1. Douglas Stevens Kari Organtini Waters Corporation
Tandem Quadrupole MS/MS v. Magnetic Sector HRMS in the Analysis of Mixed-Halogenated Dioxins and Furans: Firefighter Occupational Exposure
Douglas Stevens
Kari Organtini
Waters Corporation

2. Overview
Atmospheric Pressure Chemical Ionisation Gas Chromatography (APGC aka GC-APCI) source on HP TQ can achieve extremely low limits of detection, e.g. <100ag for 2,3,7,8 TCDD
Comparisons with magnetic sector and 2D GC EI TOF was performed and will be discussed

3. 100fg 2,3,7,8 TCDD on Magnetic Sector
Signal to Noise = 125:1
[PtP using 10 peak widths of noise]

4. APGC on Xevo TQ-XS

5. 100fg 2,3,7,8 TCDD on TQ-XS Signal to Noise = 5888:1
[PtP using 10 peak widths of noise]

6. Wellington Labs TCDD-MXB Standard
100fg 2,3,7,8-TCDD
Rxi-5Sil MS 60m x 0.25mm x 0.25µm Column
50fg 1,2,3,4-TCDD
25fg 1,4,7,8-TCDD
10fg 1,3,7,8-TCDD
5fg 1,3,7,9-TCDD
2fg 1,3,6,8-TCDD

7. TCDD-MXB Standard diluted 10:1
500ag
200ag

8. TCDD sensitivity GC APCI MS/MS v EI HRMS
Quantification of peak areas in terms of number of ions detected
APGC MS/MS
100fg gives 63,500 ions
[This equates to 7060 ions when monitoring 9 MRM transitions]
EI HRMS
100fg gives 9900 ions
[1100 ions when monitoring 9 SIR masses]

9. Ion losses in HRMS Total Losses = 94.0% “Beta” slit typical losses ~2%
“Z restrictor” typical losses ~5%
Majority of losses are at source slit to resolution
Final shaping at collector, ~10% loss
Total ion loss at 10,000 RP = 92.8%
Total Losses = 94.0%

10. Ion losses in TQ MS/MS Total Losses = 86.8%
Losses in quadrupoles typically 15% each
Estimate of losses 60%
Quadrupole 1
Collision
Cell
Quadrupole 2
Detector
Source
StepWave
Losses in collision cell are ~5%
MRM fragmentation loss = 52% [for TCDD]
Total Losses = 86.8%

11. TCDD sensitivity APGC MS/MS v EI HRMS
HRMS APGC MS/MS Analyser Ion Loss 94.0% 86.8% System Efficiency 0.018% 0.117% Ionisation Efficiency 0.30% 0.89% APGC source producing ~ +3X more ions for TCDD Further ~ +2X sensitivity from MS/MS v HRMS

12. Atmospheric Pressure Chemical Ionisation
Chemical ionisation region
Ionising corona

13. Atmospheric Pressure Chemical Ionisation
Corona discharge
needle
N2+● N2 e-
2e-
2N2
M+● M
Nn+●
M+● M
Charge Exchange is driven by Ionisation Energy
A+· + B A + B +·
ONLY IF IEA > IEB

14. Dioxins and Furans
Spectral Comparisons

15. PeCDD Spectra
GC APCI EI

16. OCDD Spectra
GC APCI EI

17. Detection Examples – 1,2,3,7,8 PeCDD
EPA1613 CSL 10:1 dilution
50fg on column
x 4400 more signal
GC APCI
EPA1613 CS2
10pg on column EI

18. Detection Examples - OCDD
EPA1613 CS1 10:1 dilution
500fg on column
GC APCI
X 8400 more signal
EPA1613 CS5
2ng on column EI

19. Background – Why fix what is not broken?
GC-HRMS (sector) has been the “gold standard” for decades
Existing systems are robust (relatively) and integrated systems
Reference data all based on HRMS data
Small range of target compounds allows for simple descriptors (17-TEF compounds)
Robust prep method removes most (?) interferences

20. Is anything “broken”
17 TEF’s may not be enough for a true TEQ determination
The current targets are certainly not enough to truly characterize a source or environmental impact
There is the potential to have increasing levels of mixed halo and poly-Br compound formation in more modern samples

21. What is something they all have in common?Δ
+ source of chlorine Δ
+ source of chlorine
Source of chlorine = PVC and plastics. If the product is old enough it might even have Chlorinated Flame Retardants in it also.
De novo synthesis vs Synthesis from precursors  precursor pathway more dominant when coming from BFRs 5 4 6 3 7
Br, Cl
Br, Cl
Br, Cl
Br, Cl 2 8 1 10 9
Dibenzofuran
Dibenzo-p-dioxin

22. Do we care about other compounds?
Very few studies of the mixed halo congeners
Analytical
Biochemical
Do they follow similar chemistry as the polychloro analogs?

23. To Investigate Further
Analytical Approach?
What technique/s
Reference Materials?
Very few are available
Sample Accessibility?
How do we obtain “real” samples

24. Simulated burn studies

25. Simulated burn studies
Household Fire
Mattress
Sofa Chair
Vinyl / Wood Chair
Carpet
Pillows
Television
Electronics Fire
Televisions
Microwave
Printers
Computer monitors
Laptop
Cables/Wires

26. Simulated burn studies

27. Characterizing Fire Debris Samples: Mixed Halogenated Dioxins

28. Reference Standards BrCl2 dibenzofuran BrCl3 dibenzofuran
PXDF standards

29. Wire Extract BrCl dibenzofuran BrCl2 dibenzofuran Br2Cl dibenzofuran
TICs for each Br/Cl congener group in FESTI wire extract

30. No peaks of this congener class detected on TOF
Household Fire
BrCl dibenzofuran
No peaks of this congener class detected on TOF
PtP S:N = 6:1
Electronics Fire
Wires
BrCl dibenzofuran
An example of the sensitivity of the instrument. The top cgram (firefighter helmet wipe from the household fire) is an extract that I didn’t see any polyBr or BrCl compounds in on the TOF. The bottom cgram (electronics fire burnt wires) I did see this furan in, but at low levels. I picked an arbitrary peak that looked like it was in both samples and did a peak to peak S:N calculation on them.
PtP S:N = 38:1

31. Only one peak of this congener class detected on TOF
Electronics Fire
Dry Wall
Br2Cl dibenzofuran
PtP S:N = 233
PtP S:N = 178
Only one peak of this congener class detected on TOF
PtP S:N = 134
This is the the Br2Cl DFs in the dry wall extract. I picked this one because I only saw one peak for this compound in the TOF data. The S:N values are probably a little high because there wasn’t much “noise” area to use to calculate. You may want to remove them from the slide. I was just curious how different the S:N values were since I only saw one peak on the TOF. I’m thinking the peaks at and could have co-eluted on the TOF and that was the single peak I saw.

32. Comparing the number of each PXDF congener identified in the wires sample on both the TOF and Xevo. It was hard to count exactly the number in the Xevo data because they were mostly unresolved. I erred on the low side.

33. Conclusions
APGC-TQ-S allows for considerable improvement in sensitivity
20-40 X versus Autospec
Mixed-halo congeners can be quantified (though not identified)
They are found in most all fire debris samples studied so far…

34. Acknowledgements Penn State University Frank L. Dorman
Ontario Ministry of Environment
Eric Reiner, Terry Kolic, Karl Jobst
FESTI
Brian Ross, Mike Hutchison
Sam Marshall, Pike Krpan
Restek
Purdue University
Steve Ayrton
Waters
David Douce, Murray Booth
Dave Gordon