Orbital Trapping Mass

Subscriber access provided by Kaohsiung Medical University

New Analytical Methods

Application of a High-Resolution Quadrupole/Orbital Trapping Mass Spectrometer coupled to a Gas Chromatograph for the Determination of Persistent Organic Pollutants in Cow and Human Milk Douglas G. Hayward, Jeffery C Archer, Sue Andrews, Russell Fairchild, James Gentry, Roy Jenkins, Michelle McLain, Udaya Nasini, and Sina Shojaee J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.8b03721 • Publication Date (Web): 15 Oct 2018 Downloaded from http://pubs.acs.org on October 24, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 24

Journal of Agricultural and Food Chemistry

1

Application of a High-Resolution Quadrupole/Orbital Trapping Mass Spectrometer

2

Coupled to a Gas Chromatograph for the Determination of Persistent Organic Pollutants in

3

Cow and Human Milk

4

Douglas G. Hayward*1, Jeffery C. Archer2, Sue Andrews2, Russell D. Fairchild2, James Gentry2,

5

Roy Jenkins†2, Michelle McLain2, Udaya Nasini††2, Sina Shojaee2

6 7 8

1U.

9

2U.

S. Food and Drug Administration, 5001 Campus Dr, HFS-706, College Park, MD 20740. Tel. (240) 402-1654, Fax: (240) 402-2332, E-mail: [email protected] S Food and Drug Administration, 3900 NCTR Dr. Jefferson AR 72079

ACS Paragon Plus Environment


Page 2 of 24

10 11

Abstract: A quadrupole/orbital trapping mass spectrometer or Q-Exactive™ (QE) interfaced with

12

a gas chromatograph (GC) was optimized for measuring polychlorinated dibenzo-p-dioxins,

13

dibenzofurans (PCDD/Fs) and polychlorinated biphenyls (PCBs) in foods. Figures of merit

14

include: (1) an instrument detection limit (IDL) for 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)

15

of 9 femtograms (fg), (2) quantitative mass resolution from PCDD interferences (e.g. PCBs,

16

methoxy-PCBs

17

polychlorodibenzofuran and polychlorodibenzothiophenes) and (3) mass accuracy < 1 ppm at the

18

IDL. The QE measured the concentrations of PCDD/Fs and PCBs in whole cow’s milk with no

19

known source of contamination (e.g. TCDD 33 fg/g fat). A National Institute of Standards and

20

Technology (NIST) unfortified human milk standard reference material (SRM) 1953 was

21

measured determining 27 PCDD/F and PCB congeners with an average difference of 7.6% from

22

the certified results. The QE-GC is a benchtop instrument, easy to service, easy to operate, and

23

requires no lock masses, mass preselection or chemical ionization conditions. The QE-GC

24

demonstrated that it can be an alternative to the double focusing magnetic sector instruments

25

(sector) for the high-resolution measurement of PCDD/Fs and PCBs in dairy products.

26

Keywords: dioxins, PCBs, Orbitrap, human milk, cow milk, HRMS,

27

Introduction: Polychlorinated dibenzo-p-dioxins and dibenzofurans are two classes of related

28

environmental pollutants produced through diverse sources, known to strongly bio-accumulate,

29

and can produce multiple toxic endpoints in animals and humans.1,2 The 2,3,7,8-TCDD isomer has

30

been classified as a group 1A human carcinogen by the International Agency for Research on

31

Cancer (IARC).2 The two classes of chemicals are listed along with other chemical classes as

32

persistent organic pollutants (POPs) by the Stockholm convention. POPs are designated for world-

DDTs,

polychlorodibenzylphenyl

ethers,


polychloroxanthenes,

methyl-

Page 3 of 24


33

wide reduction from the environment1. Every practical means is being employed to limit exposures

34

and reduce the environmental loadings of POPs. The primary route of non-occupational human

35

exposure to PCDD/Fs is due to consumption of animal based foods, often via animal feeds or free

36

ranging practices.3-9 Established monitoring programs measure a wide array of foods and animal

37

feeds to assess efforts to reduce and maintain low concentrations in foods.5 The programs require

38

sensitive and selective mass spectrometry methods for the isomer specific measurement of

39

congener profiles in foods and feeds.5,6,8 Congener pattern recognition can lead to source

40

attribution from the findings in animal foods, animal feed and feed components. For example,

41

Hoogenboom et al.10 found that heat treatment of feed components (sugar beets and maize) used

42

for cow feed constitute a significant source of carry-over to cow’s milk. Continual foods and feeds

43

surveillance has proven to aid the reduction in the contamination of foods.4-10

44

The European Union (EU) has established an exposure limit of 14 pg World Health Organization

45

toxic equivalence (WHO-TEQ)11 per kg body weight per week for the sum of PCDD/Fs and PCBs.

46

Also, EU has suggested maximum levels (MLs) for animal foods such as beef, pork, sheep, dairy

47

(2.5 pg PCDD/F TEQ/g milk fat), chicken and fish.5 The U. S. Environmental Protection Agency

48

(USEPA) has established a reference dose of 0.7 pg PCDD/Fs WHO-TEQ/kg body weight per

49

day.12 Monitoring results from the United States and Canada were recently used to estimate that

50

current exposures to U. S. residents are most likely below this amount.13 Critical to these efforts

51

are mass spectrometry based confirmatory methods capable of detecting these chemicals at the

52

low concentrations required.

53

Mass spectrometry methods using high mass resolution were reported for measuring TCDD as

54

early as 1973 by Baughman and Meselson14 in biota samples from Vietnam and by Crummett and

55

Stehl15 in herbicide mixtures. Mass spectrometry measurements were accomplished using a



56

magnetic sector single or double focusing instrument with off-line14 or online15 low resolution gas

57

chromatographic separation. The limit of detection (LOD) in matrix was in the low picogram range

58

(6 pg) which is sufficient for some types of matrices, but is not low enough for measurement in

59

human serum16 or food.17 By the 1990s, sector instrumental improvements in electronic, source

60

and detector design decreased LODs for TCDD 3 orders of magnitude lower than reported by

61

Baughman and Meselson.13 TCDD measurements were now possible in human serum or foods at

62

environmentally relevant concentrations.16-20

63

Several mass spectrometry techniques were investigated during the early development of TCDD

64

measurement. Most approaches used quadrupole mass spectrometers equipped with positive

65

electron ionization (EI+), negative ion, or atmospheric pressure chemical ionization (APCI) ion

66

sources and sometimes using tandem mass spectrometry. While sensitivity for some applications

67

proved acceptable, they were limited for human serum and food applications. Negative ion using

68

methane enhanced electron capture resulted in good sensitivity for most PCDD/F (100 pg LOD).21,22 Mitchum et al.23 improved TCDD

70

sensitivity with negative ion by using atmospheric pressure ionization with oxygen to produce

71

fragments with greater sensitivity for the 2,3,7,8-TCDD isomer, although less sensitivity than

72

achievable by sector instruments. Early comparisons between triple quadrupoles using MS/MS

73

and sector instruments proved favorable achieving similar or at most 2-fold less sensitivity by

74

MS/MS as reported by McGurvin et al.24 when using a high resolution double focusing magnetic

75

sector and a triple quadrupole mass spectrometer. Tandem mass spectrometry, while very selective

76

and sensitive, typically results in at least 10 times less sensitivity for TCDD than reported with

77

newer sector instruments.17,25-28 In the early 1990s, high resolution mass spectrometry selected ion


Page 4 of 24

Page 5 of 24


78

recording (SIR) using a sector instrument became the method of choice for measurement of

79

PCDD/Fs and was specified in USEPA methods.19,20

80

Recent improvements to triple quadrupole instruments have narrowed the difference in sensitivity

81

with sector instruments such that the EU has established performance criteria for the application

82

of transmission quadrupole MS/MS methods.29-31 These improvements have broadened the

83

application of MS/MS assuming the guidelines for data acceptance are followed and a sector

84

instrument is available for confirmation when necessary.32 Improvements for TCDD sensitivity

85

have also been reported using a GC inlet to an APCI source (atmospheric pressure GC (APGC)

86

interfaced to a triple quadrupole.33,34 The APGC triple quadrupole instrument has demonstrated

87

sensitivities that match a sector instrument for measuring foods.34

88

More recently a GC has been interfaced to an Orbitrap™ mass spectrometer initially described by

89

Makarov35 based on the Kingdom trap design described elsewhere.35,36 A noncommercial version

90

of the Orbitrap™ interfaced to a GC system was described for detection of some organic

91

compounds including PCDD/Fs by Peterson et al.36 suggesting low quantities might be detected.

92

To our knowledge, the commercialized Orbitrap™ GC system has not been described for this

93

application. The goal of this study is to investigate the feasibility of using the QE with an electron

94

ionization (EI) source available on GC-MS systems for the measurement of PCDD/Fs and PCBs

95

in dairy foods.

96

Materials and Methods: PCDD/F and PCB standards were purchased from Wellington

97

Laboratories Inc., Guelph, Ontario, Canada and Cambridge Isotope Laboratories Inc, (CIL),

98

Tewksbury, MA, USA. Mixtures used were NK-LCS-A, DF-ST-B, DF-IS-200, MBP-CP, BP-

99

CP81, BP-MO, MBP-MO. Also, TF-TCDD-MXB, TF-TCDD-MXD mixed TCDD isomers 2-



100

1,000 pg/mL from Wellington Laboratories were used. CIL mixtures were EDF-9999-1-5, EC-

101

4187, EC-4986, CIL-5179, EDF-8999, and EDF-5999. All solvents were high purity HPLC grade

102

or pesticide grade. Toluene, cyclohexane, hexane, (Fisher Scientific, Fair Lane, NJ),

103

dichloromethane and ethanol (Omnisolv EMD Millipore Corporation, Billerica, MA, USA),

104

diethyl ether (ACROS Chemicals Geel, Belgium). Nonane and silica gel 60 70-230 mesh were

105

from Sigma-Aldrich, St Louis, MO, USA and Alumina B Activity Super I for dioxin analysis MP

106

EcoChrom™ was from MP Biomedicals, Santa Ana CA, USA.

107

Two PCDD/F calibration curves were constructed. One curve was prepared using a Wellington

108

Laboratories mixture of all 17 2,3,7,8-13C12 labeled PCDD/F (NK-LCS-A, mass labeled PCDD/Fs)

109

(used with Maryland milks analyses) mixed with all calibration standards to give 10 ng/mL for all

110

congeners (OCDD 20 ng/mL). The 17 2,3,7,8-substituted PCDD/Fs (Wellington Laboratories, DF-

111

ST-B) were mixed to provide 0.1, 0.2, 1, 5, 10 or 20 ng/mL for TCDD/F, PeCDD/Fs (Pentachloro-)

112

and 0.2, 0.4, 2, 10, 20 or 40 ng/mL for HxCDD/Fs (Hexachloro-), HpCDD/Fs(Heptachloro-) with

113

OCDD/F at 0.5, 1, 5, 25 50 or 100 ng/mL. Also added were PCBs 77, 81, 126 and 169 (1, 2, 5, 20,

114

100 or 1000 ng/mL) along with 13C12-PCBs 77, 81, 126 and 169 (10 ng/mL). The second curve

115

used CIL mixtures for PCDD/Fs, EDF-9999-1-5, diluting these 1 to 10 (used for Arkansas milks

116

and NIST SRM 1953). PCBs 77, 81, 126 and 169 were added during dilution to give x = 0.25, 1,

117

5, 20 or 100 ng/mL the same as PCDD/Fs except TCDD/F (1/5x) and OCDD/F (2x).

118 119

Milk collection: Whole cow’s milk was purchased at retail, collected from 3 brands in Arkansas

120

(AR), USA and 3 brands in Maryland (MD), USA. Human milk was purchased from NIST,

121

Gaithersburg, MD, USA. The human milk SRM number 1953 was composed of pooled unfortified


Page 6 of 24

Page 7 of 24


122

human milk collected from several states in 2004. All cow’s milk brands and NIST human milks

123

were kept frozen at -20 oC prior to analysis.

124

Milk Preparation: Arkansas milks (90-100 g aliquots) were extracted and cleaned up as described

125

elsewhere.37 Before extraction, 15

126

(Cambridge Isotope Laboratories) were added at 100 pg each. Maryland milks (100 g aliquots)

127

were extracted and cleaned up as previously described.17,22 Before extraction of Maryland milks,

128

all 17 2,3,7,8-substituted 13C12-PCDD/Fs and 13C12-PCBs 77, 81, 126 and 169 were added at 50

129

pg for the PCDD/Fs and 125 pg for the PCBs. For the NIST human milk 14 additional 13C12-PCB

130

congeners (coplanar + marker) were added 200 pg each. The PCDD/Fs containing fractions were

131

reconstituted in 10 µL of nonane. Concentrations were lipid adjusted using the fat determined in

132

each cow’s milk brand for Maryland milks. Fat was determined gravimetrically after extraction of

133

crude fat.22

134

Safety: PCDD/Fs and PCBs are highly toxic, carcinogenic and should be handled to avoid skin

135

contact and inhalation of dust or aerosols. Internal standard amounts in milk were 3x the daily

136

allowable intake from food.12

137

Sector Mass Spectrometry: Arkansas milk extracts were measured using a Waters Technologies

138

Corporation, Milford MA, AutoSpec Premier™ interfaced to an Agilent 7890A GC with a 60 m x

139

0.25 mm ID DB-5 ms column equipped with a 4 mm ID split/splitless liner. One µL splitless injections

140

were made at 280 oC onto the 60 DB-5 ms column with 1.0 mL/min He constant flow. Mass ions were

141

acquired in SIR lock mass mode at >10,000 resolution following USEPA 1613.20

142

Orbital Trapping Mass Spectrometry: All Cow and human milk extracts were measured using

143

a TRACE 1300 GC interfaced to a Q-Exactive™ (Orbitrap™) mass spectrometer (Thermo Fisher

13C

12-

PCDD/Fs and

13C


12-PCBs

77, 81, 126 and 169


144

Scientific, San Jose, CA). A 40 m x 0.18 mm ID DB-5 ms column (Agilent Santa Clara, CA) was

145

used with 1.5 mL/min constant flow and with an injector equipped with a split/splitless, 4 mm ID

146

Restek (Bellefonte, PA) sky liner. Two µL splitless injections were made using a pressure pulse

147

of 70 PSIG at 300o C. GC was programmed as previously described.16 The QE-GC was operated

148

in full MS SIM mode with 6 quadrupole filter settings for PCDD/Fs and two for other PCBs.

149

Settings used were 285-340, 300-340, 335-354, 335-380, 369-410, 435-480 m/z for PCDD/Fs

150

(PCBs 77, 81, 126, 169) and 250-350, 300-410 for 14 other PCBs using the same GC column and

151

temperature program. An offset of -2 V was set between the source and the C-Trap. Other electron

152

ionization parameters were default and AGC was used with a target of 1E6. The resolution at full

153

width half maximum height (RFWHM) was set to 120,000. No lock masses were used. Transfer

154

line and source temperatures were 250 oC and 300 oC, respectively.

155

Quality Assurance: The Q-Exactive™ was calibrated daily using perfluorinated tributyl amine

156

(FC-43). All milk measurements were completed in duplicate on separate days. Unfortified milk

157

amounts were subtracted from fortified milks.

158

TraceFinder™ version 3.3 using the Wellington or CIL calibration curves. IDLs were estimated

159

from the standard deviation of 9 injections using the lowest calibrated level (LCL) (Wellington

160

standard curve).

161

Results and Discussion

162

QE optimization for PCDD/Fs: The first part of the study focused on following the work done

163

by Peterson et al. 2010.36 Progressively smaller amounts of PCDD/F standards were measured

164

with the QE using Full MS-SIM mode with a quadrupole filter set to m/z 235-480 to the

165

Orbitrap™. Manufacturer supplied default tuning and calibration settings were used as a starting

Q-Exactive™ files were processed in


Page 8 of 24

Page 9 of 24


166

point and resulted in unacceptable sensitivity. TCDD signals were undetectable at slightly less

167

than 100 femtograms on column (data not shown). An examination of parameters that might affect

168

response was conducted.38 We investigated electron energy, emission current, source temperature,

169

tune parameters, resolving power, acquisition type (Full MS-SIM, Targeted-SIM, parallel reaction

170

monitoring) and C-Trap offset.

171

Five parameters were helpful for maximizing TCDD response. They were in order of

172

effectiveness: (1) tuning/source cleaning, (2) resolving power, (3) source temperature, (4) air

173

background and (5) C-Trap offset. The QE-GC allows an operator to use one of four resolving

174

power settings which are RFWHM = 15,000, 30,000, 60,000, 120,000. Figure 1 provides an

175

example of TCDD isomer responses as a function of resolving power. The area for the 2,3,7,8-

176

TCDD isomer M+ ion at 250 fg on column mass is the same (55-63K) at all four resolution settings.

177

As the on-column mass of TCDDs is lowered to 10 fg m/z 320 is no longer detected at 15,000 and

178

30,000 settings (Figure 1). Source temperature optimized at 300 oC for PCDD/Fs. A modest

179

increase (< 2 times) in signal intensity for TCDD was observable as the temperature was raised

180

from 250 oC to 300 oC, but no further increase was observed up to the maximum, 350 oC. Source

181

cleaning and tuning have impacts on sensitivity. While the repeller voltage clearly has the main

182

influence on the TIC of the calibration compound, there does not seem to be an exact 1 to 1

183

correlation between FC43 TIC and sensitivity. Air background arises almost entirely from the

184

connection of the column to the interface as is typical in these systems and this connection must

185

be done properly or response may be greatly affected.



186 187 188 189 190

Figure 1. Extracted ion chromatograms from an isomer mixture (4 injections superimposed of TFTCDD-MXD); area responses for TCDDs 1,3,7,8; 1,4,7,8; 1,2,3,4; 2,3,7,8 at 10, 25, 100 and 250 fg/µL, respectively, (1 µL injections) with resolution settings of 15,000, 30,000, 60,000 and 120,000.

191

Among the tuning parameters there is an offset for the C-Trap. The C-Trap stores ions temporarily

192

before introduction into the Orbitrap™ detector. The C-Trap offset is set to zero volts when tuning.

193

We found that setting a -2 V offset improved response for TCDD by 50%. Setting the C-Trap

194

offset up to the maximum (-4 V) had no added benefit. The C-Trap offset creates fragmentation

195

between the source and the C-Trap evident by changes in the ion intensities of FC-43. Lower

196

molecular ion response was not observed with a -2 V offset, but rather the ion intensity increased.

197

When tuning the repeller voltage produces the largest effect on the TIC of FC-43 and is always

198

lowered from the default.


Page 10 of 24

Page 11 of 24


199 200 201 202

Figure 2. Extracted ion chromatogram for TCDD isomers 1,3,6,8; 1,3,7,9; 1,3,7,8; 1,4,7,8; 1,2,3,4; 2,3,7,8; at 2, 5, 10, 25, 50 or 100 fg/µL, respectively (1 µL injected of TF-TCDD-MXB on a 40 M x 0.18 mm id DB-5MS column). Accurate masses with an error range from +0.21 to -1.3 ppm.

203 204

Figure 2 illustrates an optimized response from a series of TCDD isomers showing the extracted

205

ion chromatogram for the ion M+ m/z 319.89599 at concentrations from 2-100 fg/µL. The

206

conditions were a -2 V offset to the C-Trap and low oxygen background ≤ 2 % of m/z 69 calibration

207

ion. An ion was detected at each of the six TCDD isomer concentrations and each had the correct

208

mass within < 1 ppm, except at 5 fg (-1.3 ppm). The chlorine isotope ratios are within ± 15% of

209

the theoretical, except at 2 and 10 fg.

210

QE stability: Confirmatory methods should be reliable when measuring concentrations of

211

PCDD/Fs in foods. The PCDD/F identifications require exact mass within ±5 ppm, correct ion

212

ratio ±15% of theoretical19 and preferably during 100% of the measurement attempts. The QE-GC

213

was tested for reliability by measuring the LCL and ½ the LCL (Wellington Laboratories curve)



214

with six sequential injections. The determinations at ½ LCL resulted in mass errors that were less

215

< 1 ppm for the 17 PCDD/F congeners in 98% of the measurements.

216

Table 1. Mean measurements by QE chlorinated dibenzo-p-dioxins/furans (CDD/Fs) of ½ LCL (n = 6) 1 µL

217

injections Wellington Laboratory standard, 0.05, 0.1 or 0.25 ng/mL. Mean accurate mass; absolute Δ m (ppm);

218

isotope ratio (M+2/M+4) (red > ±15% theoretical); % ratios meeting criteria; IDLs (n=9) and LODs in milk fat.

219 mean

mean mean

%

congener

Observed

Δm

Ratio

met

IDL LODs** pg/g fg fat

2,3,7,8-CDD*

321.89287 355.85394 389.81501 389.81490 389.81499 423.77641 459.73430 305.89803 339.85901 339.85898 373.81993 373.81995 373.82002 373.82003 407.78111 407.78122 443.73903

0.52 0.43 0.24 0.5 0.27 0.68 0.23 0.33 0.43 0.49 0.66 0.63 0.45 0.41 0.31 0.25 0.58

0.76 0.48 1.28 1.26 1.33 0.99 0.90 0.78 1.64 1.66 1.27 1.26 1.30 1.31 1.11 1.06 0.93

100 17 83 100 100 100 100 100 100 67 100 100 83 100 100 83 83

9 10 37 29 12 41 55 10 9 10 13 17 9 17 23 26 47

0.014 0.015 0.056 0.044 0.018 0.062 0.083 0.015 0.014 0.015 0.020 0.026 0.014 0.026 0.035 0.039 0.071

220

1,2,3,7,8-CDD 1,2,3,4,7,8-CDD 1,2,3,6,7,8-CDD 1,2,3,7,8,9-CDD 1,2,3,4,6,7,8CDD OCDD 2,3,7,8-CDF* 1,2,3,7,8-CDF 2,3,4,7,8-CDF 1,2,3,4,7,8-CDF 1,2,3,6,7,8-CDF 2,3,4,6,7,8-CDF 1,2,3,7,8,9-CDF 1,2,3,4,6,7,8-CDF 1,2,3,4,7,8,9-CDF OCDF *(M/M+2)

221

**LODs limits of detection estimates based on the milk fat injected and the IDL.

222 223

Each determination resulted in a congener ion ratio measurement within ± 15% of theoretical20 or

224

nearly so (83 or 100%), with some exceptions. Only 17% of ion ratio measurements for PeCDD

225

were within ±15% (Table 1) with the mean ion ratio low by 20%. The fraction correctly measured

226

increased to 83% at the LCL. Low PeCDD ion ratios were unrelated to a potential mass inference


Page 12 of 24

Page 13 of 24


227

from PCB 169 co-eluting PeCDD with a common nominal mass of 358 specified for PeCDD in

228

USEPA methods20 (Figure S-1). The 2,3,4,7,8-CDF had acceptable ratios for 67% of the

229

determinations and the fraction increased to 83% at the LCL. The IDLs were as low as 9 fg for

230

TCDD to as much as 55 fg for OCDD.

231

Verification in Cow’s Milk: The U. S. FDA Arkansas Laboratory conducted a verification

232

study.35 Two aliquots from each brand of Arkansas milk were fortified with TCDD/Fs (0.8 pg),

233

PeCDD/Fs/HxCDD/Fs/HpCDD/Fs (4 pg) and OCDD/F (8 pg). Six more aliquots were also

234

fortified at 2.4, 12 and 24 pg (Figure S-2). Fortified and unfortified milk extracts were then

235

shipped to the U. S. FDA laboratory in College Park, MD after measurement using the sector in

236

the Arkansas Laboratory.

237 238 239 240

Figure 3. Mean amounts in pg measured in fortified cow’s milk, standard deviations indicated by error bars (n=6) corrected for the unfortified milk, plotted on a log scale. TCDD and TCDF were 0.68 pg, 0.65 pg by QE and 0.86, 0.83 pg by sector, respectively.

241

The QE and sector agreed quite well conforming to accepted criteria.18 Mean measurements from

242

the sector were higher than the predicted by 2-16% in all, but one congener (HpCDD) at the lower

243

fortification (Figure 3). Mean measurements from QE were evenly distributed compared to



244

predicted by 1-19% higher or lower (Figure 3). The QE mean for PCDD/F TEQ (9.35 pg, sd 0.43)

245

was closer to the predicted (9.13 pg) than the sector (9.82 pg, sd 0.27) 7.6% high. The means

246

measured at the higher fortification were all within 0.1-14% and 0.6-11% of the predicted for

247

sector and QE, respectively (Figure S-2). The PCDD/F TEQs measured by sector and QE were

248

only 0.3% different (28.14 pg, sd 0.54 and 28.24 pg, sd 0.58 respectively) and both instruments

249

were 8% higher than predicted. Nearly all fortified measurements were within ±20% of the

250

predicted (98% sector, 95% QE).

251 252

Cow’s Milk Measurements: PCDD/Fs and PCBs are environmental contaminants with diverse

253

sources and therefore detectable concentrations are to be expected. Even as sources become

254

controlled there will still be reservoir sources from past emissions and natural reservoirs.4 Mass

255

spectrometry based methods are required to measure the concentrations in dairy with no known

256

source of contamination so that emerging contamination can be recognized. On-going exposure

257

assessments are enhanced by having an accurate measure of current dairy concentrations. Three

258

brands of Maryland milk were measured 2 times each and one 3 times for a total of 7

259

measurements. Table 2 provides the mean values across the 3 milk brands collected in Maryland.

260

The QE reported 9 congeners with 1-7 determinations not detected or with incorrect isotope ratios.

261

The QE reported no 1,2,3,7,8,9-HxCDF ion responses in any milk. The important question

262

addressed in Table 2 is whether the results would provide a sufficient measure of milk with no

263

known source of contamination. Table 2 results suggest that for dairy products this is true. For

264

example, the two MD milks listed separately resulted in a < 20% difference between the upper

265 266

Table 2. Measurements for cow’s milk collected in Maryland (MD) by QE. Means for three milk sources; # ND = number not detected; pg/g fat, blank corrected, TEQs upper bound (ND = LOD).


Page 14 of 24

Page 15 of 24


n=1a

n=1a

n=7

#

MD

MD

MD

ND

0.019b

0.033

0.034

3b

1,2,3,7,8-CDD

0.09

0.15

0.16

1b

1,2,3,4,7,8-CDD

0.075

0.14

0.14

0

1,2,3,6,7,8-CDD

0.22

0.37

0.38

0

1,2,3,7,8,9-CDD

0.094

0.16

0.17

0

1,2,3,4,6,7,8-CDD

0.61

1.61

1.1

1b

OCDD

1.4

7.02

3.1

0

2,3,7,8-CDF

Orbital Trapping Mass

Recommend Documents