Hg compound-specific isotope analysis at ultra-trace levels using an

Subscriber access provided by Kaohsiung Medical University

Technical Note

Hg compound-specific isotope analysis at ultra-trace levels using an on line gas chromatographic pre-concentration and separation strategy coupled to multicollector-ICP-MS Sylvain Bouchet, Sylvain Berail, and David Amouroux Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b04555 • Publication Date (Web): 07 Jun 2018 Downloaded from http://pubs.acs.org on June 8, 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 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

1

Hg compound-specific isotope analysis at ultra-trace levels using an on line

2

gas chromatographic pre-concentration and separation strategy coupled to

3

multicollector-ICP-MS

4 5

Sylvain Bouchet†*, Sylvain Bérail and David Amouroux

6 7

CNRS / Univ Pau & Pays Adour, Institut des sciences analytiques et de physico-chimie pour

8

l’environnement et les matériaux, UMR5254, 64000, Pau, France

9 10

†

Present address: ETH Zürich, D-USYS department, Universitätstrasse 16, CH-8092 Zürich, Switzerland

11 12 13

*

Corresponding author: [email protected]; +41 58 765 5461

14 15

1 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 18

16

ABSTRACT

17

Stable Hg isotope analyses are nowadays widely employed to discriminate Hg sources and understand its

18

biogeochemical cycle. Up to now, total Hg isotopic compositions have been mainly used but Hg compound-

19

specific isotopic analysis (CSIA) methodologies are emerging. On-line Hg-CSIA were limited to samples

20

containing high concentrations but in this work we overcome this limitation for the measurement of inorganic (IHg)

21

and monomethylmercury (MMHg) by gas chromatography hyphenated to MC-ICP-MS (GC/MC-ICP-MS) through

22

the use of an automated on-line pre-concentration strategy, allowing injection volumes up to 100 times larger than

23

usual. The pre-concentration of Hg species and subsequent transfer to the column were achieved by a

24

programmed temperature vaporization (PTV) injector fitted with a packed liner. The PTV parameters were first

25

optimized using a quadrupole ICP-MS and then its suitability for Hg-CSIA was evaluated with long-term replicate

26

analysis of various standards and reference materials (RMs). The large preconcentration capability enables

27

analyses with Hg concentrations in the organic solvent two orders of magnitude lower than the previous

28

conventional GC/MC-ICP-MS method but a compound specific standard bracketing procedure was required for

29

MMHg in order to correct for the differential behavior of Hg species in the liner. The external reproducibility of the

30

method ranged from 0.19 to 0.39 ‰ for ∆199Hg and δ202Hg (as 2 SD, n = 143-167) depending on the species. The

31

analysis of various RMs demonstrated the applicability to environmental samples with species concentrations

32

down to about 150 ng.g-1. This new methodology opens the way for a much wider range of on-line Hg-CSIA

33

measurements that will improve our understanding of the Hg biogeochemical cycle.

34


Page 3 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


35

INTRODUCTION

36

The fate and impact of Hg in the environment is heavily driven by its chemical speciation, i.e. the particular forms

37

under which Hg exists. This has boosted the development of many speciation methodologies over the last 4

38

decades but Hg isotopic composition measurements have recently emerged as a major tool to further elucidate

39

sources and transformations of Hg1. When Hg signatures between sources and receiving ecosystems are

40

contrasted, it allows a clearer source identification and apportionment as well as a better understanding of Hg

41

dispersion and bioaccumulation in ecosystems2,3. Hg mass dependent (MDF) and mass independent (MIF)

42

isotope fractionation arise from virtually all Hg species transformations, whether biotic or abiotic and dark or

43

photochemically induced1. Large Hg MDF may originate from various biotic/abiotic reactions whereas only

44

photochemical processes have been shown to induce large Hg MIF for both IHg and MMHg until now4,5 and these

45

species-specific signatures are then recorded in compartments where Hg accumulates such as aquatic biota.

46

Measurements of Hg isotopic composition in environmental samples until now have been performed mostly on

47

total Hg (HgT) because sensitivity and transient signals are major issues with MC-ICP-MS instruments6–8.

48

Continuous flow Cold Vapor Generation (CCVG) has been the most widely used introduction system for 15

49

years9,10 but this introduction technique is actually restricted to samples with Hg concentrations above 1 µg.L-1 in

50

solution. Therefore various off-line pre-concentration techniques for HgT have been developed to overcome this

51

limitatione.g. 11–13 and more recently also an online one using cold vapor generation and dual gold-amalgamation

52

(CVG-DGA/MC-ICP-MS)14 allowing analyses at the ng.L-1 level. However, the HgT isotopic composition

53

represents the weighted average isotopic composition of individual Hg species present in the sample and

54

important information can be retrieved from the species own isotopic signatures as already demonstrated in the

55

study of natural biogeochemical cycles of lighter elementse.g. 15 or degradation of micropollutantse.g. 16,17. For Hg, it

56

has been shown to be a powerful tool to track the metabolic pathways of Hg detoxification in top predators18 but it

57

may also be used to better understand the uptake and bioamplification in the lower food webs levels where

58

organisms have a more balanced Hg speciation. Off-line selective extraction methods (SEM) for MMHg19,20 have

59

been developed to overcome this problem but they remain time and labor consuming as well as limited to provide

60

information on other species21. On the other hand, on-line Hg CSIA with GC/MC-ICP-MS have been

61

developed22,23 and the recent implementation of new data treatment strategy adapted to transient signals24 have

62

brought the uncertainties associated to such measurements down to reasonable levels (2 SD within 0.2-0.5 ‰) 3 ACS Paragon Plus Environment


Page 4 of 18

63

compared to the usual 4-5 ‰ variations range observed in the environment1. Despite this methodological

64

improvement, on-line Hg CSIA have been so far limited to laboratory experiments25–28 or naturally concentrated

65

samples, i.e. hair24, tissues and organs from humans, fish and aquatic mammals18 containing ppm levels of Hg

66

due to the inherent MC-ICP-MS sensitivity limitation.

67

A critical point for detection limits (DLs) when using GC after derivatization and liquid-liquid extraction lies in the

68

limited amount of solvent, typically 1-2 µL that can be accommodated by isothermal split/splitless inlets. A few

69

Large Volume Injection (LVI) techniques do however exist to alleviate this problem among which the Programmed

70

Temperature Vaporization (PTV) injection is the most popular since it is an easy, flexible and cost effective

71

technique to implement while being also rugged for dirty samples29. PTV injection actually comprises several sub-

72

techniques that differ mainly in the kind of liner used (straight or baffled, packed or empty) and the sample

73

introduction method (‘at-once’, multiple or speed-controlled injection). On one hand, packed liners remove the

74

need for cryo-cooling and the maximum volume injectable ‘at-once’ is larger but risks of analyte degradation,

75

irreversible adsorption or poor transfer to the column are also higher than with empty ones29. The use of PTV

76

injection is common for the quantification and CSIA of trace organicse.g. 29,30 but there have been relatively fewer

77

applications for metal species, mainly for Sn and Pbe.g. 31,32 and none for Hg isotopic analysis until now.

78

In this work, we relieved the sensitivity limitation of the GC/MC-ICP-MS through the use of a LVI injection

79

technique. We selected a polymeric resin as a packing material and first optimized its amount and the PTV

80

injector parameters in order to obtain the best Hg species preconcentration and transfer to the GC column. Then,

81

we evaluated the suitability of the technique for Hg species isotopic composition measurements using long-term

82

standard analyses and environmental Reference Materials (RMs) representative of various environmental

83

matrices and Hg concentrations. To the best of our knowledge, this study is the first to report on the development

84

and potential applications of the PTV injection technique hyphenated with GC/MC-ICP-MS analysis.

85


Page 5 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


86

EXPERIMENTAL SECTION

87

Chemicals, standards and samples preparation.

88

All solutions were prepared using ultrapure water (18 MΩ cm, Millipore). Chemicals were at least of analytical

89

grade and used without any further purification. High-purity HNO3 and HCl were from Fisher chemical (Optima or

90

Trace Metals grade); ammonium acetate, isooctane, methanol and hexane were from Sigma Aldrich, acetic acid

91

and ammonium hydroxide from J.T. Baker (Phillipsburg, NJ, USA). Sodium tetrapropylborate (NaBPr4, purity ≥

92

98%) were purchased from Galab (Geesthacht, Germany). Working solutions of Tl were prepared by dilution of a

93

stock solution of the standard reference material (SRM) NIST SRM-997 (Thallium Isotopic, NIST, Gaithersburg,

94

MD). NIST SRM-3133 and 8610 (formerly known as UM-Almaden) were also purchased from NIST and

95

standards were prepared daily from the stock solutions by dilution in 1% HCl. (Note that Hg is a potent toxic that

96

must be handled with appropriate protections under a fume hood).

97

Reference Materials (Table SI-1) were extracted using a focused microwave oven (Discover, CEM corporation,

98

Mathews, NC, USA) by a fixed temperature method33: 80°C during 4 min for 200 mg of CRM in 5 mL HNO3 (6 M,

99

Trace Metal Grade). Extracts were then centrifuged at 4000 rpm for 5 min to recover supernatants. All samples

100

and standards, were derivatized by propylation that results in heavier species compared to ethylation as follow:

101

first, the pH was adjusted to 4 ± 0.1 by addition of 5 mL of an acetate buffer (0.5 M), then appropriate amounts of

102

sodium tetrapropylborate (NaBPr4) and hexane were added and derivatized Hg species are finally recovered in

103

hexane after vigorous shaking. Hg species concentrations in RM extracts were determined by isotope dilution34

104

using 199IHg and 201MMHg enriched stable isotopes (ISC Science, Oviedo, Spain) both with and without PTV.

105 106

Instrumental setups.

107

PTV and GC setups. Straight liners (2 mm ID, 2.75 mm OD, 120 mm length, ThermoFisher Scientific, France)

108

were packed with a bulk sorbent (Bondesil-ENV, 125 µm, Agilent Technologies) made of a styrene-

109

divinylbenzene (SDVB) polymeric resin. Selection criteria for this material were a large specific area (500 m2.g-1)

110

with an acceptable thermo-stability (maximum temperature of 245°C, above boiling points for Hg species). A

111

recent work35 demonstrated its better capacity to retain alkylated Hg species compared to other common

112

commercial sorbents, such as Tenax® or Carbotrap®. Two tight glass wool plugs of about 0.5 cm each (Supelco,

113

Sigma Aldrich) were used to maintain it roughly centered in the liner. Liners were prepared with 20 to 30 mg of 5 ACS Paragon Plus Environment


Page 6 of 18

114

sorbent and pre-conditioned (in the GC, without the column connected) under an inert atmosphere (He, 20

115

ml.min-1) as follow before use: a first flushing step to eliminate O2 (15 min), then a step by step incremental

116

heating phase (50°C per step for 15 min each) from room temperature to 245 °C and eventually five high-speed

117

cycles from room temperature to the maximum temperature were performed to complete the process.

118

A Trace Ultra GC (ThermoFisher Scientific, France) offering both regular split/splitless and PTV injection ports

119

(BEST PTV InjectorTM) was used in combination with a Triplus RSH autosampler fitted with either 10 or 100 µL

120

syringes (Hamilton, Switzerland) and a MXT-1 capillary column (0.53 mm ID, 1 µm thick coating, 30 m length,

121

Restek, France). Hyphenation to either the quadrupole ICP-MS instrument34 (X7 Series II, ThermoFisher

122

Scientific, France) or the multi-collector ICP-MS instrument (Figure SI-1) was achieved through a commercial

123

heated interface (ThermoFisher Scientific, France). The operating conditions for the PTV inlet, GC and ICP-MS

124

instruments are given in Table 1.

125 126

Multi-collector ICP-MS. A Nu Plasma HR (Nu instruments, UK) has been used throughout this work. The GC was

127

interfaced to the MC-ICP-MS through a commercial dual inlet glass torch as described previously22,24 in order to

128

run the instrument under wet plasma conditions as recommended for such applications23. This set up allows for

129

the simultaneous analysis of an isotopically certified Tl solution (200 µg.L-1 in 2% HNO3) to correct for

130

instrumental mass bias. This solution is continuously introduced in the second entry of the plasma torch via a 200

131

µL.min-1 micro-concentric nebulizer and a cyclonic spray chamber (cinnabar, Glass Expansion). All signals were

132

acquired using the Time Resolved Analysis (TRA) mode of the instrument with an integration time of 0.5 s.

133

Analysis followed a Sample Standard Bracketing (SSB) sequence were the SRM NIST SRM-3133 (IHg) and

134

STREM (MMHg) were used as primary standards and matched to the sample concentrations within 25%24.

135 136

Peak areas, isotopic ratios and delta value calculations

137

To determine the species-specific response of the PTV/GC-MC-ICP-MS, peak areas were integrated for the 202Hg

138

isotope (V202Hg.s) using the software AZUR edited by Datalys (France). For the calculation of isotopic ratios, the

139

Linear Regression Slope (LRS) method7,24,36 was used with an integration windows of 30 s, usually centered on

140

the peak apex or slightly shifted to the right when the peak tailing was too pronounced. The instrumental mass

141

bias was corrected using the measured

205/203Tl

and an exponential law as previously described24. Hg isotopic 6

ACS Paragon Plus Environment

Page 7 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


142

compositions are commonly reported as delta values relative to the NIST SRM-3133 IHg standard10. For IHg, as

143

the species in the sample is the same as the reference standard, deltas were calculated as followed

144

δ

xxx

 IHg (‰) =  

(

(

xxx

xxx

Hg / 198Hg

Hg / 198Hg

)

)

IHg Sample

IHg NIST 3133

 − 1 × 1000 

(1)

145

Where xxx can be 204, 202, 201, 200 or 199 and (xxxHg/198Hg)IHg NIST SRM-3133 is the averaged isotopic ratio of the

146

two bracketing standards.

147

For MMHg, two different delta processing approaches were tested: in the first one called Compound Unspecific

148

Bracketing (CUB, equation 2) the MMHg delta value for the sample is calculated directly against the NIST SRM-

149

3133 IHg standard while in the second one called Compound Specific Bracketing (CSB, equation 3), the MMHg

150

delta value for the sample is calculated against the STREM MMHg standard and subsequently converted

151

relatively to the NIST SRM-3133 as proposed by Epov et al.24:

152

153

154

δ

xxx

 MMHgCUB (‰) =  

( (

xxx xxx

) Hg )

Hg / 198Hg

MMHg Sample

Hg / 198

IHg NIST 3133

 − 1 × 1000 

(2)

 δ xxx (STREMvs NIST3133)CCVG   δ xxx (samplevs STREM)   δ MMHgCSB (‰) =  + 1 ×  + 1 − 1 ×1000 (3) 1000 1000      xxx

155 156

where δxxx (STREM vs NIST SRM-3133)CCVG is the previously reported isotopic composition of the STREM MMHg

157

standard relative to the NIST SRM-3133 measured by CCVG / MC-ICP-MS24 and δxxx (sample vs STREM) is the

158

isotopic composition of the sample MMHg versus the STREM MMHg standard calculated as follows:

159

δ

xxx

 Sample vs STREM (‰) =  

( (

xxx xxx

) Hg )

Hg / 198Hg

MMHg Sample

Hg / 198

MMHg STREM

 − 1 × 1000 

(4)

160

The ∆ notation is used to express the mass independent fractionation (MIF), calculated as ∆xxxHg = δxxxHg-

161

βkin×δ202Hg where βkin = ln(m198/mxxx)/ln(m198/m202)10. The HgT isotopic composition was calculated by weighing

162

the isotopic composition of each Hg species by its contribution to the total Hg concentration18. Following previous

163

recommendations10, the external reproducibility of the method is reported as 2 SD of secondary standards



Page 8 of 18

164

measurements while uncertainties on unknown samples are reported as 2 SE. Statistical analyses were

165

conducted with Origin (OriginLab).


Page 9 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


166

RESULTS AND DISCUSSION

167

On-line PTV pre-concentration for isotopic ratio measurements by GC/MC-ICP-MS.

168

Transient signal properties. A typical chromatogram obtained for a mixed standard containing IHg NIST SRM-

169

3133 and MMHg STREM is given in Figure SI-5a. The pattern of the Tl signal is very similar to what has been

170

observed with GC/Q-ICP-MS under the same conditions (Figure SI-6) with multiple perturbations in the beginning

171

(< 90 s) when the successive “waves” of solvent generated by the PTV operation reach the plasma. It is however

172

stable during Hg species elution allowing accurate

173

corrections. Both Hg peaks are fronting but it is more pronounced for MMHg, likely because it is less retained

174

during the refocusing step. The MMHg peak is broader, lasting 15 s at the base compared to 11 s for IHg, and

175

therefore different concentrations of MMHg and IHg were needed to reach similar 202Hg peak height, 52 and 24

176

µg.L-1 to generate 5.3 and 5.4 V, respectively.

177

The time courses of isotopic ratios along the species elution are not consistent with each other, being either

178

stable, drifting upwards or downwards (Figure SI-5 b and c) as previously seen with conventional GC/MC-ICP-

179

MS21 or other transient introduction modes7,37. Over 565 analyses (mass injected between 330 and 1560 pg Hg)

180

the internal precision (SDint) of the

181

(corresponding to a RSD of 47 ± 15 ppm) while for IHg it averaged 1.4.10-4 ± 0.4.10-4 (RSD of 43 ± 9 ppm). It is

182

comparable to values obtained with the same GC without PTV (1.4.10-4 ± 0.7.10-4, n = 46, data not shown).

183

Despite similar ranges and average values, the internal precisions for IHg and MMHg are not correlated (r² = 10-5,

184

data not shown) demonstrating that they are not affected by the same factors. Nevertheless, they both show

185

negative trends with the injected Hg mass and peak areas as predicted by the counting statistics6,14,24 (Figure SI-

186

7).

202/198Hg

205/203Tl

isotopic ratio measurements and mass bias

ratio for the STREM MMHg standard averaged 1.3.10-4 ± 0.3.10-4

187 188

Isotopic composition accuracy and external reproducibility. Figure SI-8 presents delta values for standards NIST

189

SRM-8610 and STREM versus NIST SRM-3133 as a function of the injected volume and thereby of the Hg

190

concentration in the organic solvent (Table SI-2). In all cases, the average δ202Hg and ∆199Hg values are

191

consistent with published values (dashed lines) considering the associated uncertainties. There are no trends

192

observable with the injected volume, neither for the average values nor the associated uncertainties and

193

measurements are therefore possible down to 5 µg.L-1 of Hg in the solvent, 2 orders of magnitude lower than 9 ACS Paragon Plus Environment


Page 10 of 18

194

before24. It corresponds to Hg species concentrations of 25 ng.L-1 in aqueous samples assuming the pre-

195

concentration of 100 mL in 500 µL of solvent while, assuming the extraction of 200 mg of sample in 5 mL of

196

extractant, the theoretical Hg species concentration measurable in a solid sample is about 125 ng.g-1. For

197

comparison, previous measurements by conventional GC/MC-ICP-MS down to about 300 ng.g-1 in solid samples

198

could be achieved by increasing the solid/extractant ratio combined with a tedious offline pre-concentration under

199

Ar stream18.

200

Figure 1 presents the distribution of δ202Hg and ∆199Hg values of the NIST SRM-8610 and STREM standards for

201

a large number of measurements (n = 167 and 143, respectively) performed with Hg concentrations from 5 to 100

202

µg.L-1 and spread over 3 different sessions and 2 years (all other values in Figures SI-9). For the NIST SRM-

203

8610, both the δ202Hg and ∆199Hg mean values calculated directly against the NIST SRM-3133 (-0.56 and -0.01

204

‰, respectively) are accurate compared to values obtained by cold vapor generation. For the STREM standard

205

however, the mean δ202Hg and ∆199Hg values are respectively 0.33 and 0.23 ‰ lower than the reference ones

206

when directly calculated versus the NIST SRM-3133 (CUB) but match the reference ones (one sample t-test, p >

207

0.01) and are also less scattered when the CSB is performed. The external reproducibility for MMHg (0.31 and

208

0.19 ‰, respectively as 2 SD) is better than for IHg (0.39 and 0.26 ‰). Overall, they are comparable to other

209

transient signal methods14,24 and about 2 to 3 times higher than for HgT measurements, which is certainly good

210

enough to study isotopic fractionation during species reaction but not to detect minor variations, e.g. in Hg

211

sources.

212

Figure SI-10 displays an example of a session where the δ202Hg STREM values determined by either CUB or

213

CSB are confronted to the ratio between the species-specific responses for the STREM and NIST SRM-3133

214

standards (calculated from peak areas). Throughout the run the CUB values are always lower than the CSB ones

215

but the difference is larger when the sensitivity ratio between MMHg and IHg increases. We suggest that a slight

216

degradation of HgPr2 (the derivatized form of IHg) occurs upon desorption from the packing material, leading to

217

an enrichment of heavier IHg isotopes that explain the generally lower CUB values for MMHg. Overall, it

218

demonstrates that the use of such a pre-concentration technique where the species behavior might be different

219

absolutely requires a species-specific bracketing to reach the best accuracy and precision.

220


Page 11 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


221

Intercomparison with previous total and CSIA values.

222

The isotopic compositions of the various environmental RMs measured are presented in Table 2. First, it should

223

be noted that the uncertainties of replicate measurements (5-6) range between 0.04 and 1.12 ‰ for δ202Hg and

224

∆199Hg (as 2 SE, respectively and NRC TORT-2 excluded). More specifically, they are typically below or close to

225

0.2 ‰ for the predominant species and HgT but usually between 0.4 and 0.8 ‰. for the minor species. The

226

uncertainties associated to IHg in the NRC TORT-2 are obviously much higher than others (2.35 ‰ for δ202HgIHg),

227

which largely impact the HgT values. The Tl signal during IHg elution is strongly drifting, likely due to the

228

simultaneous elution of biogenic organic compounds (Figure SI-11). This perturbation is very reproducible over

229

time but not seen for the subsequent standards (data not shown). It is also observed but to a much lesser extent

230

with the ERM BCR-414 that present similar concentrations, suggesting a matrix dependent effect. From all the

231

analyses performed with natural waters, sediments or biota, we never observe this perturbation during MMHg

232

elution, suggesting that the biogenic organic compounds leading to this perturbation all have higher boiling points.

233

Even though the MMHg peak of the NRC TORT-2 is not impacted, the δ202HgMMHg value is significantly higher

234

than the previously published one19. Together, these two RMs clearly points to the limitations of our methodology

235

in the low concentration range (below about 150 ng.g-1) but further analyses of samples with intermediate

236

concentrations are required to precisely define the thresholds.

237

Nevertheless, most of our measured CSIA and calculated HgT values are in agreement with previous ones based

238

on SEM, conventional GC/MC-ICP-MS and CVG/MC-ICP-MS (unpaired t-test, p-values > 0.01) with some

239

exceptions. Significant differences are found for (i) the ∆199HgIHg of the ERM CE-464 that is lower than the value

240

reported by Perrot et al.18, (ii) the ∆199HgMMHg of the NIST SRM-1947 and ERM BCR-414 that are slightly but

241

significantly higher than previous values19 and (iii) the IHg MDF values of the IAEA-450 that are also slightly lower

242

than HgT measurements. As none of these values are currently certified, this will also require further

243

investigations to decipher which are the correct ones.

244

As previously pointed out by Masbou et al.19, the calculation of the isotopic composition of a minor species by

245

mass balance results in large errors if the species contribution is not balanced or if their isotopic compositions are

246

not very contrasted. As can be seen in table 2, the uncertainties associated to IHg values obtained by SEM are

247

indeed higher than those of direct measurement for samples with low IHg contribution, e.g. ERM CE-464 and the



Page 12 of 18

248

NIST SRM 1947. Moreover, the δ202HgIHg values are neither consistent with previous direct measurements18,24 nor

249

with the variations expected between MMHg and IHg during in vivo demethylation18. Both values and

250

uncertainties become similar or better when the proportion of IHg increases (> 30-50 %). Altogether, it implies

251

that SEM measurements are of limited usefulness for samples dominated by MMHg while the PTV-GC/MC-ICP-

252

MS technique enables the direct measurement of both the MMHg and IHg isotopic composition in samples where

253

HgT is below 1 µg.g-1 and IHg < 30 % (down to 150 ng.g-1) even though the associated uncertainties can be

254

relatively high. These conditions are met for many fish and other organisms and are well exemplified by the NIST

255

SRM-1947 (trout fish muscle) where the δ202HgIHg is lower than the δ202HgMMHg by about 2.2 ‰ as anticipated

256

from in vivo degradation18 but interestingly the MIF values are also significantly lower than their MMHg

257

counterparts (1.70 vs 5.56 ‰, respectively for ∆199Hg). It may indicate the contribution of different Hg source to

258

this fish population since in vivo pathways do not lead to significant MIF18.

259 260

CONCLUSION

261

A novel method for Hg-CSIA at ultra-trace levels through an on line pre-concentration technique combined with

262

GC/MC-ICP-MS was successfully developed and validated. Using a programmed temperature vaporization

263

injector, Hg species are efficiently preconcentrated in a packed liner while separated from the solvent and

264

subsequently transferred to the analytical column. With a compound specific standard bracketing procedure for

265

MMHg, the method demonstrates accurate results for both IHg and MMHg over long term standard analyses with

266

associated uncertainties similar to other on-line CSIA methods. The potential and limitations of the method

267

towards real sample analysis were demonstrated with environmental RMs exhibiting Hg species concentrations

268

down to about 150 ng.g-1, more than ten times better than conventional GC/MC-ICP-MS and similar to SEM

269

although the precision is inferior. The method brings an opportunity to a more comprehensive understanding

270

about the species-specific processes and sources affecting Hg in the environment.

271 272

SUPPLEMENTARY INFORMATION

273

Supplemental Methods including figures of the GC/MC-ICP-MS coupling and PTV operating conditions;

274

optimization of the PTV conditions and Hg species calibration curves; Hg species concentrations in RMs; typical


Page 13 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


275

chromatograms for GC/MC-ICP-MS and GC/Q-ICP-MS; internal precisions on isotopic ratios and isotopic

276

compositions of standards; injected volumes and corresponding concentrations of Hg; complimentary long-term

277

MIF values for standards; examples of delta values calculated by CUB and CSB; typical chromatogram for NRC

278

TORT-2.

279 280

ACKNOWLEDGMENT

281

This work is a contribution to the LA PACHAMAMA project (ANR CESA program, No ANR-13-CESA-0015-01). E.

282

Tessier and J. Barre (IPREM, UPPA/CNRS, France) are thanked for technical assistance during GC coupling and

283

operation of the MC/ICP-MS instrument. J. Masbou (GET, Univ. Toulouse, France) is acknowledged for providing

284

uncertainties associated to SEM measurements.

285



286

REFERENCES

287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)

(16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37)

Page 14 of 18

Blum, J. D.; Johnson, M. W. Rev. Mineral. Geochem. 2017, 82 (1), 733–757. Foucher, D.; Ogrinc; Hintelmann, H. Environ. Sci. Technol. 2009, 43 (1), 33–39. Estrade, N.; Carignan, J.; Donard, O. F. X. Environ. Sci. Technol. 2011, 45 (4), 1235–1242. Bergquist, B. A.; Blum, J. D. Science 2007, 318 (5849), 417–420. Sherman, L. S.; Blum, J. D.; Johnson, K. P.; Keeler, G. J.; Barres, J. A.; Douglas, T. A. Nat. Geosci. 2010, 3 (3), 173– 177. Vanhaecke, F.; Balcaen, L.; Malinovsky, D. J. Anal. At. Spectrom. 2009, 24 (7), 863. Gourgiotis, A.; Bérail, S.; Louvat, P.; Isnard, H.; Moureau, J.; Nonell, A.; Manhès, G.; Birck, J.-L.; Gaillardet, J.; Pécheyran, C.; Chartier, F.; Donard, O. F. X. J. Anal. At. Spectrom. 2014, 29 (9), 1607. Claverie, F.; Hubert, A.; Berail, S.; Donard, A.; Pointurier, F.; Pécheyran, C. Anal. Chem. 2016, 88 (8), 4375–4382. Foucher, D.; Hintelmann, H. Anal. Bioanal. Chem. 2006, 384 (7–8), 1470–1478. Blum, J. D.; Bergquist, B. A. Anal. Bioanal. Chem. 2007, 388 (2), 353–359. Štrok, M.; Hintelmann, H.; Dimock, B. Anal. Chim. Acta 2014, 851, 57–63. Fu, X.; Heimbürger, L.-E.; Sonke, J. E. J. Anal. At. Spectrom. 2014, 29 (5), 841. Sun, R.; Enrico, M.; Heimbürger, L.-E.; Scott, C.; Sonke, J. E. Anal. Bioanal. Chem. 2013, 405 (21), 6771–6781. Bérail, S.; Cavalheiro, J.; Tessier, E.; Barre, J. P. G.; Pedrero, Z.; Donard, O. F. X.; Amouroux, D. J. Anal. At. Spectrom. 2017, 32 (2), 373–384. Greenwood, P. F.; Amrani, A.; Sessions, A.; Raven, M. R.; Holman, A.; Dror, G.; Grice, K.; McCulloch, M. T.; Adkins, J. F. In Principles and Practice of Analytical Techniques in Geosciences; Grice, K., Ed.; Royal Society of Chemistry: Cambridge, 2014; pp 285–312. Elsner, M.; Imfeld, G. Curr. Opin. Biotech. 2016, 41, 60–72. Horst, A.; Renpenning, J.; Richnow, H.-H.; Gehre, M. Anal. Chem. 2017, 89 (17), 9131–9138. Perrot, V.; Masbou, J.; Pastukhov, M. V.; Epov, V. N.; Point, D.; Bérail, S.; Becker, P. R.; Sonke, J. E.; Amouroux, D. Metallomics 2016, 8 (2), 170–178. Masbou, J.; Point, D.; Sonke, J. E. J. Anal. At. Spectrom. 2013, 28 (10), 1620–1628. Janssen, S. E.; Johnson, M. W.; Blum, J. D.; Barkay, T.; Reinfelder, J. R. Chem. Geol. 2015, 411, 19–25. Epov, V. N.; Berail, S.; Pécheyran, C.; Amouroux, D.; Donard, O. F. X. In Isotopic Analysis; Vanhaecke, F., Degryse, P., Eds.; Wiley-VCH Verlag GmbH & Co. KGaA: Weinheim, Germany, 2012; pp 495–517. Epov, V. N.; Rodriguez-Gonzalez, P.; Sonke, J. E.; Tessier, E.; Amouroux, D.; Bourgoin, L. M.; Donard, O. F. X. Anal. Chem. 2008, 80 (10), 3530–3538. Rodríguez-González, P.; Epov, V. N.; Pecheyran, C.; Amouroux, D.; Donard, O. F. X. Mass. Spectrom. Rev. 2012, 31 (4), 504–521. Epov, V. N.; Berail, S.; Jimenez-Moreno, M.; Perrot, V.; Pecheyran, C.; Amouroux, D.; Donard, O. F. X. Anal. Chem. 2010, 82 (13), 5652–5662. Rodríguez-González, P.; Epov, V. N.; Bridou, R.; Tessier, E.; Guyoneaud, R.; Monperrus, M.; Amouroux, D. Environ. Sci. Technol. 2009, 43 (24), 9183–9188. Perrot, V.; Jimenez-Moreno, M.; Berail, S.; Epov, V. N.; Monperrus, M.; Amouroux, D. Chem. Geol. 2013, 355, 153– 162. Jiménez-Moreno, M.; Perrot, V.; Epov, V. N.; Monperrus, M.; Amouroux, D. Chem. Geol. 2013, 336, 26–36. Perrot, V.; Bridou, R.; Pedrero, Z.; Guyoneaud, R.; Monperrus, M.; Amouroux, D. Environ. Sci. Technol. 2015, 49 (3), 1365–1373. Hoh, E.; Mastovska, K. J. Chromatogr. A 2008, 1186 (1–2), 2–15. Blessing, M.; Jochmann, M. A.; Haderlein, S. B.; Schmidt, T. C. Rapid. Commun. Mass Sp. 2015, 29 (24), 2349– 2360. Ceulemans, M.; Łobiński, R.; Dirkx, W. M. R.; Adams, F. C. Fresenius’ J. Anal. Chem. 1993, 347 (6–7), 256–262. Heisterkamp, M.; Adams, F. C. Fresenius’ J. Anal. Chem. 1998, 362 (5), 489–493. Pacheco-Arjona, J.; Rodriguez-Gonzalez, P.; Valiente, M.; Barclay, D.; Donard, O. F. X. Int. J. Environ. An. Ch. 2008, 88 (13), 923–932. Monperrus, M.; Tessier, E.; Veschambre, S.; Amouroux, D.; Donard, O. Anal. Bioanal. Chem. 2005, 381 (4), 854–862. Baya, P. A.; Hollinsworth, J. L.; Hintelmann, H. Anal. Chim. Acta 2013, 786, 61–69. Fietzke, J.; Liebetrau, V.; Günther, D.; Gürs, K.; Hametner, K.; Zumholz, K.; Hansteen, T. H.; Eisenhauer, A. J. Anal. At. Spectrom. 2008, 23 (7), 955. Hirata, T.; Hayano, Y.; Ohno, T. J. Anal. At. Spectrom. 2003, 18 (10), 1283.


Page 15 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


341

Table 1. Operating parameters for the programmed temperature vaporization injector, GC and ICP-MS

342

instruments.

343 344


ACS Paragon Plus Environment HgT MMHg IHg

HgT IHg HgT IHg

Dogfish liver

Lobster hepatopancreas

Contaminated marine sediment

Estuarine sediments

(Zoo)plankton

NRC DOLT-4

NRC TORT-2

NIST SRM-1944

IAEA-405

ERM BCR-414

HgT MMHg IHg

a

HgT 19 MMHg SEM 19 IHg SEM

19

276 ± 18 75.4 % 24.6 %

HgT

b

810 ± 40 99.3 %

HgT

b

3400 ± 500 99.8 %


1

270 ± 60 56.3 % 43.7 %

3 3 3

6

48

5

16

5

89 6 6

6

46 9 9 4 4

HgT 19 MMHg SEM 19 IHg SEM 18 MMHg GC-ICP-MS 18 IHg GC-ICP-MS

b

6

29 6 6

6

2580 ± 220 51.6 % 48.4 %


b

941 ± 19 91.7 % 8.3 %

1

47 6 6 7 7

SD

HgT 19 MMHg SEM 19 IHg SEM 18 MMHg GC-ICP-MS 18 IHg GC-ICP-MS

±

6

av 5240 ± 100 97.5 % 2.5 %

n

-1.02 ± 0.22 -0.37 ± 0.26 -2.99 ± 0.95

-0.56 ± 0.04

-0.64 ± 0.31

-0.68 ± 0.04

-0.93 ± 0.26

-1.74 ± 1.45 1.03 ± 0.31 -5.32 ± 3.16

-0.56 ± 0.05

-0.43 ± 0.14 0.07 ± 0.25 -0.96 ± 0.23

1.55 ± 0.14 1.80 ± 0.22 -1.14 ± 1.00

0.81 ± 0.23 0.86 ± 0.24 -0.98 ± 0.50

av ± 2 SE

δ 204 Hg ± 2 SE

± ± ± ± ±

0.01 0.05 4.36 0.04 0.19

± 0.03 ± 0.03 ± 0.91

± ± ± ± ±

0.04 0.02 0.13 0.19 0.41

± 0.04

-0.07 0.06 -0.47

± 0.11 ± 0.01 ± 0.21

0.10 ± 0.21 -0.13 ± 0.17 0.79 ± 0.45

-0.38 * ± 0.03

-0.62 * ± 0.10

-0.44

-0.57 ± 0.15

0.06 ± 0.02 0.54 * ± 0.06 -1.45 ± 0.23

-0.73 ± 1.07 0.96 * ± 0.22 -2.90 ± 2.35

-0.34 0.05 -0.47 0.18 -0.89

-0.23 ± 0.08 0.08 ± 0.14 -0.55 ± 0.12

1.18 1.11 3.28

1.05 ± 0.12 1.23 ± 0.13 -0.98 ± 0.44

0.68 0.62 5.02 1.13 -1.27

0.66 ± 0.11 0.70 ± 0.11 -1.09 ± 0.52

av

δ 202 Hg ± 2 SE

± 0.03 ± 0.05

2.58 0.77

± 0.05 ± 0.01

± 0.11 ± 0.35

1.02 -0.02

± 0.07

± 0.05

0.55 0.65

± 0.10 ± 0.02

0.72 ± 0.18 0.79 ± 0.18 0.50 ± 0.22

-0.31 * ± 0.02

-0.49 * ± 0.13

-0.32

-0.42 ± 0.20

1.24

-0.03 ± 0.82 1.64 ± 0.21 -2.18 ± 1.68

± 0.03 ± 0.02

0.61 0.98

0.73 ± 0.08 1.04 ± 0.11 0.39 ± 0.13

4.91 4.95

4.85 ± 0.13 5.25 ± 0.13 0.35 ± 0.42

± 0.07

2.35

2.43 ± 0.04 2.49 ± 0.05 0.23 ± 0.73

av

δ201Hg

a. in µg.kg dry weight; b. in-house measurements by CCVG-MC-ICP-MS; values that are significantly different are indicated by an asterisk

-1

HgT MMHg IHg

Trout tissue

NIST SRM-1947

HgT MMHg IHg

Tuna fish muscle

HgT MMHg IHg

ERM CE-464

346 concentrations and

± 2 SE

± 0.05 ± 0.14

± 0.04

± 0.02 ± 0.04

± 0.09 ± 0.25

± 0.02 ± 0.02

± 0.04

± 0.03

0.15 0.08

± 0.12 ± 0.06

0.12 ± 0.17 -0.02 ± 0.16 0.55 ± 0.59

-0.20 * ± 0.02

-0.34 * ± 0.06

-0.23

-0.23 ± 0.19

0.31

-0.34 ± 0.53 0.39 ± 0.15 -1.27 ± 1.05

0.08 -0.54

-0.14 0.06

-0.08 ± 0.08 0.04 ± 0.15 -0.22 ± 0.14

0.69 0.63

0.69 ± 0.10 0.74 ± 0.13 0.17 ± 0.76

0.71 -0.55

0.39

0.39 ± 0.10 0.43 ± 0.08 -1.00 ± 0.64

av

δ 200 Hg ± 2 SE

± 0.09 ± 0.08

± 0.05

± 0.06 ± 0.02

± 0.06 ± 0.16

± 0.02 ± 0.01

± 0.05

± 0.03

0.71 0.84

± 0.10 ± 0.03

0.93 ± 0.11 1.17 ± 0.13 0.20 ± 0.57

-0.13 * ± 0.01

-0.28 * ± 0.06

-0.11

-0.22 ± 0.07

1.18

0.68 ± 0.36 1.17 ± 0.11 0.04 ± 0.79

0.91 0.72

0.95 1.20

1.04 ± 0.08 1.23 ± 0.10 0.84 ± 0.10

5.45 5.57

5.50 ± 0.09 5.87 ± 0.11 1.45 ± 0.60

2.49 1.66

2.50

2.45 ± 0.09 2.49 ± 0.09 0.76 ± 0.51

av

δ 199 Hg

0.02

-1.16 ± 0.42 -0.18 ± 0.27 -4.17 ± 1.12

0.28 ± 0.35

-0.01

-0.08 ± 0.08

-0.09 ± 0.02

-0.66 ± 0.30 -0.40 ± 0.13 -0.99 ± 0.65

-0.10 ± 0.08 -0.05 ± 0.08 -0.15 ± 0.11

-0.01 ± 0.15 -0.04 ± 0.17 0.33 ± 0.52

-0.09 ± 0.01

-0.17 ± 0.17 -0.19 ± 0.18 0.65 ± 0.50

av ± 2 SE

∆204Hg ± 2 SE

± 0.01 ± 0.03

± 0.04 ± 0.05

± 0.02 ± 0.01

± 0.01 ± 0.05

± 0.02

± 0.01

0.61 ± 0.14 0.61 * ± 0.01

0.65 ± 0.15 0.89 * ± 0.10 -0.09 ± 0.33

-0.03

-0.02 ± 0.07

-0.02

0.01 ± 0.14

0.59 0.83

0.52 ± 0.16 0.92 ± 0.10 0.00 ± 0.33

0.88 0.65

0.87 0.94

0.90 ± 0.04 0.99 ± 0.05 0.80 ± 0.05

4.02 ± 0.04 4.12 * ± 0.02

4.06 ± 0.08 4.33 * ± 0.07 1.09 ± 0.33

1.73 ± 0.06 1.72 * ± 0.09

1.97 1.88

1.94 ± 0.11 1.96 ± 0.12 1.05 * ± 0.80

av

∆201 Hg

0.13 ± 0.08 0.05 ± 0.05

0.07 ± 0.21 0.04 ± 0.14 0.15 ± 0.68

-0.01 ± 0.02

-0.03 ± 0.09

0.01 ± 0.02

0.06 ± 0.12

0.06 ± 0.01 0.04 ± 0.02

0.03 ± 0.35 -0.10 ± 0.16 0.19 ± 0.77

-0.01 ± 0.19 -0.09 ± 0.05

0.03 ± 0.01 0.04 ± 0.01

0.03 ± 0.06 0.01 ± 0.10 0.06 ± 0.10

0.10 ± 0.01 0.08 ± 0.03

0.17 ± 0.06 0.12 ± 0.07 0.67 ± 0.68

0.15 ± 0.06 0.09 ± 0.05

0.08 ± 0.01 0.08 ± 0.02

0.06 ± 0.06 0.08 ± 0.06 -0.45 ± 0.45

av ± 2 SE

∆200 Hg

± 2 SE

± ± ± ± ±

0.01 0.04 3.22 0.11 0.03

± ± ± ± ±

0.02 0.01 0.09 0.01 0.06

± 0.01 ± 0.05 ± 0.13

± 0.02

± 0.02

0.73 ± 0.11 0.82 * ± 0.03 0.45 ± 0.25

0.91 ± 0.14 1.20 * ± 0.11 0.00 ± 0.63

-0.04

-0.13 ± 0.06

0.00

-0.08 ± 0.05

0.75 1.04 0.54

0.86 ± 0.36 0.93 ± 0.12 0.77 ± 0.86

1.03 1.19 0.67 0.86 0.95

1.10 ± 0.06 1.21 ± 0.09 0.98 ± 0.07

5.15 ± 0.05 5.29 * ± 0.02 4.93 ± 0.61

5.24 ± 0.07 5.56 * ± 0.09 1.70 ± 0.55

2.40 2.34 0.34 2.21 1.98 *

2.28 ± 0.06 2.32 ± 0.06 1.04 * ± 0.41

av

∆199Hg

345

Material

Certified

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

SRM

Analytical Chemistry Page 16 of 18

Table 2. Compilation of speciation and isotopic values for the various Reference Materials tested.

16

Page 17 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


347

348 349 350 351 352 353 354 355

Figure 1. Distribution of isotopic compositions measurements over 2 years of the NIST SRM-8610 and Strem standards (values given as average ± 2SD). For the MMHg Strem, results from both Compound Unspecific Bracketing (CUB) and Compound Specific Bracketing (CSB) calculation methods are presented. Dashed lines represent values determined with CCVG/MC-ICP-MS (Blum and Bergquist, 2007 for the NIST SRM-8610 and Epov et al., 2010 for the Strem standard).



356 357

Page 18 of 18

For TOC only

358


Hg compound-specific isotope analysis at ultra-trace levels using an

Recommend Documents