent were added. All tea test portions were

153

extracted as previously described3.

and

10

mL

extraction

solvent

(0.060

mg/L


of

the

internal

standard,


154

Solid-phase Extraction Cleanup. The SPE system consisted of a manifold fitted with 24

155

conditioned SPE columns, with a rack containing 15 mL glass centrifuge tubes. These SPE

156

cartridges were conditioned with 3 column volumes of acetone and remained wetted with

157

acetone. The collection rack, consisting of the 15 mL disposable centrifuge tubes was inserted

158

in the vacuum manifold. An aliquot (1.25 mL) of the mostly acetonitrile containing upper layer

159

of the centrifuged tea leaves/salt/water/acetonitrile mixture was loaded and allowed to

160

percolate onto a tandem SPE column, “ECPSA” topped with anhydrous sodium sulfate. The

161

columns were rinsed with 1 mL of acetone and eluted and collected with 12 mL 3:1

162

acetone:toluene. The cleaned extract was reduced to approximately 100 µL with a gentle

163

nitrogen stream and a water bath (50 °C) using a nitrogen evaporator (N-Evap, Organomation,

164

Associates, Berlin, MA). Toluene (0.5 mL), QC standards (50 µL of deuterated polycyclic

165

aromatic hydrocarbons mixture, 0.5 mg/L), and 25 mg of magnesium sulfate were added and

166

the extracts were centrifuged3 at 3000 rpm x 5 min. The toluene extracts were divided

167

between two GC vials with 250 μL vial inserts (Restek Corp, Bellefonte PA) in order to create a

168

duplicate reserve set of vials.

169

The TPT SPE columns were operated exactly as the ECPSA columns. The COA SPE

170

columns were operated exactly as described for the ECPSA columns, except the COA columns

171

were eluted with 12 mL of either acetone or ethyl acetate in place of 12 mL 3:1

172

acetone:toluene. Incurred teas were cleanup using the COA column with acetone elution.

173

Fortification Studies. Black tea and green tea test portions were fortified at 10, 25, 100

174

and 500µg/kg. Test portions (1.0 g) were fortified by adding 100 µL of the appropriate

175

fortification solution (0.10, 1.0, and 5.0 mg/L standards prepared in toluene) to the dry test


Page 8 of 30

Page 9 of 30


176

portion producing a final concentration of 10, 100, or 500 μg/kg, respectively. For the 25 µg/kg

177

fortification, 50 µL of a 0.5 mg/L was added. Each centrifuge tube with fortified tea test portion

178

was vigorously vortexed to distribute the pesticides, then allowed to rest for 10 min.

179

In addition, the entire fortification study for black and green teas was repeated three

180

times from the beginning (i.e., samples were spiked, extracted and cleaned). The first

181

replication used cleanup on the Agela “Cleanert” (TPT) SPE columns. The second replication

182

used cleanup on the United Science Dual Phase QuEChERS Carbon X Plus (COA) using COA

183

eluting with acetone. The third replication was identical to the second but eluted the COA

184

columns with ethyl acetate instead of acetone. Fortification results for COA with black tea and

185

ethyl acetate were not reportable because of laboratory quality assurance issues. Each

186

fortification required two types of tea with five replicates at each of the four concentrations

187

from extraction to clean up with 20 SPE columns for each tea type so that in total the

188

fortifications included 160 final fortified extracts, 135 of which were reportable. Blanks are not

189

included in this number. Four blank tea portions were processed with each batch of 20.

190

Quantification was performed by using the peak area ratio responses of the analyte to

191

that of the internal standard, tris-(1,3-dichloroisopropyl)phosphate or tributyl-phosphate-D27

192

and calculating the concentration by preparing a calibration curve, using the peak area ratios of

193

matrix-matched calibration standards to that of the same internal standard. Matrix-matched

194

standards were prepared by extracting tea blanks and fortifying all tea extracts after extraction

195

and cleanup with standards in toluene. Tea standards were prepared at 0.001, 0.002, 0.005,

196

0.010, 0.02, 0.05, 0.1, 0.2, 0.5, 1.0 mg/L for use in constructing matrix matched calibration

197

standards.



Page 10 of 30

198

GC-QQQ Analysis. A TRACE gas chromatograph coupled with a TSQ Quantum triple

199

quadrupole mass spectrometer and a TriPlus auto-sampler (ThermoFisher, San Jose, CA, USA)

200

were employed for sample analysis. Analytes were separated with a 30 M x 0.25 mm ID TG-5MS

201

fused silica capillary column (ThermoFisher, San Jose, CA, USA) preceded by a deactivated guard

202

column (5 M × 0.25 mm I.D, Restek Corp., Bellefonte, PA). The column flow rate was 1.4 ml/min

203

with pressure ramps to maintain a constant flow of Helium. Temperature programming, source,

204

transfer-line temperatures, spiltless injection, multiple reaction monitoring (MRM) optimization

205

is as previously described3. Identifying pesticide residues required the measurement of two

206

precursor/ product ion transitions with the correct ion ratio and retention time during the time-

207

dependent MRM window, as previously described3. We restricted our screen to the 170

208

pesticides we had previously optimized3 that were available in the 12 custom AccuStandard

209

mixtures used in this study. Anthraquinone was weighed out separately, optimized and added

210

to the mixed standards for the GC-QQQ method. A complete list of pesticide common names,

211

CAS numbers, structures, molecular weights, retention times, dynamic range, MS/MS

212

transitions and collision energies are provided in Table 1 of the supporting information3.

213

Statistical Analysis and Calculations. Means and standard deviations from fortification-

214

incurred studies and matrix-dependent instrument quantitation limits (MDIQLs) were

215

determined using Microsoft Excel 2007. Pesticide concentrations were determined by using

216

Thermo Fisher Quanlab Forms 3.0 software to build calibration curves. MDIQLs were estimated

217

for all pesticides through 8 measurements, each from one of a series matrix-matched

218

calibration standards for both black and green tea (2, 5, 10 or 20 μg/L or 8, 20, 40 and 80 μg/kg

219

in the tea) and from using the standard deviation obtained from 8 quantifications multiplied by


Page 11 of 30


220

2.998 (t-test, 1% critical value, degrees of freedom = 7). MDIQLs calculations for each pesticide

221

in both black and green tea were verified by manual inspection to ensure sufficient signal to

222

noise, repeatability, and the presence of both transitions with the correct ion ratio and

223

retention time. Ion ratios were required to be ± 20%, or ± 25% for ions < 50% of the base peak

224

and > 20% of the base peak or ± 30% between 10-20 %, or ± 50% for ions < 10% of base peak3.

225 226

RESULTS AND DISCUSSION

227 228

Fortification of black and green teas: Both black and green teas were fortified with the

229

same mixture of standards containing most of the pesticides previously reported on tea by GC-

230

MS, but not dicofol3,5,6,11,12,18. Table 2 summarizes the mean recoveries and the mean of their

231

relative standard deviations (RSDs) of the pesticides at each concentration for in both green

232

and black teas by SPE column brand. Note the mean recoveries and RSDs in Table 2 reflect the

233

mean for only the recoveries of pesticides measured in common across all three columns.

234

Supporting Tables 2&3 contain means generated from all measured pesticides fortified on the

235

ECPSA and COA SPE columns. Of the 24 overall mean recoveries reported in Table 2, 13 did not

236

change when only pesticides in common to all SPEs were used for the mean rounded to the

237

nearest 1%. Nine others changed less than 2%, and only two at 10 ppb for black tea went down

238

by 4 or 5% (COA and TPT respectively). The corresponding RSDs were unchanged, except four at

239

10 ppb, lower by 1 or 2 percentage points.

240 241

Table 2 also reports the total number of pesticides reported at each concentration by SPE column brand and the number with recoveries between 70-120%. The fraction of pesticides



242

with recoveries within the desired recovery limits varied from 65-95% of the total reported and

243

demonstrated no relationship to SPE column or tea type or the fortifying concentration.

244

Recovery profiles for both green tea and black tea are provided in Figures 1&2, respectively, for

245

all three SPE brands. The profiles are quite similar by concentration and SPE brand.

246

The recovery distributions and RSDs are in general agreement with findings of Pang et al.10

247

except we did not find improved performance with the TPT column with respect to RSDs (Table

248

3). Table 3 gives the percentage of the pesticides with RSDs < 20% or < 10% at 100 or 500 µg/kg

249

spiking concentrations for both green and black teas. The ECPSA and COA columns both

250

demonstrate 97-99% of the pesticide RSDs

Determinations for Pesticides on Black, Green, Oolong, and White

Recommend Documents