Automated SPE for NDMA and Metaldehyde in water

Automated SPE for NDMA and Metaldehyde in water using GCQq. Q Dan Carrier, Applications Chemist www. anatune. co. uk

Purpose of this presentation “show methods NDMA & Metaldehyde fully automated using Gerstel and Agilent instrumentation”

Summary of talk • Introduction – Background NDMA and Metaldehyde – Why trace level method for both analytes • Method – Automated Sample Preparation – Chromatographic method using GC/MS/MS • Results – ~1 ng/l NDMA in extracted water – ~2 ng/l Metaldehyde in extracted water • Summary – Good overview of system

Anatune • Girton, Cambridge (March 2012) • VAR for Agilent – GC and LC products – MSD, Qq. Q, QTOF • Gerstel – MPS – DHS, Twister, ITSP • Focus - Sell and Support Solutions – Wide number of industries - Environmental, Food and Flavours, Petrochem, Pharmaceutical, Forensic, and Clinical

• Dual Head MPS Solution with LC/UV – Formaldehyde and acetaldehyde in air

Twister SBSE • Enrich trace level analytes in water – Twisters made PDMS (ethylene glycol) – Adsorption based on lipophibicity – Log K o/w

Twister SBSE – PAH solution • 100 ml water samples (2 hours) – Dried and placed in TDU tubes – SIM 16 PAH (0. 02 ug/l to 1 ug/l) – Acenaphthene 0. 999 (1 -2%)

ITSP & LC/MS/MS • Diurons in water – Enrich 10 ml – Elute in 0. 8 ml Chromatogram Monuron (quantifier transition) in standard 0. 40 μg/L after extraction Correlation coefficient Calibration after extraction Monuron 0. 9989 Isoproturon 0. 9995 Diuron 0. 9997 Linuron 0. 9984

Introduction • NDMA and Metaldehyde very polar molecules – Enrichment methods from water problematic – Developed two separate automated methods Metaldehyde NDMA

Harm - NDMA • Industrial by-product many processes • IARC (International Agency for Research on Cancer) NDMA is probably carcinogenic to humans • WHO Guideline limit NDMA of 100 ng/l in drinking water.

Metaldehyde • Widespread Pesticide • Regulatory limit 100 ng/L in drinking water

NDMA • Initial method on MSD 5975 C (before Qq. Q) – Method Detection limit 25 ng/L – AS 119 – Needed to get better sensitivity but also selectivity!

Summary of talk • Introduction – NDMA and Metaldehyde • Method – Automated Sample Preparation (focus on Gerstel instrumentation) – GC/MS/MS method • Results • Summary

Automated Sample Prep 2. 5 ml HS Syringe 10 ul Syringe • Brand this set up - Multiflex – Consists of Dual Head MPS – Thermal Desorption unit – Cold Inlet System PTV

ITSP (Instrument Top Sample Preparation) • Typically 15 -25 mg sorbent • Metaldehyde – Biotage ENV (as in literature) – NDMA – Coconut charcoal Name Metaldehyde_DCM_Standard 4 Metaldehyde. STD+IS_test_Biotage. ENV Metaldehyde. STD+IS_test_DVBENV Metaldehyde. STD+IS_test_C 18 Metaldehyde. STD+IS_test_Coconut. Charcoal Metaldehyde Results Metaldehyde_ d 16 (ISTD) Results Resp. Area 5202 4756 6678 5510 959 926 880 354 478 6392 6650 5910 6189 6115 8082 6795 857 919 779 284 266 392. 7722 7737 7172

Coconut Charcoal ITSP cartridges (NDMA) ENV (Metaldehyde) Right MPS (2. 5 ml Headspace syringe) Conditioned 750 µl dichloromethane 1000 µl of methanol Equilibrated 2000 µl of HPLC grade water Load 10 ml of sample (in water) Dried 15 minutes Eluted 400 ul dichloromethane Left MPS (10 ul) Large Volume injection X 25 concentration

Large Volume Injection– removing DCM boiling point 40 °C, Metaldehyde and NDMA both exceed 100 °C Inlet kept at 10 °C (peltier cooled) Slow injection speed at 0. 5 ul/s (to remove DCM) ramped to 250 °C (12 °C /s) NDMA (similar for Metaldehyde) - Glass beads liner gave best results

Summary of talk • Introduction – NDMA and Metaldehyde • Method – Automated Sample Preparation – GC/MS/MS method • Results • Summary

Agilent GC/Qq. Q • Agilent 7890 A GC • Agilent 7000 GC/MS triple quad – Increased Sensitivity and Selectivity Direct comparison at NDMA at 0. 125 ng/ml (without extraction) Single Ion Monitoring Multiple Reaction monitoring

GC Methods • NDMA – DB-WAX 30 m x 250 μm x 0. 5 μm – Thermal gradient from 35 deg C to 240 deg C over 10 minutes • Metaldehyde – DB 5 30 m x 250 μm x 0. 25 μm – Thermal gradient from 35 deg C to 250 deg C over 10 minutes • Different method – No reason in future to create one method – WAX

Multiple Reaction Monitoring 74. 1 44. 1 Direct comparison at NDMA at 0. 125 ng/ml

Optimised MRM Transitions for NDMA Compound NDMA-d 6 (Quant) NDMA-d 6 (Qual) NDMA (Quant) NDMA (Qual) Precursor Ion 80. 1 74. 1 Product Ion CE (v) 50. 1 46. 1 44. 1 42. 1 5 15 5 20

EI Ionisation Experiments 70 e. V Varying from 10 to 80 e. V

Varying Ion Source Temperature Varying between 150 & 350 °C

Optimised MRM Transitions for Metaldehyde Compound Precursor Ion Product Ion CE (v) Metaldehyde (Quant) 89. 0 45. 1 9 Metaldehyde (Qual) 89. 0 43. 1 42 98. 1 50. 1 12 98. 1 46. 1 39 Metaldehyde d 16 (Quant) Metaldehyde d 16 (Qual)

Summary of Method development • Metaldehyde method developed within 2 weeks • NDMA over 3 months – Contamination issues – Exposing LC/MS grade water to light!

Summary of talk • Introduction – NDMA and Metaldehyde • Method – Automated Sample Preparation – GC/MS/MS method • Results - NDMA • Summary

Extracted Water - NDMA Water spiked to build seven point calibration from 0. 25 to 15 ng/l. Correlation co-efficient of 0. 9995. X 25 concentration

NDMA comparison with blank 1 ng/l NDMA in LC/MS grade water with blank (extracted LC/MS grade water) - around detection limit

Standard Extracted Water 10 ng/l NDMA-d 6

Absolute Recovery Absolute recovery of NDMA established at 1 ng/l & 10 ng/l spikes 1 ng/L NDMA – Absolute recovery 55 % (% RSD 9. 2) 10 ng/l NDMA – Absolute recovery 61 % (% RSD 2. 6) 40 ng/l NDMA-d 6 – Absolute recovery 56 % (% RSD 7. 9)

Recovery and precision of water extractions Calculated accuracy of all calibration points 86. 5 – 107. 1 % Accuracy and precision of 1 ng/l & 10 ng/l standards (n=3) NDMA 1 ng/l 10 ng/l Mean ng/l result 0. 997 10. 329 SD 0. 005 0. 266 0. 5 2. 6 100. 2 102. 7 % RSD Ave. % Corrected Recovery

40 ng/l Metaldehyde extracted standard comparison with blank Detection limit approx 2 ng/l (based on signal to noise from this standard)

Extracted Water - Metaldehyde Water spiked to build seven point calibration from 40 to 800 ng/l. Correlation co-efficient of 0. 9993. X 25 concentration

Recovery and precision of water extractions (Metaldehyde) Amount spiked (μg/L) 0. 06080 0. 70400 Amount detected (μg/L) 0. 05734 0. 63858 Mean SD % Recovery 0. 05721 0. 71908 0. 06000 0. 70449 0. 05628 0. 72256 0. 05641 0. 72204 0. 05745 0. 70135 0. 0015 0. 035856 2. 61 5. 11 94. 49 99. 62

Summary • Developed two fully automated methods for NDMA and Metaldehyde • Good linearity and recovery for NDMA and Metaldehyde • Detection limit (after extraction) – NDMA between 0. 5 to 1 ng/l – Metaldehyde approx 2 ng/l

Acknowledgements • Anatune – Paul Roberts – Anais Maury – Matthew Carson • Rick Youngblood • Ken Brady

Any Questions?

Additional Work on MRM method • Work was completed to assess the impact of running MS 1 & MS 2 in unit, wide and widest resolution modes. • Increased sensitivity as resolution decreased, but possibility of interferences increases • Unit 0. 7 amu at half height • Wide 1. 2 amu at half height • Widest 2. 5 amu at half height • Default QQQ Collision Cell conditions for Collision Cell and Quench Gas