Accurate Mass Spectrometer to Expand the Scope of EPA ......Octadecyl (C-18) functional group after adding Surrogates. The eluent from SPE is concentrated to dryness, and then diluted

www.eatonanalytical.com

Use of a High Resolution Accurate Mass Spectrometer to Expand the Scope of EPA

Methods for Emerging Contaminants (EPA 539)

October 2015

Andy Eaton, PhD, BCESTechnical Director

Ali Wolfgang HaghaniTechnical R&D Manager

Eurofins Eaton Analytical Monrovia, CA 91016

www.eatonanalytical.com

Hormones as Emerging Contaminants (EC) is NOT a New Concept

•~1950 as the TOF-MS and GC-MS concepts were introduced at Harvard, scientists were already speculating on fate and transport of hormones in water streams.

• ~ 1970-80’s New analytical technologies made studies possible using GC-MS or TOF-MS but still detection limits were inadequate for needed levels.

• 2001 USGS; Koplin et al. published their Water-Quality Data for Pharmaceuticals, Hormones, and Other Organic Wastewater Contaminants in U.S. Streams, 1999-2000 - some at low ppt levels, using LCMS as SIM for two of their 5 methods.

• 2010 EPA 539 by Glynda Smith et al. was published for UCMR 3 monitoring (2013-2015) using Tandem MS for selected hormones at low to sub ppt levels. The lowest reporting limit was set for Testosterone at 0.1ppt.

•Directive 2013/39/EU (Europa made of 28 nations) – Estrone, Estradiol and Ethynylestradiol at 1.0, 0.3 and 0.03ppt in water streams such as effluents.

2

2011 Q Exactive: a Hybrid HRAM is Added to the Orbi Exactive Family

3From Trussell, Clarke Prize Lecture 2013

2005 LTQ2007 Exactive2011 Q Exactive…2015 PLUS & HF

EPA 539 summary

MS requirements of the method:Summary : A 1 Liter of dechlorinated sample with Omadine as a biocide is extracted through SPE Octadecyl (C-18) functional group after adding Surrogates. The eluent from SPE is concentrated to dryness, and then diluted to 1 ml with 50/50% methanol/water. An aliquot is injected to the LC-MS/MS after adding internal standards and quantified against the IS.

3.16 MULTIPLE REACTION MONITORING (MRM) – A mass spectrometric technique in which a precursor ion is first isolated, then subsequently fragmented into a product ion(s). Quantitation is accomplished by monitoring a specific product ion. As described in Section 10.1.2, MS parameters must be optimized for each precursor ion (Sect. 3.17) and product ion (Sect. 3.19).

6.12.3 ELECTROSPRAY IONIZATION-TANDEM MASS SPECTROMETER (ESI-MS/MS) – The mass spectrometer must be capable of rapid switching between negative ion and positive ion electrospray ionization modes. The system must be capable of performing MS/MS to produce unique product ions (Sect. 3.19) for the method analytes within specified retention time segments. At least 10 – 15 scans across the chromatographic peak are needed to ensure adequate precision.

Orbitrap™Technology meets these requirements

4

Why Consider Orbitrap Technology for Method 539? We are becoming more and more concerned

about “non-target” analytes. Traditional LC-MS/MS systems are focused

entirely on target analytes. Orbitrap and Q-TOF (time of flight) sensitivity has

improved over the last decade Orbitrap has resolution of >70,000, allowing one to

also watch for false positives and false negatives. The UCMR 3 database shows a lot (>5% of PWS) of

“sporadic” results for hormones. It is important to ensure they are real.

5

Evaluating the QE in Parallel Reaction Mode (PRM) for EPA 539 Initial Demonstration of capability (IDC) from the method: 1. DEMONSTRATION OF LOW SYSTEM BACKGROUND <1/3 MRL 2. DEMONSTRATION OF PRECISION – 4-7 LFB’s RSD </= 20% 3. DEMONSTRATION OF ACCURACY – LFB’s recovery diff. </= 30% of T.V. 4. MINIMUM REPORTING LEVEL (MRL) CONFIRMATION or doing LCMRL

Targeted compounds and their MRLs:Analyte DL Fortified Concentration (ng/L) DL (ng/L) LCMRL (ng/L)Estriol 1.10 0.24 0.28Estrone 1.05 0.19 4.017β-Estradiol 1.30 0.39 0.3217α-Ethynylestradiol 1.75 0.33 1.3Androstenedione 0.50 0.20 0.37Testosterone 0.50 0.04 0.062Equilin 1.25 2.94 3.0

6

EPA 539 Is Written For LC-MS/MS So The QE PRM Mode Is Truest To The Method

7

Quadrupole-equipped HR/AM instrumentsHR/AM analyzer permits parallel detection of all target product ions in one concerted high resolution mass analysis

Serial monitoring

Parallel monitoring

Drag And Drop Method Development For NEG/POS/NEG On The Same Run for EPA 539

8

All Calibration Coefficients > 0.99 and Good Dynamic Range and Linearity

9

EE2 range 0.225 – 7.2ppt

An Example of a Calibration for EE2 Using TraceFinder 3.2

Compound Calibration Report

Lab Name: EEA Method:021215-539-PRM-IDOC-LINEAR_EPA-539-PRMInstrument: Q EXACTIVE EPA-539-PRMUser: Thermo Cali File:021215-539-PRM-IDOC-LINEAR.calx

Batch:021215-539-PRM-IDOC-LINEAR

Compound Name: 17 alpha-Ethynylestradiol

LinearPass

LevelStd

Amount Std Area IS Amount IS AreaResp factor/

ratio

CalcAmount Units % CV % RSD

0 0 419 100 9244925 0 0.002 ng/L N/A N/A

ICAL-1 0.225 63224 100 10063239 0.006 0.231 ng/L N/A N/A

ICAL-2 0.45 110807 100 9324158 0.012 0.436 ng/L N/A N/A

ICAL-3 0.9 243598 100 9755408 0.025 0.915 ng/L N/A N/A

ICAL-4 1.8 491433 100 9920757 0.05 1.815 ng/L N/A N/A

ICAL-5 3.6 942018 100 9557799 0.099 3.612 ng/L N/A N/A

ICAL-6 5.4 1415265 100 9788761 0.145 5.298 ng/L N/A N/A

ICAL-7 7.2 1850842 100 9334500 0.198 7.266 ng/L N/A N/A

10

Determination of Minimum Reporting Limit (MRL) using LCMRL calculation

11

The LCMRL is defined as the lowest spiking concentration at which recovery of between 50 and 150 percent is expected 99 percent of the time by a single analyst. The procedure requires, at a minimum, four replicates at each of seven fortification levels. Four laboratory reagent blanks should also be included. All must be processed through the entire method procedure

http://water.epa.gov/scitech/drinkingwater/labcert/analyticalmethods_ogwdw.cfm

NB – Calculator only works with 32 bit computers.

LCMRL<0.19 ppt For EE2. It Could Not Be Calculated Because It Needed A Lower Point

12

0.05ppt as critical level

Area and Height of the Lowest CAL Stdfor EE2 – Using Trace Finder 3.2

Compound Name: 17 alpha-EthynylestradiolSample Conc: 0.225 ng/LRetention time: 14.96Area (Quan): 61535Height (Quan): 5855

13

Can Directive 2013/39/EU be achieved at 0.03ppt based on Area & Height ?

HRAM is MUCH More Sensitive than the Published LCMRL for EPA539Laboratory: Eurofins Eaton Analytical - EEA

Extraction: Concentrated 500 ml of solutions to final 1 ml - per EPA 539 preservatives/extraction procedure

Instrument: Q- Exactive, hybrid quadrupole Mass Spectrometry HRAM,Resolution for collected data : 70000 (mz 200 FWHM)AGC 2e5Max IT 200 msIsolation window 1.0 m/z

Scans/peak collected >10-15 scans per peaks

HPLC: Thermo RS Ultimate UHPLC 3000, binary pump, autosampler and column heater with 100 ul sample loopColumn: Acclaim Polar Advantage II, 2.1x150 mm, 3 um, 120AEluents: A) 1 mM Ammonium Fluoride B) 50%/50% ACN/MeoH

Gradient flow at 0.3ml/min with a 21.4 minutes run.Injection volume: 50 ul

Method 539 UCMR3 Proposed Analyte UCMR MRL (ng/L) EPA 539 published LCMRL

(ng/L)

Q Exactive HRAM 70000 Res.

LCMRL (ng/L)

QE-LCMRL Calc -DL (ng/L)

17α-ethynylestradiol 0.9 1.3 Critical level 0.05 0.117β-estradiol 0.4 0.32 0.17 0.047equilin 4 0.28 Critical level 0.23 0.48estriol 0.8 3 0.27 0.2estrone 2 4 0.84 0.48testosterone 0.1 0.062 0.033 0.0274-androstene-3,17-dione 0.3 0.37 0.19 0.08

14

This got EPA (Glynda Smith’s) one thumb up of approval and, as she predicted, it was much more sensitive.

Precision and Accuracy for LFB (10xMRL) and a Spiked Sample at MRL

QE-HRAM-MS2 70000 resolution PRECISION AND ACCURACY IN FORTIFIED REAGENT WATER (n=4)Analyte Fortified Concentration (ng/L) Avg. %Recovery %RSD17α-ethynylestradiol 7.2 82% 417β-estradiol 3.2 84% 3equilin 32 81% 3estriol 6.4 100% 4estrone 16 83% 4testosterone 0.8 87% 5

4-androstene-3,17-dione 2.4 85% 8

QE-HRAM-MS2 70000 resolution PRECISION AND ACCURACY IN Sample 20150108002 (2) Low Level spikeAnalyte Fortified Concentration (ng/L) Avg. %Recovery %RSD17α-ethynylestradiol 0.72 95% 2.317β-estradiol 0.32 87% 0.6equilin 3.2 92% 8.2estriol 0.64 101% 4.0estrone 1.6 95% 3.4testosterone 0.08 99% 0.1

4-androstene-3,17-dione 0.24 118% 0.1

15

Continued: Spiked Field Samples at Reporting Levels. Samples Had No Hits.

QE-HRAM-MS2 70000 resolution PRECISION AND ACCURACY IN Sample 20150108002 (2) Low Level spikeAnalyte Fortified Concentration (ng/L) Avg. %Recovery %RSD17α-ethynylestradiol 0.72 95% 2.317β-estradiol 0.32 87% 0.6equilin 3.2 92% 8.2estriol 0.64 101% 4.0estrone 1.6 95% 3.4testosterone 0.08 99% 0.14-androstene-3,17-dione 0.24 118% 0.1

QE-HRAM-MS2 70000 resolution PRECISION AND ACCURACY IN Sample 201501090001 (2) Low Level spikeAnalyte Fortified Concentration (ng/L) Avg. %Recovery %RSD17α-ethynylestradiol 0.72 97% 1617β-estradiol 0.32 95% 13equilin 3.2 92% 11estriol 0.64 107% 13estrone 1.6 95% 14testosterone 0.08 97% 194-androstene-3,17-dione 0.24 104% 20

16

This got Glynda Smith’s second thumbs up. Matrix spike recovery at the MRL is a truer measure of performance than the LCMRL.

Second Approach - How About Looking For Non Targeted And Unknowns While Quantifying For EPA 539 Compounds Simultaneously ?

17

DIA # 1: 50 amuisolation window, loop 6DIA # 2: Estriolwith isolation window 1 amu, NCE 75Loop 1DIA # 3: 50 amuisolation window loop 1DIA # 4 : 200 amuisolation window loop 2

Why adding a PRM for estriol?

Estriol Problematic For the Full-MS (70K) Method.No Signal Collected for Cal1&2 (0.2 & 0.4ppt) Stds

CAL 1 & 2ICAL-3 = MRL

Estriol At Ultra Low Concentrations Shows Selectivity Issues @ 70000 res(TSIM/ddMS2)

19

5 ppm accuracy filter

Internal Std Estriol-d2 Spiked At 10x In DI 289.7773 Amu And Contaminated Interference 287.16171 As A Suspect Causing Non Selectivity >10ppm For Estriol 287.16528

C:\TraceFinderData\...\Data\030515-071 03/25/15 20:59:58 IS

RT: 5.78 - 8.09 SM: 5B

6.0 6.5 7.0 7.5 8.0Time (min)

10

20

30

40

50

60

70

80

90

100

Rel

ativ

e A

bund

ance

6.61

6.197.657.22 7.30

NL: 1.73E7TIC F: FTMS - p ESI Full ms2 [email protected] [50.00-325.00] MS 030515-071

RT: 5.78 - 8.09 SM: 5B

6.0 6.5 7.0 7.5 8.0Time (min)

10

20

30

40

50

60

70

80

90

100

Rel

ativ

e A

bund

ance

6.61

6.197.657.22 7.30

NL: 1.73E7TIC F: FTMS - p ESI Full ms2 [email protected] [50.00-325.00] MS 030515-071

030515-071 #3243 RT: 6.61 AV: 1 NL: 1.82E7F: FTMS - p ESI Full ms2 [email protected] [50.00-325.00]

285 286 287 288 289 290 291 292 293 294m/z

0

10

20

30

40

50

60

70

80

90

100

Rel

ativ

e A

bund

ance

289.17773R=16706

290.18115R=16702

288.17154R=15502

291.18369R=15002

287.16171R=15102

030515-071 #3243 RT: 6.61 AV: 1 NL: 2.84E5F: FTMS - p ESI Full ms2 [email protected] [50.00-325.00]

286.6 286.8 287.0 287.2 287.4 287.6 287.8 288.0m/z

0

10

20

30

40

50

60

70

80

90

100

Rel

ativ

e A

bund

ance

287.16171R=15102

Possible interfering unknown

Full MS Does Not See 287.16528. DIA/PRM Sees Precursors In MS2 Plus The Product Ion – Spiked Sample At MRL

C:\TraceFinderData\32\DIA-4 05/16/15 18:16:48

RT: 4.90 - 7.67 SM: 5B

5.0 5.2 5.4 5.6 5.8 6.0 6.2 6.4 6.6 6.8 7.0 7.2 7.4 7.6Time (min)

50

1000

50

1000

50

100

Re

lative

Ab

un

da

nce

50

1005.935.91 5.965.79 6.05 6.13 6.386.33 6.445.77 6.475.575.40 6.555.465.325.26 6.876.645.74 7.045.23 6.93 7.10 7.13 7.21 7.335.62 7.415.15 7.565.125.03

7.02 7.086.99 7.116.946.596.56 6.62 6.766.37 7.166.34 7.28 7.34 7.42 7.456.25 7.576.175.975.945.775.755.69

7.067.03 7.09

7.116.977.146.94 7.176.91 7.266.636.57 6.80 7.607.557.406.406.345.225.16

5.50 5.945.33 6.005.895.445.27 5.80 6.035.24 5.55 5.58 6.14 6.205.16 6.31 6.42 6.484.99 6.565.63 6.65 7.056.976.85 7.11 7.14 7.22 7.31 7.54 7.57

NL: 1.21E6Base Peak m/z= 287.16383-287.16671 F: FTMS - p ESI Full ms [100.00-1000.00] MS DIA-4

NL: 9.44E4Base Peak m/z= 287.16383-287.16671 F: FTMS - p ESI Full ms2 [email protected] [50.00-315.00] MS DIA-4NL: 7.70E3Base Peak m/z= 287.16383-287.16671 F: FTMS - p ESI Full ms2 [email protected] [50.00-310.00] MS DIA-4NL: 1.23E5Base Peak m/z= 287.16383-287.16671 F: FTMS - p ESI Full ms2 [email protected] [50.00-365.00] MS DIA-4

DIA-4 #2881 RT: 7.01 AV: 1 NL: 6.81E5T: FTMS - p ESI Full ms [100.00-1000.00]

286.98 287.00 287.02 287.04 287.06 287.08 287.10 287.12 287.14 287.16 287.18 287.20 287.22 287.24 287.26 287.28m/z

0

10

20

30

40

50

60

70

80

90

100R

ela

tive

Ab

un

da

nce

287.16611

287.12924

287.22327

287.14124 287.18661 287.20203287.09473

C:\TraceFinderData\32\DIA-4 05/16/15 18:16:48

RT: 4.90 - 7.67 SM: 5B

5.0 5.2 5.4 5.6 5.8 6.0 6.2 6.4 6.6 6.8 7.0 7.2 7.4 7.6Time (min)

50

1000

50

1000

50

100

Re

lativ

e A

bu

nd

an

ce

50

1005.935.91 5.965.79 6.05 6.13 6.386.33 6.445.77 6.475.575.40 6.555.465.325.26 6.876.645.74 7.045.23 6.93 7.10 7.13 7.21 7.335.62 7.415.15 7.565.125.03

7.02 7.086.99 7.116.946.596.56 6.62 6.766.37 7.166.34 7.28 7.34 7.42 7.456.25 7.576.175.975.945.775.755.69

7.067.03 7.09

7.116.977.146.94 7.176.91 7.266.636.57 6.80 7.607.557.406.406.345.225.16

5.50 5.945.33 6.005.895.445.27 5.80 6.035.24 5.55 5.58 6.14 6.205.16 6.31 6.42 6.484.99 6.565.63 6.65 7.056.976.85 7.11 7.14 7.22 7.31 7.54 7.57

NL: 1.21E6Base Peak m/z= 287.16383-287.16671 F: FTMS - p ESI Full ms [100.00-1000.00] MS DIA-4

NL: 9.44E4Base Peak m/z= 287.16383-287.16671 F: FTMS - p ESI Full ms2 [email protected] [50.00-315.00] MS DIA-4NL: 7.70E3Base Peak m/z= 287.16383-287.16671 F: FTMS - p ESI Full ms2 [email protected] [50.00-310.00] MS DIA-4NL: 1.23E5Base Peak m/z= 287.16383-287.16671 F: FTMS - p ESI Full ms2 [email protected] [50.00-365.00] MS DIA-4

DIA-4 #2888 RT: 7.03 AV: 1 NL: 1.50E4F: FTMS - p ESI Full ms2 [email protected] [50.00-310.00]

170.60 170.65 170.70 170.75 170.80 170.85 170.90 170.95 171.00 171.05 171.10 171.15 171.20 171.25 171.30 171.35m/z

0

10

20

30

40

50

60

70

80

90

100

Re

lativ

e A

bu

nd

an

ce

171.08087

171.04462171.11725

Product 171.0808

Precursor-287.16528

2.8ppm

PRM PRM

DIA

DIA DIA

DIA

Full Full

Estriol Lowest Standard Using Trace Finder 3.2 Spectral Library Match / Identified

22

Full-MS/DIA gives the same result as PRM experiment. No selectivity issues in our experimental set up or we have not gone low enough to test it. Precision/Accuracy of LFB and Matrix spikes matched very closely to PRM experiment.

Extraction: Concentrated 500 ml of solutions to final 1 ml - per EPA 539 preservatives/extraction procedure

Instrument: Q- Exactive, hybrid quadrupole Mass Spectrometry HRAM, Q- Exactive, hybrid quadrupole Mass Spectrometry HRAM,

Resolution for collected data : 70000 (mz 200 FWHM)Full-MS - Resolution for collected data : 70000 (mz@200 FWHM)

AGC 2e5DIA 17500 (mz@200FWHM)

Max IT 200 ms Estriol PRM as a line with isolation window 1 amu NCE=75

Isolation window 1.0 m/zDIA experiment one 100-520 60 amu isolation window, loop 7DIA experiment #2 510-1100 200 amu isolation window, loop 2

HPLC: Thermo RS Ultimate UHPLC 3000,, binary pump, autosampler and column heater with 100 ul sample loop Step NCE 35, 80

Column: Acclaim Polar Advantage II, 2.1x150 mm, 3 um, 120A

Eluents: A) 1 mM Ammonium Fluoride B) 50%/50% ACN/MeoH

Gradient flow at 0.3ml/min with a 21.4 minutes run.

Injection volume: 50 ul

PRM PRMFULL-

MS/DIA(EstriolPRM)

FULL-MS/DIA(Estriol

PRM)

Method 539 UCMR3 Proposed Analyte UCMR MRL (ng/L) EPA 539 published LCMRL

(ng/L)

Q ExactiveHRAM 70000 Res. LCMRL

(ng/L)

QE-LCMRL Calc -DL (ng/L)

Q Exactive HRAM 70000 Res. LCMRL

(ng/L)

QE-LCMRL Calc -DL (ng/L)

17α-ethynylestradiol 0.9 1.3Critical level

0.05 0.1 Critical level 0.05 0.1

17β-estradiol 0.4 0.32 0.17 0.047 0.16 0.12

equilin 4 0.28Critical level

0.23 0.48 Critical level 0.23 0.44

estriol 0.8 3 0.27 0.2 Critical level 0.046 0.068estrone 2 4 0.84 0.48 0.7 0.45testosterone 0.1 0.062 0.033 0.027 NA NA

4-androstene-3,17-dione 0.3 0.37 0.19 0.08 NA NA23

Using Full-MS data with Sieve Software for Site Specific Emerging Contaminants

FrameTable: 209 Items CompID ID CompMW Formula MZ Time HitCount

10 10 211.1428815 C9H24N1[32]S2 210.1356049 5.512476921 1

DetailTable: 1 Items

CSID Name Formula SMILES MW AvgMass MonoisotopicMass SearchMass DeltaPPM Adduct

14620 Atraton C_{9}H_{17}N_{5}O CC/N=c/1\[nH]/c(=N/C(C)C)/[nH]c(n1)OC

211.2642 211.2642 211.143311 211.1428815 2.03436459 comp

11 11 232.016676 C1H5O9N5 231.0093994 7.780558586 1

DetailTable: 1 Items

CSID Name Formula SMILES MW AvgMass MonoisotopicMass SearchMass DeltaPPM Adduct

3008 Diuron C_{9}H_{10}Cl_{2}N_{2}O CN(C)/C(=N\c1ccc(c(c1)Cl)Cl)/O

233.0945 233.0945 232.017014 232.016676 1.45673122 comp

12 12 234.1254712 233.1181946 5.222514153 2

DetailTable: 2 Items

CSID Name Formula SMILES MW AvgMass MonoisotopicMass SearchMass DeltaPPM Adduct

4E+06 Stiripentol C_{14}H_{18}O_{3} CC(C)(C)C(/C=C/c1ccc2c(c1)OCO2)O

234.29092 234.2909 234.125595 234.1254712 0.528861389 comp

24

Summary and Questions

The HRAM performance has better sensitivity than LC-MS/MS for EPA 539 with additional confirmation such as product ions and a spectral library match.

Further the same injection can be used for identification of non-target and unknowns simultaneously, using Sieve software and/or looking for formation and degradation products using Compound discovery software.

Dr. Andrew Eaton ([email protected])Ali W. Haghani ([email protected])