Sensitive simultaneous determination of 19 fluorobenzoic acids in saline waters by solid-phase extraction and liquid chromatography–tandem mass spectrometry

Accepted Manuscript Title: Sensitive simultaneous determination of 19 fluorobenzoic acids in saline waters by solid-phase extraction and LC-MS/MS Author: Paweł Kubica Herv´e Garraud Joanna Szpunar Ryszard Lobinski PII: DOI: Reference:

S0021-9673(15)01308-4 http://dx.doi.org/doi:10.1016/j.chroma.2015.09.024 CHROMA 356846

To appear in:

Journal of Chromatography A

Received date: Revised date: Accepted date:

17-5-2015 4-9-2015 8-9-2015

Please cite this article as: P. Kubica, H. Garraud, J. Szpunar, R. Lobinski, Sensitive simultaneous determination of 19 fluorobenzoic acids in saline waters by solid-phase extraction and LC-MS/MS, Journal of Chromatography A (2015), http://dx.doi.org/10.1016/j.chroma.2015.09.024 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

*Highlights (for review)

ce pt

ed

M

an

us

cr

ip t

Oil reservoir tracers were determined in salt-rich waters 19 fluorinated benzoic acidswere simultaneously preconcentrated by SPE Previously reported LC-MS/MS detection limits were improved 10-20 times

Ac

  

Page 1 of 49

*Manuscript

[Tapez un texte]

2

Sensitive simultaneous determination of 19 fluorobenzoic acids in saline waters by solid-phase extraction and LC-MS/MS

3

Paweł Kubica,aHervé Garraudb, Joanna Szpunarc* and Ryszard Lobinskic,d

1

4

7 8 9 10

Department of Analytical Chemistry, Faculty of Chemistry, Gdansk University of Technology, 11/12 Narutowicza Str, Gdańsk, Poland b

SOBEGI, Laboratoire Contrôle et Environnement, Pôle 4, Av. du Lac, 64150 Mourenx, France

ip t

6

a

c

CNRS/UPPA, Laboratoire de Chimie Analytique Bio-inorganique et Environnement (LCABIE-IPREM), Hélioparc, 2, AvenuePr. Angot, 64053 Pau, France

cr

5

11 12

d

13

[email protected], tel: +33 559 40 77 55, fax: +33 559 40 77 82

an

us

Department of Analytical Chemistry , Warsaw University of Technology, ul. Noakowskiego 3, 00-664 Warsaw, Poland

Ac

ce pt

ed

M

14

1 Page 2 of 49

[Tapez un texte]

Abstract

16

30

A solid-phase extraction (SPE) procedure using C18 stationary phase was optimized for the preconcentration of 19 fluorinated derivatives of benzoic acid (FBA): mono-, ditri- and tetrafluorosubstituted in the ring, trifluoromethylbenzoic acid and 3,5bistrifluoromethyl benzoic acid from undiluted salt-rich (>20%) reservoir waters. Quantitative (>90%) retention/elution of 16 out of 19 analyte compounds was achieved allowing a 4-fold preconcentration factor accompanied by the elimination of >99% of salt. For the three most polar compounds (2,6-dFBA, 2,3,6-tFBA and 2,4,6tFBA) the non-quantitative recoveries(>70%)were corrected by dedicated customsynthesized deuterated internal standards. The FBAs were determined by HPLC MS/MS revisited in terms of a choice of column, elution conditions and MS/MSsignal acquisitionparameters allowing the baseline separation and a gain in sensitivity. For a sample intake of 4 mL, detection limits for all the compounds in a reservoir water sample containing more than 20% salt were between 0.01 and 0.05 ng/ml which represents a gain of a factor of 10-20 in comparison with the state-of the art LCMS/MS procedures for samples of similar complexity.

31

Keywords:fluorobenzoic acids, solid-phase extraction, reservoir water, LC MS/MS

20 21 22 23 24 25 26 27 28 29

cr

19

us

18

an

17

ip t

15

M

32

Introduction:

34

Derivatives of benzoic acid with one or more fluorine atoms, or one or more trifluoromethyl groups,attached to the aromatic ring are the most common currently used non-radioactive passive water tracers for oil field applications [1]. As a tracing campaign involves a set of several different compounds (out of more than 20 commercially available), there is a need for methods for their simultaneous determination in an oil reservoir water matrix. Low detection limits are critical as they determine the quantity of the tracers necessary to be used and thus the cost and the environmental impact of the campaign. The matrix differs depending on the sample origin but it is usually rich in salts (reaching in some cases up to 30%) and organic constituents [2].

37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

ce pt

36

Ac

35

ed

33

The lowest detection limits (down to 0.01 ng/ml) were obtained by gas chromatography (GC)- MS but lengthy (24 h) and tedious sample preparation procedures including matrix removal and derivatization were necessary[3]. The incomplete and strongly compound-dependent yields required compound specific isotope dilution calibration that was proposed for sixspecies determined to achieve accurate analysis. [4],[5]. The alternative is the use of HPLC - MS/MS analysis to eliminate the derivatization step and thus to simplify the sample processing. The original work [5 ], which was applied to simple matrices butdid not show any chromatogram reported fairly high 2 Page 3 of 49

[Tapez un texte]

61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90

Although the reported selectivity of HPLC separation of a set of usually studied 20 tracers wasgenerally high, the baseline separation of all of them was not achieved in any of the published works [3-9]. This caveat was compensated by the determination of the co-eluting compounds using different fragmentation reactions. On the other hand, the number of theoretical plates achieved in HPLC is important. Indeed, the poor specificity of fragmentation reactions (the loss of CO2) used for the quantification, in combination with the unit resolution of a quadrupole filter and matrix rich in organic acids, may lead to the increase in baseline and false positives.

ip t

60

cr

59

us

58

The above reasons spur the need for the development of methods allowing a considerable enrichment of FBAs with regard to salt and organic matrix. Solid phase extraction (SPE) is an attractive option for both matrix removal and preconcentration of analytes [10-12]prior to LC-MS/MS analysis of samples rich in salts. However, quantitative SPE of FBAs from reservoir waters is a difficult task because of the high polarity of the tracers. The problems result, on one hand, from the difficulty to trap quantitatively and simultaneously all the analytes while avoiding the retention of the matrix and, on the other hand, to release the trapped analytes quantitatively without substantial dilution. Another critical factor is the sample volume to be used for analysis as it determines the SPE time.

an

57

M

56

ed

55

ce pt

54

detection limits: 0.5-1 ng/ml for electrospray ionization (ESI) and 10-20 ng/ml for atmospheric pressure chemical ionization (APCI), respectively. The detection limits were considerably (about an order of magnitude) decreased by Serres-Pioles at al.[1] except for tFBA, for which hardly any improvement was observed. The maximum tolerated salt content of the samples allowed by the method was pretty low (1%) which required a considerable sample dilution (10-20times) drastically limiting the scope of the method applications.

As a result of an extensive optimization study, Müller et al.reported fairly satisfactory recoveries (between71% (2,5-dFBA) and 94 % (3-FBA))from tap water [7]but for reservoir waters the extraction efficiencies were generally low (down to 18% for 2,3,5,6-tetraFBA and 2,6-dFBA)and strongly compound-dependent [3]. Moreover, relatively large sample volumes (100 ml) processed [3, 7] resulted in long analysis times. The recovery problems were (for sixselected compounds) addressed by the use of custom synthetized deuterated internal standards[4]which were used in the analysis of reservoir and ground water [8].

Ac

53

The main goal of this work was the development of a rapid (small sample volume) quantitative SPE method allowing a direct multi-tracer (19 compounds) analysis in salt-rich (>20% salt) reservoir water samples with an objective to reach at least an order of magnitude in terms of detection limits over the direct injection procedure [1].

91 3 Page 4 of 49

[Tapez un texte]

Experimental conditions

93

Samples Collection.Reservoir water samples of different origins with different salt contents: Gabon (200 g/l), Qatar (220 g/l), Russia (170 g/l), Yemen (80 g/l) and Congo (250 g/l) were used for the method development. The salts components were primary sodium and calcium with minor contribution of potassium and magnesium [2]. The samples were collected in 5-L glass flasks and the aqueous and organic fractions were separated by gravitation. Sub-samples of 100 mL were transported in ambient temperature in glass flasks in containers preventing the exposure to light; the samples were acidified to pH 2-3 with formic acid and stored prior to analysis at 4°C in dark;in these conditions they were stable at least 90 days.

101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127

cr

100

Reagents and standards.Acetonitrile, acetic acid, tetrahydrofuran, ammonia aq. were purchased from Sigma-Aldrich (Saint-Quentin-Fallavier, France). Ultrapure water (18 MΩ.cm) was obtained from a Milli-Q system (Millipore, Bedford, MA). The characteristics of the FBA standards used in this study are listed in Table 1.Deuterated 2,6-dFBA and 2,4,6-tFBA were a gift from Dr. K. Müller and Prof.Dr.A. Seubert (Faculty of Chemistry, Philipps-Universität, Marburg, Germany).4fluorobenzoic acid-α-13C-2,3,5,6-d4 was purchased from Sigma-Aldrich (SaintQuentin-Fallavier, France).

us

99

an

98

M

97

Materials. The SPE disposable cartridges (C18, 500 mg, 3 mL) were supplied by SigmaAldrich (Saint-Quentin-Fallavier, France). Separations were carried out using an Acquity UPLC BEH C18 column (150 mm x 2.1 mm x1.7 µm) with a matching precolumnAcquity UPLC BEH C18 VanGuard (130Å, 1.7 µm, 2.1 mm X 5 mm)(Waters, Guyancourt, France).

ed

96

ce pt

95

Instrumentation. SPE was carried out using aSupelco VisiPrep 24DL(supplied by Sigma-Aldrich).Eluates wereevaporated to dryness using an Eppendorf Concentrator Plus(Eppendorf France SAS, Montesson).An Acquity UPLC system (Waters) including a binary solvent pump, a cooled autosampler and a column oven was used. The detector was a XevoTQ (quadrupole-T-wave-quadrupole) MS with an orthogonal Zspray-electrospray interface (Waters).

Ac

94

ip t

92

Procedures

Initial sample preparation procedure.Samples were filtered through 0.2 µm (13mm) syringe filter, GHP Acrodisc(Interchim, Montluçon, France)). 4-fluorobenzoic acid-α13 C-2,3,5,6-d4) was added at 20 ng/mL as an internal standard. Deuterated 2,6-dFBA and 2,4,6-tFBA were added at 20 ng/mL if the corresponding compounds were to be determined.

4 Page 5 of 49

[Tapez un texte]

135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152

ip t

134

Measurement conditions. A 50 µL aliquot was analyzed by HPLC - MS/MS.Mobile phase was composed by mixing 0.05% CH3COOH (A) and 0.05% CH3COOH in acetonitrile (B). The elution gradient was: 0 min (13% B), 1.3 min (13% B), 9 min (28 % B) and 13 min (80 % B). The column was equilibrated for 5 min. The flow rate was 0.45 ml/min, the column temperature was 45°C and the autosampler temperature was 5°C.Tandem MS data acquisition was performed with the electrospray source operating in negative mode (ESIneg) under the MRM conditions listed in Table 2.

cr

133

us

132

an

131

Calibration. A calibration curve was constructed by plotting peak area for 7 concentrations(0.05, 0.1, 0.2, 0.5, 1, 10, 20 ng/mL).

M

130

Data processing. The Masslynx software (Waters, Milford, MA) was used to process data.

ed

129

Solid-phase extraction.The SPE cartridges were conditioned with two successive 2-ml volumes of acetonitrile followed by rinsing with two successive 2-ml volumes of water. Then, the sample was loaded as two successive 2-mL aliquots. After loading of the sample, the sorbent was rinsed with a 2-mL volume of water to remove remaining salts and polar compounds. The cartridge was dried for 3 min under the gentle stream of nitrogen (purity 99.999 %). Then, the elution was performed with two successive 2-mL volumes of acetonitrile: 10%NH4OHaq.(8:2 v/v).Thefirst portion of the eluent was kept for 3 min to facilitate the desorption of analytes.The eluate was collected and evaporated to dryness under vacuum. The residue was dissolved in 1 mL of 10% (v/v) acetonitrile and analyzed by HPLC - MS/MS.

Quality control and assurance. For the purpose of method validation three samples were prepared by spiking a reservoir water (salinity 22%) at the different concentration levels: 0.2, 1 and 10 ng/mL, respectively. The samples were analysed by procedure developed.

ce pt

128

153

155 156 157 158 159 160 161 162 163 164

Results and discussion

Ac

154

LC - MS/MS determination of FBAs The separation methods reported in the literature were based on isocratic elution inion-chromatography [7] or C18 reversed phase chromatography [5]. An improved selectivity in reversed-phase HPLC was obtained by gradient elution with slightly acidic methanol or acetonitrile [9]. The latter procedure was the starting point for the optimization of the HPLC separation conditions in this work. In order to obtain the baseline separation and to reduce the co-elution with matrix components, the length of the column was increased which tripled the number of theoretical plates in comparison tothe former work [9]and the baseline separation of all the 19 FBAs to be achievedwithin 13 min as shown in Fig. 1. 5 Page 6 of 49

[Tapez un texte]

169 170 171 172 173 174 175 176 177 178 179 180 181

ip t

168

Table 3 also shows that the obtained detection limits were in one case lower than, and in one case comparable with, the indicative values received from the manufacturers for QTOF systems operated in the MRM mode. The higher resolution of QTOF may offer an advantage of reducing the risk of false positives in the case of more complex samples. On the other hand, the range of linearity of the triple quadrupole spectrometer was an order of magnitude larger than that of the TOF instruments.

cr

167

us

166

The calibration curvesshowed good linearity (r2>0.999) and precision below 3% (n=3) (as shown inTable 1 Supplementary Information). The detection limits calculated as 3x standard deviation of blank integrated at the corresponding retention times are summarized in Table 3. In the absence of sample matrix, the LODs depend,in particular, on the ionization efficiency. The latter waslargely affected by the low content of the organic modifier for the early eluting species (2,6-dFBA, 2,3,6-tFBA, 2,4,6tFBA and 2,3,4,5-tetraFBA) for which relatively high LODs were observed. In general, the LODs compare favorably with those published elsewhere for LC-based methods[5,7,9].The most spectacular gain (10-fold) was obtained for the triFBAswhich are very sensitive to ionization conditions.

an

165

M

182

Optimisation of SPE conditions

184

Müller at al. [3] published a comprehensive comparison study of five different SPE materials tested in a broad pH range (1-11); the best results were obtained for two of them: Oasis HLB-Plus (hydrophilic-lipophilic-balanced reversed-phase poly(divinylbenzene-co-N-vinylpyrrolidone sorbent) and Isolute ENV+ (hydroxylated polystyrenedivinylbenzene copolymer) at pH 3.4 and 1.5, respectively [3]. Preliminary tests in these conditions for salt-rich reservoir waters produced very low (often 10-20%) and irreproducible recoveries. The preliminary tests using Oasis HLB phase were not encouraging, either. Although high, quasi-quantitative recoveries of the analytes were obtained, no conditions could be found for their quantitative desorption. The most promising results were obtained with a C18sorbentsimilar to that of the column which was investigated in detail.

187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202

ce pt

186

Ac

185

ed

183

The optimization procedure included: (i) choice of the solvent for the initial conditioning step (acetonitrile or tetrahydrofuran); (ii) pH of the final condition step and sample (acidic, neutral, or alkaline); (iii) choice of the elution solvent (acetonitrile and tetrahydrofuran) and its pH. The initial experiments with MeOH were unsuccessful. The conditions tested are summarized inTable 4. The results of the recoveries obtained during the optimization are summarized in Fig. 2. The first hypothesis tested involved lowering pH to revert the dissociation of FBAs in order to increase their retention and then alkalize the solution for their elution. The 6 Page 7 of 49

[Tapez un texte]

210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228

In terms of elution conditions, the use of ammonia resulted in recovery ratios of FBAs higher than 90% for most of the analytes. Two polar organic eluting solvents (acetonitrile and THF) were testedtogether with ammonia. Recoveries from SPE procedures IX to XII were similar. Procedure X was chosen because theresulting solution (8:2 organic/aqueous) was easier to evaporate than 5:5 organic/aqueoussolution and because acetonitrile was easier to evaporate than THF.Also, the recoveries for 2,6-dFBA and 2,3,6-tFBA were significantly higher in comparison with other procedures.

ip t

209

cr

208

us

207

Fig. 2. indicates that quantitative (>90%)recoveries (retention/elution) of 16 out of 19 analyte compounds were achieved from a salt-rich water matrix. The simultaneous elimination of >99% of salt content and matrix simplification allowed a 4-fold preconcentration factor. For three compounds: 2,6-dFBA, 2,3,6-tFBA and 2,4,6-tFBA non-quantitative recoveries were observed.

an

206

M

205

The values in Fig. 2 were completed by verifying the recoveries from the water by the method developed at three different concentration levels. The data are shown in Table 5. This systematic study showed that, in fine, only two compounds were problematic in terms of recoveries (2,6-dFBA, recovery ca. 50% and 2,4,6-tFBA, recovery ca. 80%). It could also be concluded that the matrix did not practically affect the recoveries.

ed

204

acidification was initially carried out only during the conditioning step (1% acetic acid) but the recoveries were lower than when the conditioning was carried out with water (cf. e.g. procedures VI-XII). The recoveriesdropped further when acetic acid was added to the sample during the loading step (procedure II). Hence, it was decided to add acid neither during conditioning nor to the sample. Note that the recoveries in alkaline conditions (conditioning step and sample) (procedure III) were dramatically low (possibly also due to the signal suppression because of the non-retained salt).

ce pt

203

229

231 232 233 234 235 236 237 238 239 240

SPE - HPLC- MS/MS for the simultaneous multiple tracer analysis

Ac

230

Fig. 3. shows a chromatogram obtained for a concentration of 50 pg/mlFBAs added to a sample matrix containing 200 g/l of salt by the SPE method developed and the corresponding blanks. The analytes’ concentration was chosen to correspond roughly to the detection limits of the procedure based on the direct injection HPLC. The figure clearly shows peaks for all the compounds well above the background; it demonstrates not only the absence of the need for sample dilution despite the high salt content but also an effective preconcentration factor of up to 4 times resulting from the SPE. The LODs are affectedby the ionization efficiency (the degree of matrix removal and the content of acetonitrile at a given point of the chromatographic gradient), the peak shape and the baseline noise (again depending on the matrix). 7 Page 8 of 49

[Tapez un texte] 241 242 243 244

The calibration curve data obtained for the procedureand the detection and quantification limits are summarized in Table 6. They confirm a 3-4-fold gain in detection limits resulting from the preconcentration factor in addition to the absence of the need of sample dilution prior to analysis.

245

Isotope dilution correction for the non-quantitatively eluted compounds:

247

The recoveries of themost polar compounds 2,6-dFBA, 2,3,6-tFBA and 2,4,6-tFBA were not only non-quantitative(theywere not sufficiently adsorbed on the C18 sorbent and partially found in the eluate of the spiked sample)but they were also observed to vary by up to 30 % depending on the day and sample matrix. Therefore they have to be corrected for.

252 253 254 255 256 257 258 259

cr

251

us

250

A convenient method proposed by Müller et al.[4, 8]is the use of deuterated standards The chromatograms(Fig. 4) show the perfect co-elution of the doubly deuterated and non-deuterated standards which allows them to be measured in identical ionization conditions as the analyte.Table 7highlights the benefits from the isotopically-labelled internal standards showing an efficient correction of the nonquantitative recoveries. Note that a single internal standard wassufficient to correct both of 2,3,6-tFBa and 2,4,6-tFBA recoveries as these compounds elute closely and share the reaction used for their quantification.

an

249

M

248

ip t

246

ed

260

Validation of the method developed

262

In order to validate the method, three synthetic samples containing all the tracers at the different concentration levels: 0.2, 1 and 10 ng/ml were prepared and analysed according to the developed procedure. The results shown in Table 8demonstrate consistent accuracies between 90-100% and precision between 2-5%.

263 264 265

ce pt

261

267 268 269 270 271 272 273 274 275 276 277

Ac

266

Analysis of real samples: comparison with the direct analysis The developed method was compared with the method based on the direct injection of diluted samples[1].The examples of chromatograms are shown in Fig. 5. The comparison shows an increase in sensitivity over at least an order of magnitude, allowing the detection of peaks in the background not seen with the direct injection method, stabilization of the baseline, and especially the elimination of the false positives commonly encountered when integrating the peaks close to baseline using the direct injection procedure. Note that the direct injection method developed elsewhere[1] was slightly improved by diverting the chromatographic eluate off the detector for the first30 sto reduce the load of the salt on the column, as recently suggested by Bayen [13]. 8 Page 9 of 49

[Tapez un texte] 278

Conclusions

280

287

The optimization of solid phase extraction allowed an efficient and straightforward simultaneous preconcentration of 19 fluorinated derivatives of benzoic acid commonly used as oil reservoir tracers from salt-rich waters.The simultaneous elimination of the salt eliminated the need for sample dilution allowing a gain of 1020 in terms of detection limits in comparison with the figures of merit reported elsewhere in the literature for the HPLC-MS/MS analysis of similar samples.The method requires a few ml of sample only, is relatively rapid and can be readily automated.

288

Acknowledgements

289 290 291 292 293 294

The authors thank Dr. O. Arwal, TOTAL (France) for supplying the samples used for the method development and Dr. K. Müller and Prof. Dr. A. Seubert (Faculty of Chemistry, Philipps-Universität, Marburg, Germany) for the gift of the deuterated 2,6-dFBA and 2,4,6tFBA. We also thank Applied Biosystems and Bruker (Paris) for providing indicative detection limits data for the FBA standardsfor the last generation Q-TOF systems. The financial support of the mass spectrometryplatform at the LCABIE-IPREM by Aquitaine Region is acknowledged.

cr us

286

an

285

M

284

ed

283

ce pt

282

Ac

281

ip t

279

9 Page 10 of 49

[Tapez un texte] 295 296

Captions to Figures

297 298

Figure 1.HPLC-MS/MS chromatogramsobtained for 50 ng/mL standards. a) 139-->95: 1) 2-fluorobenzoic acid, 2) 3-fluorobenzoic acid, 3) 4fluorobenzoic acid;

301 302 303

b) 157-->113: 4) 2,6-difluorobenzoic acid, 5) 2,5-difluorobenzoic acid, 6) 2,3- difluorobenzoic acid, 7) 2,4-difluorobenzoic acid, 8) 3,5difluorobenzoic acid, 9) 3,4-difluorobenzoic acid;

304 305 306

c) 175-->113: 10) 2,3,6-trifluorobenzoic acid, 11) 2,4,6-trifluorobenzoic acid, 12) 2,4,5-trifluorobenzoic acid, 13) 2,3,4-trifluorobenzoic acid, 14) 3,4,5-trifluorobenzoic acid;

307 308

d) 189-->145: 15) 2-(trifluoromethyl)benzoic acid, 16) (trifluoromethyl)benzoic acid, 17)4-(trifluoromethyl)benzoic acid;

309

e) 193-->149: 18) 2,3,4,5-tetrafluorobenzoic acid;

310

f) 257-->213: 19) 3,5-bis(trifluoromethyl)benzoic acid;

311

g) 144->99: 20) 4-fluorobenzoic acid-α-13C-2,3,5,6-d4(internal standard);

cr

ip t

299 300

us

an

312

Figure 2. Analyte recoveries from a spiked reservoir water sampleobtained with the SPE procedures described in Table 4.

M

313 314

3-

315

Figure 3. HPLC-MS/MS chromatograms of a reservoir water spiked with 50 pg/mL of each FBA (top chromatogram in each subfigure) and the corresponding blank (unspiked reservoir water) analysed by the developed procedure.

319 320

a) 139-->95: 1) 2-fluorobenzoic acid, 2) 3-fluorobenzoic acid, 3) 4fluorobenzoic acid;

ce pt

ed

316 317 318

b) 157-->113: 4) 2,6-difluorobenzoic acid, 5) 2,5-difluorobenzoic acid, 6) 2,3- difluorobenzoic acid, 7) 2,4-difluorobenzoic acid, 8) 3,5difluorobenzoic acid, 9) 3,4-difluorobenzoic acid;

321 322 323

327 328

c) 175-->113: 10) 2,3,6-trifluorobenzoic acid, 11) 2,4,6-trifluorobenzoic acid, 12) 2,4,5trifluorobenzoic acid, 13) 2,3,4-trifluorobenzoic acid, 14) 3,4,5-trifluorobenzoic acid;

Ac

324 325 326

d) 189-->145: 15) 2-(trifluoromethyl)benzoic acid, 16) (trifluoromethyl)benzoic acid, 17) 4-(trifluoromethyl)benzoic acid;

329

e) 193-->149: 18) 2,3,4,5-tetrafluorobenzoic acid;

330

f) 257-->213: 19) 3,5-bis(trifluoromethyl)benzoic acid;

3-

331 332 333

Figure 4. HPLC-MS/MS chromatograms early eluting compounds with specific internal standards:

10 Page 11 of 49

[Tapez un texte]

a) 157 --> 113: 1) 2,6- difluorobenzoic acid; b) 159-->115: 2) 2,6difluorobenzoic acid -d2; c) 177 --> 131: 3) 2,3,6-tFBA, 4) 2,4,6-TFBA; d) 177-->133: 5) 2,4,6-tFBA-d2.

334 335 336 337 338 339 340

Figure 5.HPLC-MS/MS chromatograms of two (A and B) reservoir water samples. a,b Sample A. c,d - Sample B. a,c- direct injection upon dilution [9]b,d analysed by the SPE-HPLC-MS/MS procedure developed.

341

Ac

ce pt

ed

M

an

us

cr

ip t

342

11 Page 12 of 49

[Tapez un texte] 343

References

345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382

[1] C. Serres-Piole, A. Commarieu, H. Garraud, R. Lobinski, H. Preud'Homme, New passive water tracers for oil field applications, Energy and Fuels, 25 (2011) 4488-4496. [2] C. Serres-Piole, New water tracers for water reservoirs. A contribution to the fundamental understanding of tracer behaviour to enhance nanoscale monitoring in advanced reservoir exploitation by LC - tandem MS., PhD Thesis, University of Pau, France (2011). [3] K. Müller, A. Seubert, Ultra trace determination of fluorobenzoic acids in tap and reservoir water using solid-phase extraction and gas chromatography-mass spectrometry, Journal of Chromatography A, 1260 (2012) 9-15. [4] K. Müller, A. Seubert, Synthesis of deuterium-labelled fluorobenzoic acids to be used as internal standards in isotope dilution mass spectrometry, Isotopes in Environmental and Health Studies, 50 (2014) 88-93. [5] R.K. Juhler, A.P. Mortensen, Analysing fluorobenzoate tracers in groundwater samples using liquid chromatography-tandem mass spectrometry: A tool for leaching studies and hydrology, Journal of Chromatography A, 957 (2002) 11-16. [6] T. Isemura, F. Kitagawa, K. Otsuka, Separation of complex mixtures of fluorobenzoic acids by capillary electrophoresis, Journal of Separation Science, 32 (2009) 381-387. [7] K. Müller, A. Seubert, Separation and determination of fluorobenzoic acids using ion chromatography-electrospray mass spectrometry, Journal of Chromatography A, 1270 (2012) 96-103. [8] K. Müller, A. Seubert, Ultra trace determination of fluorobenzoic acids in reservoir and ground water using isotope dilution gas chromatography mass spectrometry, Isotopes in Environmental and Health Studies, 50 (2014) 277-284. [9] C. Serres-Piole, N. Moradi-Tehrani, R. Lobinski, H. Preud'homme, Direct sensitive simultaneous determination of fluorinated benzoic acids in oil reservoir waters by ultra highperformance liquid chromatography-tandem mass spectrometry, Journal of Chromatography A, 1218 (2011) 5872-5877. [10] M. Concheiro, S. Anizan, K. Ellefsen, M.A. Huestis, Simultaneous quantification of 28 synthetic cathinones and metabolites in urine by liquid chromatography-high resolution mass spectrometry, Analytical and Bioanalytical Chemistry, 405 (2013) 9437-9448. [11] V. Gabet-Giraud, C. Miege, B. Herbreteau, G. Hernandez-Raquet, M. Coquery, Development and validation of an analytical method by LC-MS/MS for the quantification of estrogens in sewage sludge, Analytical and Bioanalytical Chemistry, 396 (2010) 1841-1851. [12] M.J. Whiting, Simultaneous measurement of urinary metanephrines and catecholamines by liquid chromatography with tandem mass spectrometric detection, Annals of Clinical Biochemistry, 46 (2009) 129-136. [13] S. Bayen, X. Yi, E. Segovia, Z. Zhou, B.C. Kelly,Analysis of selected antibiotics in surface freshwater and seawater using direct injection in liquid chromatography electrospray ionization tandem mass spectrometry, Journal of Chromatography A, 1338 (2014) 38-43.

cr

us

an

M

ed

ce pt

Ac

383

ip t

344

12 Page 13 of 49

Ac

ce

pt

ed

M

an

us

cr

ip t

Figure

Fig. 1 HPLC-MS/MS chromatograms obtained for 50 ng/mL standards. Page 14 of 49

Ac

ce

pt

ed

M

an

us

cr

ip t

Figure

Fig. 2 Analyte recoveries from a spiked reservoir water sample obtained with the SPE procedures described in Table 4. Page 15 of 49

Intensity [cps]

Intensity [cps]

Time [min]

Time [min]

Intensity [cps]

(b)

i

(d)

cr

Intensity [cps]

Intensity [cps]

us Intensity [cps]

M an

Intensity [cps]

Time [min]

Intensity [cps]

ed

Intensity [cps]

(a)

Intensity [cps]

Intensity [cps]

ce pt

(c)

Ac

Intensity [cps]

Figure

Time [min]

(e)

Time [min]

(f)

Time [min]

Page 16 of 49

Fig.3. HPLC-MS/MS chromatograms of a reservoir water spiked with 50 pg/mL of each FBA

Ac

ce

pt

ed

M

an

us

cr

ip t

Figure

Fig. 4. HPLC-MS/MS chromatograms early eluting compounds

Page 17 of 49

Ac

ce

pt

ed

M

an

us

cr

ip t

Figure

Fig. 5 HPLC-MS/MS chromatograms of two (A and B) reservoir water samples

Page 18 of 49

Tables

[Tapez un texte]

Table 1. Standard compounds used in this study Formula

Purity [%]

2-fluorobenzoic acid

2-FBA

C7H5O2F

99

3-fluorobenzoic acid

3-FBA

C7H5O2F

99

4-fluorobenzoic acid

4-FBA

C7H5O2F

98

2,6-difluorobenzoic acid

2,6-dFBA

C7H4O2F2

98

2,5-difluorobenzoic acid

2,5-dFBA

C7H4O2F2

98

2,3-difluorobenzoic acid

2,3-dFBA

C7H4O2F2

98

2,4-difluorobenzoic acid

2,4-dFBA

C7H4O2F2

99

3,5-difluorobenzoic acid

3,5-dFBA

C7H4O2F2

97

3,4- difluorobenzoic acid

3,4-dFBA

C7H4O2F2

99

2,3,6-tFBA

C7H3O2F3

99

2,4,6-tFBA

C7H3O2F3

98

2,4,5-tFBA

us

an 99.5

C7H3O2F3

98

logP

140.11

3.23

1.77

140.11

3.67

1.77

140.11

3.79

1.77

158.10

2.42

1.92

158.10

2.87

1.92

158.10

2.87

1.92

158.10

3.00

1.92

158.10

3.31

1.92

158.10

3.43

1.92

176.10

2.06

2.06

176.10

2.19

2.06

176.10

2.64

2.06

176.10

2.64

2.06

176.10

3.07

2.06

190.12

3.17

2.51

190.12

3.50

2.51

190.12

3.69

2.51

ed

C7H3O2F3

pKa

3,4,5-tFBA

C7H3O2F3

98

2-tFmBA

C9H5O2F3

98

3-tFmBA

C9H5O2F3

99

4-tFmBA

C9H5O2F3

98

C7H2O2F4

99

SigmaAldrich

194.08

2.27

2.20

3,5-bisFmBA

C9H4O2F6

98

SigmaAldrich

258.12

2.97

3.39

2,3,4,5tetrafluorobenzoic acid

Ac

3, 5-bistrifluoromethylbenzoic acid

2 3

2,3,4-tFBA

Across Organics* Across Organics SigmaAldrich** Across Organics Across Organics SigmaAldrich Across Organics SigmaAldrich Across Organics SigmaAldrich SigmaAldrich Across Organics SigmaAldrich SigmaAldrich Across Organics SigmaAldrich SigmaAldrich

Mass

ce pt

2,3,6-trifluorobenzoic acid 2,4,6-trifluorobenzoic acid 2,4,5-trifluorobenzoic acid 2,3,4-trifluorobenzoic acid 3,4,5-trifluorobenzoic acid 2-trifluoromethylbenzoic acid 3-trifluoromethylbenzoic acid 4-trifluoromethylbenzoic acid

Supplier

ip t

Abbreviation

cr

Name

M

1

2,3,4,5-

tetraFBA

*Across Organics supplied by Fisher Scientific SAS,( Illkirch, France), **Sigma-Aldrich (Saint-Quentin-Fallavier, France)

4

1 Page 19 of 49

[Tapez un texte]

Table 2. Reaction monitoring parameters and operating parameters of ESI ion source

1 2 3

22 18 22 20 14 16 10 14 20 14 14 12 14 10 20 20 26 22 15 22

12 10 12 14 12 10 8 10 14 10 12 8 12 8 12 12 14 14 5 16

175.1 -> 131.1

an

us

157.1 -> 113.0

Collision [V]

ip t

144.0 -> 99.1

Cone [V]

cr

139.1 -> 95.0

M

2-FBA 3-FBA 4-FBA 4-FBAiso 2,3-dFBA 2,4-dFBA 2,6-dFBA 2,5-dFBA 3,4-dFBA 3,5-dFBA 2,3,4-tFBA 2,3,6-tFBA 2,4,5-tFBA 2,4,6-tFBA 3,4,5-tFBA 2-tFmBA 3-tFmBA 4-tFmBA 2,3,4,5-tetraFBA 3,5-bistFmBA

Ion transition

189.2 -> 145.1 193.2 -> 149.1 257.2 -> 213.1

ed

Name

ce pt

Ion source parameters

Capillary [kV] 1.4 4

Cone gas [L/h] 50

Desolvation gas [L/h] 900

Ac

5

Desolvation temp. [˚C] 550

2 Page 20 of 49

[Tapez un texte] 1 Table 3. 2

AB SCIEX TripleTOF® 6600 a,d

This method

Xevo TQb

Bruker Impact II Q-TOF MSa,d

Xevo TQc

0.07

0.20

-

0.090

3-FBA

0.09

0.02

0.086

0.150

4-FBA

0.08

0.20

0.180

0.500

2,6-dFBA

0.20

0.20

-

2,5-dFBA

0.05

0.20

0.068

2,3-dFBA

0.03

2.0

0.02

2,4-dFBA

0.03

0.20

3,5-dFBA

0.04

0.20

3,4-dFBA

0.06

0.20

2,3,6-tFBA

0.17

2.0

2,4,6-tFBA

0.13

2,4,5-tFBA

0.02

2,3,4-tFBA

0.03

cr

2-FBA

0.04

ip t

Compound

HPLC-ESI MS/MS detectionlimits (ng/mL) for FBA tracersin water using different detection systems (Acquity UPLC BEH C18 1.7 µm / 2.1 x 50 mm column).

us

0.003

0.03 0.02 0.08 0.04

0.050

0.04

0,023

0.090

0.04

0.022

0.035

0.04

0.020

0.040

0.04

0.96

3

0.08

0.20

-

0.300

0.04

0.20

0.650

1

0.02

2.0

0.31

0.500

0.04

0.03

0.20

0.29

0.900

0.03

0.1

0.20

0.072

0.100

0.02

0.1

0.20

0.030

0.039

0.04

4-tFmBA

0.09

0.20

0.031

0.100

0.04

2,3,4,5-tetraFBA

0.05

nd

0.24

0.700

0.01

3,5-bisFmBA

0.04

nd

0.0004

0,003

0.01

2-tFmBA

3 4 5

a

M

ed

Ac

3-tFmBA

ce pt

3,4,5-tFBA

an

0.500

b

c

d

10 µl injection, 50 µl injection[1], 15 µl injection[1], indicative manufacturer’s values

3 Page 21 of 49

ip t

[Tapez un texte]

Sample

SPE IV

SPE V

SPE VI

2x2 mL 2x2 mL ACN ACN

2x2 mL ACN

2x2 mL ACN

2x2 mL ACN

2x2 mL ACN

2x2 mL 2x2 mL 1%AA 1%AA

2x2 mL 1%NH4OH

2x2 mL 1%AA

2x2 mL 1%AA

4 mL 1% NH4OH

4 mL

4 mL

4 mL

4 mL 1%AA

Drying in air stream 2x2 mL 2x2 mL ACN ACN

-

2x2 mL ACN:1%NH4OH (8:2)

SPE VIII

SPE IX

SPE X

SPE XI

SPE XII

2x2 mL THF

2x2 mL THF

2x2 mL ACN

2x2 mL ACN

2x2 mL THF

2x2 mL THF

2x2 mL H2O

2x2 mL H2O

2x2 mL H2O

2x2 mL H2O

2x2 mL H2O

2x2 mL H2O

2x2 mL H2O

4 mL

4 mL

4 mL

4 mL

4 mL

4 mL

2x2 mL THF

2x2ml ACN:1%NH4OH (5:5)

2x2ml ACN:10%NH4OH (8:2)

4 mL

2x2 mL ACN 1%AA

2 mL H2O** 4 min

2x2 mL ACN:1%NH4OH (8:2)

ep te

Elution

SPE VII

d

Rinsing

us

SPE III*

an

Conditioning

SPE II

M

SPE I

cr

Table 4. Experimental conditions of the SPE procedures tested

2x2ml THF:1%NH4OH (8:2)

2x2ml 2x2ml THF:1%NH4OH THF:10%NH4OH (5:5) (8:2)

Evaporation to dryness

Dissolving of residue in 1 mL of mobile phase

Ac c

* idea of the procedure was based on cleaning the sample without adsorption of analytes ** this step was in all the procedures except SPE III

ACN – acetonitrile, AA – acetic acid, NH4OH – ammonia, THF - tetrahydrofuran

4 Page 22 of 49

[Tapez un texte]

Table 5. Recoveries of FBA standards from water samples by SPE in the optimal conditions (cf.Procedure) at the different concentration levels.

Recovery of

Recovery of

0.2 ng/mL,

1ng/mL

10 ng/mL

% (SD, n=3)

% (SD, n=3)

% (SD, n=3)

2-FBA

90 (2.7)

94 (3.4)

3-FBA

95 (4.2)

96 (2.3)

4-FBA

105 (4.9)

96 (4.5)

2,6-dFBA*

52 (3.2)

51 (2.5)

2,5-dFBA

99 (1.4)

2,3-dFBA

94 (3.9)

102 (2.1)

cr

94 (3.7)

us

49 (4.0) 96 (1.2)

103 (1.8)

98 (3.9)

96 (3.9)

98 (1.2)

104 (3.3)

93 (2.2)

90 (1.8)

90 (3.5)

95 (4.1)

93 (4.6)

92 (3.4)

112 (2.9)

108 (3.5)

106 (4.1)

76 (4.5)

82 (3.6)

84 (2.8)

2,4,5-tFBA

97 (2.8)

101 (2.3)

103 (1.9)

2,3,4-tFBA

96 (2.7)

102 (2.0)

97 (3.6)

ce pt

2,4-dFBA

99 (3.4)

98 (3.2)

an

Compound

ip t

Recovery of

3,4,5-tFBA

93 (4.2)

95 (5.1)

93 (2.7)

2-tFmBA

92 (3.4)

88 (3.7)

90 (2.0)

3-tFmBA

87 (2.3)

92 (2.6)

89 (4.1)

4-tFmBA

94 (3.0)

95 (2.5)

94 (4.0)

2,3,4,5-tetraFBA

98 (3.4)

103 (2.4)

108 (5.5)

3,5-bisFmBA

94 (3.7)

100 (2.3)

101 (2.8)

3,4-dFBA 2,3,6-tFBA*

M

3,5-dFBA

Ac

ed

2,4,6-tFBA*

, n - number of measurements * early eluting compounds

5 Page 23 of 49

[Tapez un texte]

Table 6. Linearity, detection and quantification limits for the method developed applied to a reservoir water (source Quatar, >20% salt) Calibration curve equation for 1/x (8 points, n=3)

Sa

Sb

R

LOD [ng/mL]

LOQ [ng/mL]

2-FBA

y=8804x - 70

40

75

0.9987

0.03

0.09

3-FBA

y=16595x + 3262

104

162 0.9991

0.03

0.09

4-FBA

y=12234x + 936

59

89

0.02

0.06

2,6-dFBA*

y=15951x + 540

168

187 0.9986

0.04

0.12

2,5-dFBA

y=57762x + 2495

853

336 0.9998

0.02

0.06

2,3-dFBA

y=34310x + 820

140

224 0.9986

0.02

0.06

2,4-dFBA

y=53965x + 1117

251

311 0.9997

0.02

0.06

3,5-dFBA

y=79825x + 3508

416

324 0.9999

0.01

0.03

3,4-dFBA

y=69755x + 3231

877

287 0.9993

0.01

0.03

2,3,6-tFBA*

y=6518x + 230

64

84

0.9984

0.04

0.12

2,4,6-tFBA*

y=4986x – 65

11

55

0.9987

0.04

0.12

y=98181x + 3296

899

614 0.9995

0.02

0.06

y=91303x + 2057

1507 284 0.9991

0.01

0.03

1662 452 0.9989

0.01

0.03

us

cr

0.9988

an

M

ce pt

2,3,4-tFBA

ed

2,4,5-tFBA

2

ip t

Name

y=115567x + 2969

2-tFmBA

y=45379x + 6555

81

481 0.9997

0.03

0.09

3-tFmBA

y=129965x + 5599

152 1021 0.9999

0.03

0.09

4-tFmBA

y=95547x + 3384

265

841 0.9998

0.03

0.09

2,3,4,5-tetraFBA

y=8512x + 691

28

77

0.9998

0.03

0.09

3,5-bisFmBA

y=129169x + 8247

955

755 0.9997

0.02

Ac

3,4,5-tFBA

0.06 2

Sa - standard deviation of the slope, Sb - standard deviation of the intercept, R - coefficient of determination, LOD - limit of detection, LOQ - limit of quantitation, n - number of measurements * early eluting compounds

6 Page 24 of 49

[Tapez un texte]

Table 7. Recoveries of the most polar compounds and their correction using dedicated deuterated internal standards. Concentration added: 10 ng/mL

Recovery CV % (n=3) with 4FBAiso

Recovery CV % (n=3) with 26dFBAiso

Recovery with CV % (n=3) 246tFBAiso

2,6-dFBA

61.2 (3.8)

92.5 (4.2)

-

2,3,6-tFBA

113.4 (5.1)

-

94.1 (2.4)

2,4,6-tFBA

69.6 (3.5)

-

96.2 (3.9)

Ac

ce pt

ed

M

an

us

cr

ip t

Name

7 Page 25 of 49

[Tapez un texte]

Validation of the SPE-HPLC-MS/MS method developed for synthetic samples [blank reservoir water (ca. 20% salt) with FBA tracers spiked at 3 different concentrations].

2,5-dFBA

2,3-dFBA

2,4-dFBA

3,5-dFBA

3,4-dFBA

2,3,6-tFBA

ce pt

2,4-6tFBA*

2,4,5-tFBA

2,3,4-tFBA

Ac

3,4,5-tFBA

2-tFmBA 3-tFmBA

4-tFmBA

2,3,4,5-tetraFBA

3,5-bisFmBA

ip t

2,6-dFBA*

Recovery [%] 90 94 99 95 96 102 105 96 94 91 88 93 99 98 96 94 103 98 96 98 104 93 90 90 95 93 92 112 108 106 103 92 96 97 101 103 96 102 97 93 95 93 92 88 90 87 92 89 94 95 104 98 103 108 94 100 101

cr

4-FBA

us

3-FBA

Found [ng/mL] ± SD 0.180 ± 0.005 0.94 ± 0.03 9.9 ± 0.3 0.190 ± 0.008 0.96 ± 0.02 10.2 ± 0.2 0.210 ± 0.009 0.96 ± 0.05 9.4 ± 0.4 0.182 ± 0.007 0.88 ± 0.04 9.3 ± 0.4 0.198 ± 0.002 0.98 ± 0.03 9.6 ± 0.1 0.188 ± 0.007 1.03 ± 0.02 9.8 ± 0.4 0.192 ± 0.007 0.98 ± 0.01 10.4 ± 0.3 0.186 ± 0.004 0.90 ± 0.02 9.0 ± 0.4 0.190 ± 0.008 0.93 ± 0.05 9.2 ± 0.3 0.224 ± 0.005 1.08 ± 0.04 10.6 ± 0.4 0.206 ± 0.005 0.92 ± 0.06 9.6 ± 0.4 0.194 ± 0.006 1.01 ± 0.02 10.3 ± 0.2 0.192 ± 0.005 1.02 ± 0.02 9.7 ± 0.4 0.186 ± 0.008 0.95 ± 0.05 9.3 ± 0.3 0.184 ± 0.006 0.88 ± 0.04 9 ± 0.2 0.174 ± 0.005 0.92 ± 0.03 8.9 ± 0.4 0.188 ± 0.006 0.95 ± 0.03 10.4 ± 0.4 0.196 ± 0.007 1.03 ± 0.02 10.8 ± 0.6 0.188 ± 0.007 1.00 ± 0.02 10.1 ± 0.3

an

2-FBA

Added [ng/ml] 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10

M

Compound

ed

Table 8.

* early eluting compounds were quantified with their corresponding internal standards

8 Page 26 of 49

[Tapez un texte]

Ac

ce pt

ed

M

an

us

cr

ip t

[1] C. Serres-Piole, N. Moradi-Tehrani, R. Lobinski, H. Preud'homme, Direct sensitive simultaneous determination of fluorinated benzoic acids in oil reservoir waters by ultra highperformance liquid chromatography-tandem mass spectrometry, Journal of Chromatography A, 1218 (2011) 5872-5877.

9 Page 27 of 49

Ac ce p

te

d

M

an

us

cr

ip t

Electronic Supplementary Material (online publication only)

Page 28 of 49

Miscellaneous

[Tapez un texte]

Table 1. Standard compounds used in this study Formula

Purity [%]

2-fluorobenzoic acid

2-FBA

C7H5O2F

99

3-fluorobenzoic acid

3-FBA

C7H5O2F

99

4-fluorobenzoic acid

4-FBA

C7H5O2F

98

2,6-difluorobenzoic acid

2,6-dFBA

C7H4O2F2

98

2,5-difluorobenzoic acid

2,5-dFBA

C7H4O2F2

98

2,3-difluorobenzoic acid

2,3-dFBA

C7H4O2F2

98

2,4-difluorobenzoic acid

2,4-dFBA

C7H4O2F2

99

3,5-difluorobenzoic acid

3,5-dFBA

C7H4O2F2

97

3,4- difluorobenzoic acid

3,4-dFBA

C7H4O2F2

99

2,3,6-tFBA

C7H3O2F3

99

2,4,6-tFBA

C7H3O2F3

98

2,4,5-tFBA

us

an 99.5

C7H3O2F3

98

logP

140.11

3.23

1.77

140.11

3.67

1.77

140.11

3.79

1.77

158.10

2.42

1.92

158.10

2.87

1.92

158.10

2.87

1.92

158.10

3.00

1.92

158.10

3.31

1.92

158.10

3.43

1.92

176.10

2.06

2.06

176.10

2.19

2.06

176.10

2.64

2.06

176.10

2.64

2.06

176.10

3.07

2.06

190.12

3.17

2.51

190.12

3.50

2.51

190.12

3.69

2.51

ed

C7H3O2F3

pKa

3,4,5-tFBA

C7H3O2F3

98

2-tFmBA

C9H5O2F3

98

3-tFmBA

C9H5O2F3

99

4-tFmBA

C9H5O2F3

98

C7H2O2F4

99

SigmaAldrich

194.08

2.27

2.20

3,5-bisFmBA

C9H4O2F6

98

SigmaAldrich

258.12

2.97

3.39

2,3,4,5tetrafluorobenzoic acid

Ac

3, 5-bistrifluoromethylbenzoic acid

2 3

2,3,4-tFBA

Across Organics* Across Organics SigmaAldrich** Across Organics Across Organics SigmaAldrich Across Organics SigmaAldrich Across Organics SigmaAldrich SigmaAldrich Across Organics SigmaAldrich SigmaAldrich Across Organics SigmaAldrich SigmaAldrich

Mass

ce pt

2,3,6-trifluorobenzoic acid 2,4,6-trifluorobenzoic acid 2,4,5-trifluorobenzoic acid 2,3,4-trifluorobenzoic acid 3,4,5-trifluorobenzoic acid 2-trifluoromethylbenzoic acid 3-trifluoromethylbenzoic acid 4-trifluoromethylbenzoic acid

Supplier

ip t

Abbreviation

cr

Name

M

1

2,3,4,5-

tetraFBA

*Across Organics supplied by Fisher Scientific SAS,( Illkirch, France), **Sigma-Aldrich (Saint-Quentin-Fallavier, France)

4

1 Page 29 of 49

[Tapez un texte]

Table 2. Reaction monitoring parameters and operating parameters of ESI ion source

1 2 3

22 18 22 20 14 16 10 14 20 14 14 12 14 10 20 20 26 22 15 22

12 10 12 14 12 10 8 10 14 10 12 8 12 8 12 12 14 14 5 16

175.1 -> 131.1

an

us

157.1 -> 113.0

Collision [V]

ip t

144.0 -> 99.1

Cone [V]

cr

139.1 -> 95.0

M

2-FBA 3-FBA 4-FBA 4-FBAiso 2,3-dFBA 2,4-dFBA 2,6-dFBA 2,5-dFBA 3,4-dFBA 3,5-dFBA 2,3,4-tFBA 2,3,6-tFBA 2,4,5-tFBA 2,4,6-tFBA 3,4,5-tFBA 2-tFmBA 3-tFmBA 4-tFmBA 2,3,4,5-tetraFBA 3,5-bistFmBA

Ion transition

189.2 -> 145.1 193.2 -> 149.1 257.2 -> 213.1

ed

Name

ce pt

Ion source parameters

Capillary [kV] 1.4 4

Cone gas [L/h] 50

Desolvation gas [L/h] 900

Ac

5

Desolvation temp. [˚C] 550

2 Page 30 of 49

[Tapez un texte] 1 Table 3. 2

AB SCIEX TripleTOF® 6600 a,d

This method

Xevo TQb

Bruker Impact II Q-TOF MSa,d

Xevo TQc

0.07

0.20

-

0.090

3-FBA

0.09

0.02

0.086

0.150

4-FBA

0.08

0.20

0.180

0.500

2,6-dFBA

0.20

0.20

-

2,5-dFBA

0.05

0.20

0.068

2,3-dFBA

0.03

2.0

0.02

2,4-dFBA

0.03

0.20

3,5-dFBA

0.04

0.20

3,4-dFBA

0.06

0.20

2,3,6-tFBA

0.17

2.0

2,4,6-tFBA

0.13

2,4,5-tFBA

0.02

2,3,4-tFBA

0.03

cr

2-FBA

0.04

ip t

Compound

HPLC-ESI MS/MS detectionlimits (ng/mL) for FBA tracersin water using different detection systems (Acquity UPLC BEH C18 1.7 µm / 2.1 x 50 mm column).

us

0.003

0.03 0.02 0.08 0.04

0.050

0.04

0,023

0.090

0.04

0.022

0.035

0.04

0.020

0.040

0.04

0.96

3

0.08

0.20

-

0.300

0.04

0.20

0.650

1

0.02

2.0

0.31

0.500

0.04

0.03

0.20

0.29

0.900

0.03

0.1

0.20

0.072

0.100

0.02

0.1

0.20

0.030

0.039

0.04

4-tFmBA

0.09

0.20

0.031

0.100

0.04

2,3,4,5-tetraFBA

0.05

nd

0.24

0.700

0.01

3,5-bisFmBA

0.04

nd

0.0004

0,003

0.01

2-tFmBA

3 4 5

a

M

ed

Ac

3-tFmBA

ce pt

3,4,5-tFBA

an

0.500

b

c

d

10 µl injection, 50 µl injection[1], 15 µl injection[1], indicative manufacturer’s values

3 Page 31 of 49

ip t

[Tapez un texte]

Sample

SPE IV

SPE V

SPE VI

2x2 mL 2x2 mL ACN ACN

2x2 mL ACN

2x2 mL ACN

2x2 mL ACN

2x2 mL ACN

2x2 mL 2x2 mL 1%AA 1%AA

2x2 mL 1%NH4OH

2x2 mL 1%AA

2x2 mL 1%AA

4 mL 1% NH4OH

4 mL

4 mL

4 mL

4 mL 1%AA

Drying in air stream 2x2 mL 2x2 mL ACN ACN

-

2x2 mL ACN:1%NH4OH (8:2)

SPE VIII

SPE IX

SPE X

SPE XI

SPE XII

2x2 mL THF

2x2 mL THF

2x2 mL ACN

2x2 mL ACN

2x2 mL THF

2x2 mL THF

2x2 mL H2O

2x2 mL H2O

2x2 mL H2O

2x2 mL H2O

2x2 mL H2O

2x2 mL H2O

2x2 mL H2O

4 mL

4 mL

4 mL

4 mL

4 mL

4 mL

2x2 mL THF

2x2ml ACN:1%NH4OH (5:5)

2x2ml ACN:10%NH4OH (8:2)

4 mL

2x2 mL ACN 1%AA

2 mL H2O** 4 min

2x2 mL ACN:1%NH4OH (8:2)

ep te

Elution

SPE VII

d

Rinsing

us

SPE III*

an

Conditioning

SPE II

M

SPE I

cr

Table 4. Experimental conditions of the SPE procedures tested

2x2ml THF:1%NH4OH (8:2)

2x2ml 2x2ml THF:1%NH4OH THF:10%NH4OH (5:5) (8:2)

Evaporation to dryness

Dissolving of residue in 1 mL of mobile phase

Ac c

* idea of the procedure was based on cleaning the sample without adsorption of analytes ** this step was in all the procedures except SPE III

ACN – acetonitrile, AA – acetic acid, NH4OH – ammonia, THF - tetrahydrofuran

4 Page 32 of 49

[Tapez un texte]

Table 5. Recoveries of FBA standards from water samples by SPE in the optimal conditions (cf.Procedure) at the different concentration levels.

Recovery of

Recovery of

0.2 ng/mL,

1ng/mL

10 ng/mL

% (SD, n=3)

% (SD, n=3)

% (SD, n=3)

2-FBA

90 (2.7)

94 (3.4)

3-FBA

95 (4.2)

96 (2.3)

4-FBA

105 (4.9)

96 (4.5)

2,6-dFBA*

52 (3.2)

51 (2.5)

2,5-dFBA

99 (1.4)

2,3-dFBA

94 (3.9)

102 (2.1)

cr

94 (3.7)

us

49 (4.0) 96 (1.2)

103 (1.8)

98 (3.9)

96 (3.9)

98 (1.2)

104 (3.3)

93 (2.2)

90 (1.8)

90 (3.5)

95 (4.1)

93 (4.6)

92 (3.4)

112 (2.9)

108 (3.5)

106 (4.1)

76 (4.5)

82 (3.6)

84 (2.8)

2,4,5-tFBA

97 (2.8)

101 (2.3)

103 (1.9)

2,3,4-tFBA

96 (2.7)

102 (2.0)

97 (3.6)

ce pt

2,4-dFBA

99 (3.4)

98 (3.2)

an

Compound

ip t

Recovery of

3,4,5-tFBA

93 (4.2)

95 (5.1)

93 (2.7)

2-tFmBA

92 (3.4)

88 (3.7)

90 (2.0)

3-tFmBA

87 (2.3)

92 (2.6)

89 (4.1)

4-tFmBA

94 (3.0)

95 (2.5)

94 (4.0)

2,3,4,5-tetraFBA

98 (3.4)

103 (2.4)

108 (5.5)

3,5-bisFmBA

94 (3.7)

100 (2.3)

101 (2.8)

3,4-dFBA 2,3,6-tFBA*

M

3,5-dFBA

Ac

ed

2,4,6-tFBA*

, n - number of measurements * early eluting compounds

5 Page 33 of 49

[Tapez un texte]

Table 6. Linearity, detection and quantification limits for the method developed applied to a reservoir water (source Quatar, >20% salt) Calibration curve equation for 1/x (8 points, n=3)

Sa

Sb

R

LOD [ng/mL]

LOQ [ng/mL]

2-FBA

y=8804x - 70

40

75

0.9987

0.03

0.09

3-FBA

y=16595x + 3262

104

162 0.9991

0.03

0.09

4-FBA

y=12234x + 936

59

89

0.02

0.06

2,6-dFBA*

y=15951x + 540

168

187 0.9986

0.04

0.12

2,5-dFBA

y=57762x + 2495

853

336 0.9998

0.02

0.06

2,3-dFBA

y=34310x + 820

140

224 0.9986

0.02

0.06

2,4-dFBA

y=53965x + 1117

251

311 0.9997

0.02

0.06

3,5-dFBA

y=79825x + 3508

416

324 0.9999

0.01

0.03

3,4-dFBA

y=69755x + 3231

877

287 0.9993

0.01

0.03

2,3,6-tFBA*

y=6518x + 230

64

84

0.9984

0.04

0.12

2,4,6-tFBA*

y=4986x – 65

11

55

0.9987

0.04

0.12

y=98181x + 3296

899

614 0.9995

0.02

0.06

y=91303x + 2057

1507 284 0.9991

0.01

0.03

1662 452 0.9989

0.01

0.03

us

cr

0.9988

an

M

ce pt

2,3,4-tFBA

ed

2,4,5-tFBA

2

ip t

Name

y=115567x + 2969

2-tFmBA

y=45379x + 6555

81

481 0.9997

0.03

0.09

3-tFmBA

y=129965x + 5599

152 1021 0.9999

0.03

0.09

4-tFmBA

y=95547x + 3384

265

841 0.9998

0.03

0.09

2,3,4,5-tetraFBA

y=8512x + 691

28

77

0.9998

0.03

0.09

3,5-bisFmBA

y=129169x + 8247

955

755 0.9997

0.02

Ac

3,4,5-tFBA

0.06 2

Sa - standard deviation of the slope, Sb - standard deviation of the intercept, R - coefficient of determination, LOD - limit of detection, LOQ - limit of quantitation, n - number of measurements * early eluting compounds

6 Page 34 of 49

[Tapez un texte]

Table 7. Recoveries of the most polar compounds and their correction using dedicated deuterated internal standards. Concentration added: 10 ng/mL

Recovery CV % (n=3) with 4FBAiso

Recovery CV % (n=3) with 26dFBAiso

Recovery with CV % (n=3) 246tFBAiso

2,6-dFBA

61.2 (3.8)

92.5 (4.2)

-

2,3,6-tFBA

113.4 (5.1)

-

94.1 (2.4)

2,4,6-tFBA

69.6 (3.5)

-

96.2 (3.9)

Ac

ce pt

ed

M

an

us

cr

ip t

Name

7 Page 35 of 49

[Tapez un texte]

Validation of the SPE-HPLC-MS/MS method developed for synthetic samples [blank reservoir water (ca. 20% salt) with FBA tracers spiked at 3 different concentrations].

2,5-dFBA

2,3-dFBA

2,4-dFBA

3,5-dFBA

3,4-dFBA

2,3,6-tFBA

ce pt

2,4-6tFBA*

2,4,5-tFBA

2,3,4-tFBA

Ac

3,4,5-tFBA

2-tFmBA 3-tFmBA

4-tFmBA

2,3,4,5-tetraFBA

3,5-bisFmBA

ip t

2,6-dFBA*

Recovery [%] 90 94 99 95 96 102 105 96 94 91 88 93 99 98 96 94 103 98 96 98 104 93 90 90 95 93 92 112 108 106 103 92 96 97 101 103 96 102 97 93 95 93 92 88 90 87 92 89 94 95 104 98 103 108 94 100 101

cr

4-FBA

us

3-FBA

Found [ng/mL] ± SD 0.180 ± 0.005 0.94 ± 0.03 9.9 ± 0.3 0.190 ± 0.008 0.96 ± 0.02 10.2 ± 0.2 0.210 ± 0.009 0.96 ± 0.05 9.4 ± 0.4 0.182 ± 0.007 0.88 ± 0.04 9.3 ± 0.4 0.198 ± 0.002 0.98 ± 0.03 9.6 ± 0.1 0.188 ± 0.007 1.03 ± 0.02 9.8 ± 0.4 0.192 ± 0.007 0.98 ± 0.01 10.4 ± 0.3 0.186 ± 0.004 0.90 ± 0.02 9.0 ± 0.4 0.190 ± 0.008 0.93 ± 0.05 9.2 ± 0.3 0.224 ± 0.005 1.08 ± 0.04 10.6 ± 0.4 0.206 ± 0.005 0.92 ± 0.06 9.6 ± 0.4 0.194 ± 0.006 1.01 ± 0.02 10.3 ± 0.2 0.192 ± 0.005 1.02 ± 0.02 9.7 ± 0.4 0.186 ± 0.008 0.95 ± 0.05 9.3 ± 0.3 0.184 ± 0.006 0.88 ± 0.04 9 ± 0.2 0.174 ± 0.005 0.92 ± 0.03 8.9 ± 0.4 0.188 ± 0.006 0.95 ± 0.03 10.4 ± 0.4 0.196 ± 0.007 1.03 ± 0.02 10.8 ± 0.6 0.188 ± 0.007 1.00 ± 0.02 10.1 ± 0.3

an

2-FBA

Added [ng/ml] 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10 0.200 1 10

M

Compound

ed

Table 8.

* early eluting compounds were quantified with their corresponding internal standards

8 Page 36 of 49

[Tapez un texte]

Ac

ce pt

ed

M

an

us

cr

ip t

[1] C. Serres-Piole, N. Moradi-Tehrani, R. Lobinski, H. Preud'homme, Direct sensitive simultaneous determination of fluorinated benzoic acids in oil reservoir waters by ultra highperformance liquid chromatography-tandem mass spectrometry, Journal of Chromatography A, 1218 (2011) 5872-5877.

9 Page 37 of 49

Miscellaneous

[Tapez un texte]

2

Sensitive simultaneous determination of 19 fluorobenzoic acids in saline waters by solid-phase extraction and LC-MS/MS

3

Paweł Kubica,aHervé Garraudb, Joanna Szpunarc* and Ryszard Lobinskic,d

1

4

7 8 9 10

Department of Analytical Chemistry, Faculty of Chemistry, Gdansk University of Technology, 11/12 Narutowicza Str, Gdańsk, Poland b

SOBEGI, Laboratoire Contrôle et Environnement, Pôle 4, Av. du Lac, 64150 Mourenx, France

ip t

6

a

c

CNRS/UPPA, Laboratoire de Chimie Analytique Bio-inorganique et Environnement (LCABIE-IPREM), Hélioparc, 2, AvenuePr. Angot, 64053 Pau, France

cr

5

11 12

d

13

[email protected], tel: +33 559 40 77 55, fax: +33 559 40 77 82

an

us

Department of Analytical Chemistry , Warsaw University of Technology, ul. Noakowskiego 3, 00-664 Warsaw, Poland

Ac

ce pt

ed

M

14

1 Page 38 of 49

[Tapez un texte]

Abstract

16

30

A solid-phase extraction (SPE) procedure using C18 stationary phase was optimized for the preconcentration of 19 fluorinated derivatives of benzoic acid (FBA): mono-, ditri- and tetrafluorosubstituted in the ring, trifluoromethylbenzoic acid and 3,5bistrifluoromethyl benzoic acid from undiluted salt-rich (>20%) reservoir waters. Quantitative (>90%) retention/elution of 16 out of 19 analyte compounds was achieved allowing a 4-fold preconcentration factor accompanied by the elimination of >99% of salt. For the three most polar compounds (2,6-dFBA, 2,3,6-tFBA and 2,4,6tFBA) the non-quantitative recoveries(>70%)were corrected by dedicated customsynthesized deuterated internal standards. The FBAs were determined by HPLC MS/MS revisited in terms of a choice of column, elution conditions and MS/MSsignal acquisitionparameters allowing the baseline separation and a gain in sensitivity. For a sample intake of 4 mL, detection limits for all the compounds in a reservoir water sample containing more than 20% salt were between 0.01 and 0.05 ng/ml which represents a gain of a factor of 10-20 in comparison with the state-of the art LCMS/MS procedures for samples of similar complexity.

31

Keywords:fluorobenzoic acids, solid-phase extraction, reservoir water, LC MS/MS

20 21 22 23 24 25 26 27 28 29

cr

19

us

18

an

17

ip t

15

M

32

Introduction:

34

Derivatives of benzoic acid with one or more fluorine atoms, or one or more trifluoromethyl groups,attached to the aromatic ring are the most common currently used non-radioactive passive water tracers for oil field applications [1]. As a tracing campaign involves a set of several different compounds (out of more than 20 commercially available), there is a need for methods for their simultaneous determination in an oil reservoir water matrix. Low detection limits are critical as they determine the quantity of the tracers necessary to be used and thus the cost and the environmental impact of the campaign. The matrix differs depending on the sample origin but it is usually rich in salts (reaching in some cases up to 30%) and organic constituents [2].

37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52

ce pt

36

Ac

35

ed

33

The lowest detection limits (down to 0.01 ng/ml) were obtained by gas chromatography (GC)- MS but lengthy (24 h) and tedious sample preparation procedures including matrix removal and derivatization were necessary[3]. The incomplete and strongly compound-dependent yields required compound specific isotope dilution calibration that was proposed for sixspecies determined to achieve accurate analysis. [4],[5]. The alternative is the use of HPLC - MS/MS analysis to eliminate the derivatization step and thus to simplify the sample processing. The original work [5 ], which was applied to simple matrices butdid not show any chromatogram reported fairly high 2 Page 39 of 49

[Tapez un texte]

61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90

Although the reported selectivity of HPLC separation of a set of usually studied 20 tracers wasgenerally high, the baseline separation of all of them was not achieved in any of the published works [3-9]. This caveat was compensated by the determination of the co-eluting compounds using different fragmentation reactions. On the other hand, the number of theoretical plates achieved in HPLC is important. Indeed, the poor specificity of fragmentation reactions (the loss of CO2) used for the quantification, in combination with the unit resolution of a quadrupole filter and matrix rich in organic acids, may lead to the increase in baseline and false positives.

ip t

60

cr

59

us

58

The above reasons spur the need for the development of methods allowing a considerable enrichment of FBAs with regard to salt and organic matrix. Solid phase extraction (SPE) is an attractive option for both matrix removal and preconcentration of analytes [10-12]prior to LC-MS/MS analysis of samples rich in salts. However, quantitative SPE of FBAs from reservoir waters is a difficult task because of the high polarity of the tracers. The problems result, on one hand, from the difficulty to trap quantitatively and simultaneously all the analytes while avoiding the retention of the matrix and, on the other hand, to release the trapped analytes quantitatively without substantial dilution. Another critical factor is the sample volume to be used for analysis as it determines the SPE time.

an

57

M

56

ed

55

ce pt

54

detection limits: 0.5-1 ng/ml for electrospray ionization (ESI) and 10-20 ng/ml for atmospheric pressure chemical ionization (APCI), respectively. The detection limits were considerably (about an order of magnitude) decreased by Serres-Pioles et al.[1] except for tFBA, for which hardly any improvement was observed. The maximum tolerated salt content of the samples allowed by the method was pretty low (1%) which required a considerable sample dilution (10-20 times) drastically limiting the scope of the method applications.

As a result of an extensive optimization study, Müller et al.reported fairly satisfactory recoveries (between71% (2,5-dFBA) and 94 % (3-FBA))from tap water [7]but for reservoir waters the extraction efficiencies were generally low (down to 18% for 2,3,5,6-tetraFBA and 2,6-dFBA)and strongly compound-dependent [3]. Moreover, relatively large sample volumes (100 ml) processed [3, 7] resulted in long analysis times. The recovery problems were (for sixselected compounds) addressed by the use of custom synthetized deuterated internal standards[4]which were used in the analysis of reservoir and ground water [8].

Ac

53

The main goal of this work was the development of a rapid (small sample volume) quantitative SPE method allowing a direct multi-tracer (19 compounds) analysis in salt-rich (>20% salt) reservoir water samples with an objective to reach at least an order of magnitude in terms of detection limits over the direct injection procedure [1].

91 3 Page 40 of 49

[Tapez un texte]

Experimental conditions

93

SamplesCollection.Reservoir water samples of different origins with different salt contents: Gabon (200 g/l), Qatar (220 g/l), Russia (170 g/l), Yemen (80 g/l) and Congo (250 g/l) were used for the method development. The salts components were primary sodium and calcium with minor contribution of potassium and magnesium [2]. The samples were collected in 5-L glass flasks and the aqueous and organic fractions were separated by gravitation. Sub-samples of 100 mL were transported in ambient temperature in glass flasks in containers preventing the exposure to light; the samples were acidified to pH 2-3 with formic acid and stored prior to analysis at 4°C in dark;in these conditions they were stable at least 90 days.

101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127

cr

100

Reagents and standards.Acetonitrile, acetic acid, tetrahydrofuran, ammonia aq. were purchased from Sigma-Aldrich (Saint-Quentin-Fallavier, France). Ultrapure water (18 MΩ.cm) was obtained from a Milli-Q system (Millipore, Bedford, MA). The characteristics of the FBA standards used in this study are listed in Table 1.Deuterated 2,6-dFBA and 2,4,6-tFBA were a gift from Dr. K. Müller and Prof.Dr.A. Seubert (Faculty of Chemistry, Philipps-Universität, Marburg, Germany).4fluorobenzoic acid-α-13C-2,3,5,6-d4 was purchased from Sigma-Aldrich (SaintQuentin-Fallavier, France).

us

99

an

98

M

97

Materials. The SPE disposable cartridges (C18, 500 mg, 3 mL) were supplied by SigmaAldrich (Saint-Quentin-Fallavier, France). Separations were carried out using an Acquity UPLC BEH C18 column (150 mm x 2.1 mm x1.7 µm) with a matching precolumnAcquity UPLC BEH C18 VanGuard (130Å, 1.7 µm, 2.1 mm X 5 mm)(Waters, Guyancourt, France).

ed

96

ce pt

95

Instrumentation. SPE was carried out using aSupelco VisiPrep 24DL(supplied by Sigma-Aldrich).Eluates wereevaporated to dryness using an Eppendorf Concentrator Plus(Eppendorf France SAS, Montesson).An Acquity UPLC system (Waters) including a binary solvent pump, a cooled autosampler and a column oven was used. The detector was a XevoTQ (quadrupole-T-wave-quadrupole) MS with an orthogonal Zspray-electrospray interface (Waters).

Ac

94

ip t

92

Procedures

Initial sample preparation procedure.Samples were filtered through 0.2 µm (13mm) syringe filter, GHP Acrodisc(Interchim, Montluçon, France)). 4-fluorobenzoic acid-α13 C-2,3,5,6-d4) was added at 20 ng/mL as an internal standard. Deuterated 2,6-dFBA and 2,4,6-tFBA were added at 20 ng/mL if the corresponding compounds were to be determined.

4 Page 41 of 49

[Tapez un texte]

135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152

ip t

134

Measurement conditions. A 50 µL aliquot was analyzed by HPLC - MS/MS.Mobile phase was composed by mixing 0.05% CH3COOH (A) and 0.05% CH3COOH in acetonitrile (B). The elution gradient was: 0 min (13% B), 1.3 min (13% B), 9 min (28 % B) and 13 min (80 % B). The column was equilibrated for 5 min. The flow rate was 0.45 ml/min, the column temperature was 45°C and the autosampler temperature was 5°C.Tandem MS data acquisition was performed with the electrospray source operating in negative mode (ESIneg) under the MRM conditions listed in Table 2.

cr

133

us

132

an

131

Calibration. A calibration curve was constructed by plotting peak area for 7 concentrations(0.05, 0.1, 0.2, 0.5, 1, 10, 20 ng/mL).

M

130

Data processing. The Masslynx software (Waters, Milford, MA) was used to process data.

ed

129

Solid-phase extraction.The SPE cartridges were conditioned with two successive 2-ml volumes of acetonitrile followed by rinsing with two successive 2-ml volumes of water. Then, the sample was loaded as two successive 2-mL aliquots. After loading of the sample, the sorbent was rinsed with a 2-mL volume of water to remove remaining salts and polar compounds. The cartridge was dried for 3 min under the gentle stream of nitrogen (purity 99.999 %). Then, the elution was performed with two successive 2-mL volumes of acetonitrile: 10%NH4OHaq.(8:2 v/v).Thefirst portion of the eluent was kept for 3 min to facilitate the desorption of analytes.The eluate was collected and evaporated to dryness under vacuum. The residue was dissolved in 1 mL of 10% (v/v) acetonitrile and analyzed by HPLC - MS/MS.

Quality control and assurance. For the purpose of method validation three samples were prepared by spiking a reservoir water (salinity 22%) at the different concentration levels: 0.2, 1 and 10 ng/mL, respectively. The samples were analysed by procedure developed.

ce pt

128

153

155 156 157 158 159 160 161 162 163 164

Results and discussion

Ac

154

LC - MS/MS determination of FBAs The separation methods reported in the literature were based on isocratic elution inion-chromatography [7] or C18 reversed phase chromatography [5]. An improved selectivity in reversed-phase HPLC was obtained by gradient elution with slightly acidic methanol or acetonitrile [9]. The latter procedure was the starting point for the optimization of the HPLC separation conditions in this work. In order to obtain the baseline separation and to reduce the co-elution with matrix components, the length of the column was increased which tripled the number of theoretical plates in comparison tothe former work [9]and the baseline separation of all the 19 FBAs to be achievedwithin 13 min as shown in Fig. 1. 5 Page 42 of 49

[Tapez un texte]

169 170 171 172 173 174 175 176 177 178 179 180 181

ip t

168

Table 3 also shows that the obtained detection limits were in one case lower than, and in one case comparable with, the indicative values received from the manufacturers for QTOF systems operated in the MRM mode. The higher resolution of QTOF may offer an advantage of reducing the risk of false positives in the case of more complex samples. On the other hand, the range of linearity of the triple quadrupole spectrometer was an order of magnitude larger than that of the TOF instruments.

cr

167

us

166

The calibration curvesshowed good linearity (r2>0.999) and precision below 3% (n=3) (as shown inTable 1 Supplementary Information). The detection limits calculated as 3x standard deviation of blank integrated at the corresponding retention times are summarized in Table 3. In the absence of sample matrix, the LODs depend,in particular, on the ionization efficiency. The latter waslargely affected by the low content of the organic modifier for the early eluting species (2,6-dFBA, 2,3,6-tFBA, 2,4,6tFBA and 2,3,4,5-tetraFBA) for which relatively high LODs were observed. In general, the LODs compare favorably with those published elsewhere for LC-based methods[5,7,9].The most spectacular gain (10-fold) was obtained for the triFBAswhich are very sensitive to ionization conditions.

an

165

M

182

Optimisation of SPE conditions

184

Müller at al. [3] published a comprehensive comparison study of five different SPE materials tested in a broad pH range (1-11); the best results were obtained for two of them: Oasis HLB-Plus (hydrophilic-lipophilic-balanced reversed-phase poly(divinylbenzene-co-N-vinylpyrrolidone sorbent) and Isolute ENV+ (hydroxylated polystyrenedivinylbenzene copolymer) at pH 3.4 and 1.5, respectively [3]. Preliminary tests in these conditions for salt-rich reservoir waters produced very low (often 10-20%) and irreproducible recoveries. The preliminary tests using Oasis HLB phase were not encouraging, either. Although high, quasi-quantitative recoveries of the analytes were obtained, no conditions could be found for their quantitative desorption. The most promising results were obtained with a C18sorbentsimilar to that of the column which was investigated in detail.

187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202

ce pt

186

Ac

185

ed

183

The optimization procedure included: (i) choice of the solvent for the initial conditioning step (acetonitrile or tetrahydrofuran); (ii) pH of the final condition step and sample (acidic, neutral, or alkaline); (iii) choice of the elution solvent (acetonitrile and tetrahydrofuran) and its pH. The initial experiments with MeOH were unsuccessful. The conditions tested are summarized inTable 4. The results of the recoveries obtained during the optimization are summarized in Fig. 2. The first hypothesis tested involved the lowering pH to revert the dissociation of FBAs in order to increase their retention and then alkalize the solution for their 6 Page 43 of 49

[Tapez un texte]

210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229

ip t

209

In terms of elution conditions, the use of ammonia resulted in recovery ratios of FBAs higher than 90% for most of the analytes. Two polar organic eluting solvents (acetonitrile and THF) were testedtogether with ammonia. Recoveries from SPE procedures IX to XII were similar. Procedure X was chosen because theresulting solution (8:2 organic/aqueous) was easier to evaporate than 5:5 organic/aqueoussolution and because acetonitrile was easier to evaporate than THF.Also, the recoveries for 2,6-dFBA and 2,3,6-tFBA were significantly higher in comparison with other procedures.

cr

208

us

207

an

206

Fig. 2. indicates that quantitative (>90%)recoveries (retention/elution) of 16 out of 19 analyte compounds were achieved from a salt-rich water matrix. The simultaneous elimination of >99% of salt content and matrix simplification allowed a 4-fold preconcentration factor. For three compounds: 2,6-dFBA, 2,3,6-tFBA and 2,4,6-tFBA non-quantitative recoveries were observed.

M

205

ed

204

elution. The acidification was initially carried out only during the conditioning step (1% acetic acid) but the recoveries were lower than when the conditioning was carried out with water (cf. e.g. procedures VI-XII). The recoveriesdropped further when acetic acid was added to the sample during the loading step (procedure II). Hence, it was decided to add acid neither during conditioning nor to the sample. Note that the recoveries in alkaline conditions (conditioning step and sample) (procedure III) were dramatically low (possibly also due to the signal suppression because of the non-retained salt).

The values in Fig. 2 were completed by verifying the recoveries from the water by the method developed at three different concentration levels. The data are shown in Table 5. This systematic study showed that, in fine, only two compounds were problematic in terms of recoveries (2,6-dFBA, recovery ca. 50% and 2,4,6-tFBA, recovery ca. 80%). It could also be concluded that the matrix did not practically affect the recoveries.

ce pt

203

231 232 233 234 235 236 237 238 239

Ac

230

SPE - HPLC- MS/MS for the simultaneous multiple tracer analysis Fig. 3. shows a chromatogram obtained for a concentration of 50 pg/mlFBAs added to a sample matrix containing 200 g/l of salt by the SPE method developed and the corresponding blanks. The analytes’ concentration was chosen to correspond roughly to the detection limits of the procedure based on the direct injection HPLC. The figure clearly shows peaks for all the compounds well above the background; it demonstrates not only the absence of the need for sample dilution despite the high salt content but also an effective preconcentration factor of up to 4 times resulting from the SPE. The LODs are affectedby the ionization efficiency (the degree of matrix

7 Page 44 of 49

[Tapez un texte] 240 241 242 243 244 245

removal and the content of acetonitrile at a given point of the chromatographic gradient), the peak shape and the baseline noise (again depending on the matrix). The calibration curve data obtained for the procedureand the detection and quantification limits are summarized in Table 6. They confirm a 3-4-fold gain in detection limits resulting from the preconcentration factor in addition to the absence of the need of sample dilution prior to analysis.

ip t

246

Isotope dilution correction for the non-quantitatively eluted compounds:

248

The recoveries of themost polar compounds 2,6-dFBA, 2,3,6-tFBA and 2,4,6-tFBA were not only non-quantitative(theywere not sufficiently adsorbed on the C18 sorbent and partially found in the eluate of the spiked sample)but they were also observed to vary by up to 30 % depending on the day and sample matrix. Therefore they have to be corrected for.

252 253 254 255 256 257 258

A convenient method proposed by Müller et al.[4, 8]is the use of deuterated standards The chromatograms(Fig. 4) show the perfect co-elution of the doubly deuterated and non-deuterated standards which allows them to be measured in identical ionization conditions as the analyte.Table 7highlights the benefits from the isotopically-labelled internal standards showing an efficient correction of the nonquantitative recoveries. Note that a single internal standard wassufficient to correct both of 2,3,6-tFBa and 2,4,6-tFBA recoveries as these compounds elute closely and share the reaction used for their quantification.

262

Validation of the method developed

263

In order to validate the method, three synthetic samples containing all the tracers at the different concentration levels: 0.2, 1 and 10 ng/ml were prepared and analysed according to the developed procedure. The results shown in Table 8demonstrate consistent accuracies between 90-100% and precision between 2-5%.

264 265 266 267 268 269 270 271 272 273 274 275

Ac

260

ce pt

261

ed

259

us

251

an

250

M

249

cr

247

Analysis of real samples: comparison with the direct analysis The developed method was compared with the method based on the direct injection of diluted samples[1].The examples of chromatograms are shown in Fig. 5. The comparison shows an increase in sensitivity over at least an order of magnitude, allowing the detection of peaks in the background not seen with the direct injection method, stabilization of the baseline, and especially the elimination of the false positives commonly encountered when integrating the peaks close to baseline using the direct injection procedure. Note that the direct injection method developed 8 Page 45 of 49

[Tapez un texte] 276 277 278

elsewhere[1] was slightly improved by diverting the chromatographic eluate off the detector for the first30 sto reduce the load of the salt on the column, as recently suggested by Bayen [13].

279

Conclusions

281

288

The optimization of solid phase extraction allowed an efficient and straightforward simultaneous preconcentration of 19 fluorinated derivatives of benzoic acid commonly used as oil reservoir tracers from salt-rich waters.The simultaneous elimination of the salt eliminated the need for sample dilution allowing a gain of 1020 in terms of detection limits in comparison with the figures of merit reported elsewhere in the literature for the HPLC-MS/MS analysis of similar samples.The method requires a few ml of sample only, is relatively rapid and can be readily automated.

289

Acknowledgements

290 291 292 293 294 295

The authors thank Dr. O. Arwal, TOTAL (France) for supplying the samples used for the method development and Dr. K. Müller and Prof. Dr. A. Seubert (Faculty of Chemistry, Philipps-Universität, Marburg, Germany) for the gift of the deuterated 2,6-dFBA and 2,4,6tFBA. We also thank Applied Biosystems and Bruker (Paris) for providing indicative detection limits data for the FBA standardsfor the last generation Q-TOF systems. The financial support of the mass spectrometryplatform at the LCABIE-IPREM by Aquitaine Region is acknowledged.

cr

us

287

an

286

M

285

ed

284

ce pt

283

Ac

282

ip t

280

9 Page 46 of 49

[Tapez un texte] 296 297

Captions to Figures

298 299

Figure 1.HPLC-MS/MS chromatogramsobtained for 50 ng/mL standards. a) 139-->95: 1) 2-fluorobenzoic acid, 2) 3-fluorobenzoic acid, 3) 4fluorobenzoic acid;

302 303 304

b) 157-->113: 4) 2,6-difluorobenzoic acid, 5) 2,5-difluorobenzoic acid, 6) 2,3- difluorobenzoic acid, 7) 2,4-difluorobenzoic acid, 8) 3,5difluorobenzoic acid, 9) 3,4-difluorobenzoic acid;

305 306 307

c) 175-->113: 10) 2,3,6-trifluorobenzoic acid, 11) 2,4,6-trifluorobenzoic acid, 12) 2,4,5-trifluorobenzoic acid, 13) 2,3,4-trifluorobenzoic acid, 14) 3,4,5-trifluorobenzoic acid;

308 309

d) 189-->145: 15) 2-(trifluoromethyl)benzoic acid, 16) (trifluoromethyl)benzoic acid, 17)4-(trifluoromethyl)benzoic acid;

310

e) 193-->149: 18) 2,3,4,5-tetrafluorobenzoic acid;

311

f) 257-->213: 19) 3,5-bis(trifluoromethyl)benzoic acid;

312

g) 144->99: 20) 4-fluorobenzoic acid-α-13C-2,3,5,6-d4(internal standard);

cr

ip t

300 301

us

an

313

Figure 2. Analyte recoveries from a spiked reservoir water sampleobtained with the SPE procedures described in Table 4.

M

314 315

3-

316

Figure 3. HPLC-MS/MS chromatograms of a reservoir water spiked with 50 pg/mL of each FBA (top chromatogram in each subfigure) and the corresponding blank (unspiked reservoir water) analysed by the developed procedure.

320 321

a) 139-->95: 1) 2-fluorobenzoic acid, 2) 3-fluorobenzoic acid, 3) 4fluorobenzoic acid;

ce pt

ed

317 318 319

b) 157-->113: 4) 2,6-difluorobenzoic acid, 5) 2,5-difluorobenzoic acid, 6) 2,3- difluorobenzoic acid, 7) 2,4-difluorobenzoic acid, 8) 3,5difluorobenzoic acid, 9) 3,4-difluorobenzoic acid;

322 323 324

328 329

c) 175-->113: 10) 2,3,6-trifluorobenzoic acid, 11) 2,4,6-trifluorobenzoic acid, 12) 2,4,5trifluorobenzoic acid, 13) 2,3,4-trifluorobenzoic acid, 14) 3,4,5-trifluorobenzoic acid;

Ac

325 326 327

d) 189-->145: 15) 2-(trifluoromethyl)benzoic acid, 16) (trifluoromethyl)benzoic acid, 17) 4-(trifluoromethyl)benzoic acid;

330

e) 193-->149: 18) 2,3,4,5-tetrafluorobenzoic acid;

331

f) 257-->213: 19) 3,5-bis(trifluoromethyl)benzoic acid;

3-

332 333 334

Figure 4. HPLC-MS/MS chromatograms early eluting compounds with specific internal standards:

10 Page 47 of 49

[Tapez un texte]

a) 157 --> 113: 1) 2,6- difluorobenzoic acid; b) 159-->115: 2) 2,6difluorobenzoic acid -d2; c) 177 --> 131: 3) 2,3,6-tFBA, 4) 2,4,6-TFBA; d) 177-->133: 5) 2,4,6-tFBA-d2.

335 336 337 338 339 340 341

Figure 5.HPLC-MS/MS chromatograms of two (A and B) reservoir water samples. a,b Sample A. c,d - Sample B. a,c- direct injection upon dilution [9]b,d analysed by the SPE-HPLC-MS/MS procedure developed.

342

Ac

ce pt

ed

M

an

us

cr

ip t

343

11 Page 48 of 49

[Tapez un texte] 344

References

346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383

[1] C. Serres-Piole, A. Commarieu, H. Garraud, R. Lobinski, H. Preud'Homme, New passive water tracers for oil field applications, Energy and Fuels, 25 (2011) 4488-4496. [2] C. Serres-Piole, New water tracers for water reservoirs. A contribution to the fundamental understanding of tracer behaviour to enhance nanoscale monitoring in advanced reservoir exploitation by LC - tandem MS., PhD Thesis, University of Pau, France (2011). [3] K. Müller, A. Seubert, Ultra trace determination of fluorobenzoic acids in tap and reservoir water using solid-phase extraction and gas chromatography-mass spectrometry, Journal of Chromatography A, 1260 (2012) 9-15. [4] K. Müller, A. Seubert, Synthesis of deuterium-labelled fluorobenzoic acids to be used as internal standards in isotope dilution mass spectrometry, Isotopes in Environmental and Health Studies, 50 (2014) 88-93. [5] R.K. Juhler, A.P. Mortensen, Analysing fluorobenzoate tracers in groundwater samples using liquid chromatography-tandem mass spectrometry: A tool for leaching studies and hydrology, Journal of Chromatography A, 957 (2002) 11-16. [6] T. Isemura, F. Kitagawa, K. Otsuka, Separation of complex mixtures of fluorobenzoic acids by capillary electrophoresis, Journal of Separation Science, 32 (2009) 381-387. [7] K. Müller, A. Seubert, Separation and determination of fluorobenzoic acids using ion chromatography-electrospray mass spectrometry, Journal of Chromatography A, 1270 (2012) 96-103. [8] K. Müller, A. Seubert, Ultra trace determination of fluorobenzoic acids in reservoir and ground water using isotope dilution gas chromatography mass spectrometry, Isotopes in Environmental and Health Studies, 50 (2014) 277-284. [9] C. Serres-Piole, N. Moradi-Tehrani, R. Lobinski, H. Preud'homme, Direct sensitive simultaneous determination of fluorinated benzoic acids in oil reservoir waters by ultra highperformance liquid chromatography-tandem mass spectrometry, Journal of Chromatography A, 1218 (2011) 5872-5877. [10] M. Concheiro, S. Anizan, K. Ellefsen, M.A. Huestis, Simultaneous quantification of 28 synthetic cathinones and metabolites in urine by liquid chromatography-high resolution mass spectrometry, Analytical and Bioanalytical Chemistry, 405 (2013) 9437-9448. [11] V. Gabet-Giraud, C. Miege, B. Herbreteau, G. Hernandez-Raquet, M. Coquery, Development and validation of an analytical method by LC-MS/MS for the quantification of estrogens in sewage sludge, Analytical and Bioanalytical Chemistry, 396 (2010) 1841-1851. [12] M.J. Whiting, Simultaneous measurement of urinary metanephrines and catecholamines by liquid chromatography with tandem mass spectrometric detection, Annals of Clinical Biochemistry, 46 (2009) 129-136. [13] S. Bayen, X. Yi, E. Segovia, Z. Zhou, B.C. Kelly,Analysis of selected antibiotics in surface freshwater and seawater using direct injection in liquid chromatography electrospray ionization tandem mass spectrometry, Journal of Chromatography A, 1338 (2014) 38-43.

cr

us

an

M

ed

ce pt

Ac

384

ip t

345

12 Page 49 of 49