Determination of Chlorinated Dioxins/Furans in

Mainstream Cigarette Smoke using Gas

Chromatography-Tandem Mass Spectrometry

Bill Guthery, Filtrona Technology Centre

CORESTA SSPT Meeting 2013, Seville

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Background • Chlorinated dioxins and furans (“dioxins”) have been listed

as a harmful and potentially harmful constituent (HPHC) in tobacco smoke.

• Dioxins consist of a polychlorinated dibenzo structure bridged by either 2 oxygens (PCDD) or 1 oxygen (PCDF).

• There are 210 congeners of which 17 tetra- through octa- substituted congeners have been classified as ‘significantly toxic’ by the World Health Organisation (WHO).

• Formed during manufacture and combustion processes in which a large amount of material containing chlorine is present.

• Toxic effects associated with these chemicals include skin disorders, cardiovascular disease, developmental and reproductive toxicity, immunotoxicity and cancer. 1

1 De Vito MJ, Birnbaum LS. Toxicology of dioxins and related chemicals . In: Schecter A, ed. Dioxins and health. New York: Plenum Press; 1994:139–62.

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Toxic Activity

• Toxic activity is mediated through the aryl hydrocarbon receptor (AHR) with a

high affinity for dioxins containing lateral chlorines at the 2-,3-,7-, and 8-

positions.

• ‘Total Toxic Equivalence’ (TEQ) is the sum of the concentrations multiplied by

a toxic equivalence factor (TEF) relative to the most toxic component, 2,3,7,8-

tetrachlorodibenzo-p-dioxin (2378-TCDD) (Table 1).

• United States Environmental Protection Agency (US EPA) adopted this

procedure in 1989 and it has since been adopted by WHO.

• EPA safe limit is defined as 0.7 pg TEQ/kg body weight per day. 2

• WHO limit 1-4 pg TEQ/kg body weight per day.

2,3,7,8-Tetrachlorodibenzo-p-dioxin TEF = 1

2,3,7,8-Tetrachlorodibenzo[b,d]furan TEF = 0.1

2 R. Trager, Chemistry World; Feb 23, 2012

PCDD TEF

2378-TCDD 1

12378-PCDD 1

123478-HxCDD 0.1

123678-HxCDD 0.1

123789-HxCDD 0.1

1234678-HpCDD 0.01

OCDD 0.0003

PCDF

2378-TCDF 0.1

12378-PCDF 0.03

23478-PCDF 0.3

123478-HxCDF 0.1

123678-HxCDF 0.1

234678-HxCDF 0.1

123789-HxCDF 0.01

1234678-HpCDF 0.01

1234789-HpCDF 0.01

OCDF 0.0003

Table 1. TEF Values (WHO Revised 2005)

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Dioxin Analysis

• Standard methods are available for determination of these chemicals.

- (US EPA Method 1613, US EPA Method 23, JSA JIS K 0311:2005, JSA JIS K 0312, EN 1948 : 2006)

• Analysis by Gas Chromatography (GC) / High-Resolution (>10,000) Mass Spectrometry (HRMS).

• ‘Performance-based’ methods to be adapted to suit type of matrix under investigation.

• Isotope dilution → extraction → solid-phase extraction (SPE) – multi-step

• Dioxins have been measured in tobacco smoke using HRMS. 3, 4

• Ball et al. reported dioxin concentrations in 10 German brands.

- Only two constituents were clearly resolved from the matrix (1234678-HpCDD and OCDD) with reported concentrations between 0.8 and 5.4 pg/cig.

3 M. Ball et al., Beiträge zur Tabakforschung International, 14/6, 1990 4 G. Löfroth and Y. Zebühr; Bull. Environ. Contam. Toxicol.; 48; 1992

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Aims & Objectives

• Determine whether any dioxins are present at detectable levels in

the cigarette mainstream smoke (MSS) of Kentucky and

CORESTA reference monitors.

• Evaluate a clean-up procedure using the Supelco ‘Dioxin Prep

System’. 5

• Assess the reproducibility of Gas Chromatography/Triple

Quadrupole Mass Spectrometry (GC-MS/MS).

• Develop a validated protocol with method detection limits (MDL) at

the ‘safe’ levels.

5 Maeoka, M. et al., Quick Pretreatment Prep Study for Dioxin Analysis. 11th Symposium on Japan Environmental Chemistry (June 3-5), 521- 532, 2002.

Supelco Dioxin Prep System

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Smoking Procedure

• Reference cigarettes (K3R4F and CM7) smoked under

ISO 3308 and Health Canada Intense (HCI) conditions

on a rotary machine.

• Smoke condensate from 20 Cigs (ISO) or 10 cigs (HCI)

trapped onto a Cambridge filter (CF) pad.

• Isotope labelled (13C12 PCDD/F) internal standards

(ISTDs) were spiked onto the CF pad.

• Pad is extracted into 50 ml hexane and shaken for 45

mins.

• Second extract into 50 ml hexane.

• Clean-up using Dioxin Prep System

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Dioxin Prep System

• Two-step clean up procedure consisting of

a multi-layer silica gel column and dual-

layer reversible carbon column (Fig. 1)

• Multi-layer column consists of four

activated silica gel layers which removes

many of the polar interferences from the

matrix such as sulphur compounds, bases

and lipids, acidic compounds.

• Dual-layer carbon column consists of low-

(75 m2/g) and high-(1200 m2/g) surface

area layers which trap dioxins which are

then eluted with polar solvent.

• Columns are connected by PTFE unions

during the load cycle

Fig. 1 Multi-layer silica gel column and dual-layer carbon column

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Clean-up Procedure

• Condition

- Multi-layer 200 ml hexane

- Dual-layer 20 ml toluene then 40 ml hexane

• Load

- Columns are connected using PTFE reducing unions.

- Sample extracts are loaded onto the columns via top

flasks passed through to waste.

- Dioxins are retained on the dual-layer columns.

- Wash with 3 % dichloromethane in hexane.

• Elution

- 50 ml toluene / reversed column.

- Concentrated 0.5 ml using a TurboVap (60°; ~15psi).

Transfer to amber autosampler vial for analysis.

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

GC-MS/MS Analysis

• Agilent 7890 GC / 7000B Triple Quadrupole Mass Spectrometer

• MS Analyser consists of 2 x Quads and a Hexapole Collision Cell (Fig. 2).

• Resolves ions (m/z) in 3 dimensions

- Q1 Precursor ion→Q2 Cv (N2) →Q3 Product ion

• HED Off-Axis Detector (Fig. 3). Removes interferences from ion stream e.g. e-, He+, hυ

Fig. 2 Triple stage analyser Fig. 3 HED Off-Axis Detector

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

GC-MS/MS Conditions

Instrument: Agilent 7890 GC / 7000B MS/MS

Software: Agilent MassHunter Workstation

Column: Restek Rxi-5ms (30m x 0.25mm; 0.25mm)

Inlet: 280°C; 1ml splitless (1 min)

Carrier: Helium; 1.2 ml/min constant flow

Oven: 90°C (1 min), 40°C/min to 200°C (0.5 min), 3°C/min to 310°C (3 min)

Total RT: 43.9 min

Transfer Line: 300°C

Source: 280 °C (EI; 70 eV)

Quadrupole 1 and 2: 150°C

Collision Cell Gas 1: Helium (2.3 ml/min)

Collision Cell Gas 2: Nitrogen (1.5 ml/min)

Collision Energy: Optimised to compound (Table 2)

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Multiple Reaction Monitoring (MRM)

• 17 dioxin congeners and 10 13C labelled analogues (one from each homologue group)

• Collision energies optimised using Product Ion Scan (0-60 V)

Table 2. Optimised Collision Energies

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

GC Separation/ MS Resolution

• K3R4F CF pad spiked with 125 pg of each component equivalent to 6.3 pg per cigarette

• Separation of 17 dioxin congeners (Fig. 4)

• Separation of 10 isotope labelled analogue congeners (Fig. 5)

Fig. 4 Overlaid MRM Chromatogram of spiked K3R4F extract (6.3 pg/cig)

Fig. 5 Overlaid MRM Chromatogram of 13C analogues (50 ng/ml)

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Recovery Data

• Calculated from the ratio of the 13C analogues signal response from K3R4F extracts against calibration

standards. Average of 3 x replicates.

• Recoveries of 86.3% to 125.7% (ISO) and 74.5% to 118.4% (HCI) were obtained (Table 3).

• EPA Method 23 – QC requirements are 40 – 130% (tetra-, penta-, hexa-); 25 – 130% (hepta-, octa-)

ISO

50 ng/ml TCDF 13C12 TCDD 13C12 PCDF 13C12 PCDD 13C12 HxCDF 13C12 HxCDD 13C12 HpCDF 13C12 HpCDD 13C12 OCDD 13C12 OCDF 13C12

Area recovery 86.3% 90.2% 97.0% 95.2% 114.9% 105.8% 122.4% 125.7% 125.4% 125.3%

HCI

50 ng/ml TCDF 13C12 TCDD 13C12 PCDF 13C12 PCDD 13C12 HxCDF 13C12 HxCDD 13C12 HpCDF 13C12 HpCDD 13C12 OCDD 13C12 OCDF 13C12

Area recovery 74.5% 77.4% 82.2% 80.4% 96.0% 90.0% 118.4% 109.0% 115.5% 113.0%

Table 3. 13C Internal Standard Recoveries

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Calibration Standards

• Standards were prepared in toluene

- 0.10 – 50 ng/ml tetra-, penta-, hexa- ; 0.25 – 50 ng/ml hepta-, octa- ; 50 ng/ml ISTD

- Linear curve fit (R2 >0.999) in all cases

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Dioxin Yields in Reference Cigarettes

• No dioxins were detected in K3R4F, K1R5F or CM7 cigarettes with or without filters under ISO 3308 or

HCI conditions

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Limit of Detection (LOD) / Limit of Quantification (LOQ)

• Cigarettes smoked under ISO 3308 conditions and the CF pad spiked with 125 pg of each dioxin.

• Empirical LOD calculated at 4.3 pg TEQ/cig (Table 4)

Table 4. LOD/LOQ – Nominal Pad Spike 6.3 pg/cig (n = 10)

PCDD mean SD LOD (SD*3) LOQ (SD*10) LOD TEQ pg/cig LOQ TEQ pg/cig

2378-TCDD 7.3 0.33 1.0 3.3 1.00 3.34

12378-PCDD 7.0 0.55 1.6 5.5 1.65 5.49

123478-HxCDD 7.6 1.28 3.8 12.8 0.38 1.28

123678-HxCDD 7.0 0.83 2.5 8.3 0.25 0.83

123789-HxCDD 7.0 0.55 1.7 5.5 0.17 0.55

1234678-HpCDD 6.5 1.08 3.2 10.8 0.03 0.11

OCDD 7.3 0.69 2.1 6.9 0.0006 0.0021

PCDF

2378-TCDF 7.3 0.27 0.8 2.7 0.08 0.27

12378-PCDF 7.4 0.48 1.4 4.8 0.04 0.14

23478-PCDF 7.0 0.27 0.8 2.7 0.24 0.80

123478-HxCDF 7.5 0.68 2.0 6.8 0.20 0.68

123678-HxCDF 6.7 0.29 0.9 2.9 0.09 0.29

234678-HxCDF 6.4 0.34 1.0 3.4 0.10 0.34

123789-HxCDF 6.2 0.24 0.7 2.4 0.01 0.02

1234678-HpCDF 7.4 0.56 1.7 5.6 0.02 0.06

1234789-HpCDF 7.0 1.16 3.5 11.6 0.03 0.12

OCDF 7.0 1.05 3.2 10.5 0.0009 0.0032

4.3 14.3Overall TEQ pg/cig

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Transfer Rates and Estimate of Uncertainty

• K3R4F cigarettes spiked with 500 pg in 1 ml ~20-30 mm from the front end of the rod.

• 10 replicates performed under ISO and HCI regimes.

• Measurable amounts of dioxins found for all components (Table 5).

• ISO recoveries 3.9% to 11.1% (median 9.0%).

• HCI recoveries 9.2% to 27.2 (median 22.0%).

• Estimates of Uncertainty U calculated from the relative standard deviation (RSD) of the 10 determinations.

Table 5. Dioxin transfer rates from rods spiked with 500 pg (1536 pg TEQ/cig)

Regime

PCDD

2378-TCDD 55.5 11.1% 22.3% 135.7 27.1% 14.0%

12378-PCDD 51.3 10.3% 18.3% 131.5 26.3% 17.1%

123478-HxCDD 45.7 9.1% 16.5% 116.0 23.2% 19.3%

123678-HxCDD 44.8 9.0% 17.2% 110.1 22.0% 16.0%

123789-HxCDD 33.7 6.7% 32.1% 82.3 16.5% 20.3%

1234678-HpCDD 43.3 8.7% 21.6% 109.8 22.0% 13.7%

OCDD 45.8 9.2% 32.5% 110.2 22.0% 24.6%

PCDF

2378-TCDF 53.0 10.6% 15.0% 136.0 27.2% 9.4%

12378-PCDF 51.6 10.3% 16.0% 128.2 25.6% 11.0%

23478-PCDF 25.5 5.1% 25.9% 61.4 12.3% 48.6%

123478-HxCDF 45.1 9.0% 17.3% 117.6 23.5% 8.4%

123678-HxCDF 47.6 9.5% 28.4% 129.6 25.9% 34.1%

234678-HxCDF 19.3 3.9% 21.2% 46.0 9.2% 38.0%

123789-HxCDF 23.8 4.8% 17.9% 61.5 12.3% 21.8%

1234678-HpCDF 44.2 8.8% 16.2% 111.5 22.3% 18.9%

1234789-HpCDF 35.9 7.2% 22.1% 90.7 18.1% 36.7%

OCDF 35.8 7.2% 23.5% 91.8 18.4% 20.6%

Total Dioxins pg

TEQ/cig

U (k = 2)

146.5 367.0

Mean

(pg/cig)Recovery U (k = 2)

9.5% 23.9%

ISO 3308 (n = 10) HCI (n = 10)

Mean (pg/cig) Recovery

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Transfer Rates (2)

• K3R4F rods spiked with 63 pg, 125 pg and

250 pg of each constituent equivalent to

192, 384 and 768 pg TEQ/cig (Table 6).

• Cigarettes smoked under ISO conditions

and HCI (one level 125 pg/cig).

• 250 pg spike – all components were

quantifiable 9 to 67 pg/cig equivalent to 160

pg TEQ/cig.

• 125 pg spike – all components detectable

(ISO and HCI). One <LOQ (ISO).

• 63 pg spike - 13 out of 17 components were

detected with 6 <LOQ.

Table 6. Dioxin transfer rates from rods spiked with 63, 125 and 250 pg

Regime HCI

PCDD

2378-TCDD 8.4 9.1 33.7 30.2

12378-PCDD <5.5 8.6 27.6 23.8

123478-HxCDD 6.8 14.8 28.8 23.1

123678-HxCDD <LOD 14.6 25.2 20.2

123789-HxCDD <5.5 15.4 8.7 22.5

1234678-HpCDD <LOD 5.7 24.9 24.6

OCDD 12.5 13.3 30.2 35.2

PCDF

2378-TCDF 8.7 10.5 30.6 25.0

12378-PCDF 6.8 10.9 28.5 24.8

23478-PCDF <2.7 8.5 7.6 16.8

123478-HxCDF <6.8 17.6 23.8 24.9

123678-HxCDF 5.0 18.0 24.8 22.8

234678-HxCDF <LOD <3.4 4.6 13.1

123789-HxCDF <2.4 3.1 11.0 14.0

1234678-HpCDF <5.6 10.0 24.6 22.0

1234789-HpCDF <11.6 9.5 16.4 28.7

OCDF <LOD 7.9 19.0 19.8

Recovery (%) 8.8 7.9 10.4 19.8

ISO 3308

125 pg/cig

75.9

63 pg/cig 125 pg/cig 250 pg/cig

Total Dioxins pg

TEQ/cig16.8 30.2 79.8

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Inter- and Intra- Day Precision

• Relative recoveries determined under ISO

and HCI conditions.

• Multiple determinations made in triplicate.

• Standard error (SE) was used to calculate

confidence intervals at 95% and 99%.

• Used as a monitor control

• Repeatability (CV.r) was derived from the

standard deviation from each

determination.

Regime ISO 3308 (n = 11) HCI (n = 9)

Mean

(pg/cig) SE (%) CV.r (%)

Mean

(pg/cig) SE (%) CV.r (%)

PCDD

2378-TCDD 50.7 5.2% 2.5% 100.6 4.1% 3.4%

12378-PCDD 49.0 7.8% 4.3% 98.3 3.9% 4.9%

123478-HxCDD 51.8 6.0% 4.3% 100.7 8.7% 4.0%

123678-HxCDD 49.1 6.8% 3.7% 95.9 4.1% 6.5%

123789-HxCDD 49.6 5.8% 6.6% 98.2 9.7% 7.0%

1234678-HpCDD 49.0 3.9% 5.8% 98.8 5.4% 4.6%

OCDD 52.7 6.0% 5.2% 104.5 8.0% 9.4%

PCDF

2378-TCDF 49.3 5.2% 2.8% 98.5 3.9% 4.0%

12378-PCDF 49.3 6.0% 3.0% 97.3 3.1% 4.5%

23478-PCDF 46.5 7.2% 4.9% 89.4 6.3% 4.9%

123478-HxCDF 49.8 4.9% 2.8% 98.0 6.5% 4.6%

123678-HxCDF 49.7 8.5% 6.0% 96.3 7.0% 7.5%

234678-HxCDF 44.5 12.6% 7.0% 85.3 8.0% 9.0%

123789-HxCDF 47.0 8.6% 5.5% 90.5 10.3% 1.9%

1234678-HpCDF 48.5 5.4% 3.8% 98.6 5.3% 6.1%

1234789-HpCDF 56.4 3.5% 3.2% 104.6 6.6% 6.2%

OCDF 47.9 6.8% 8.6% 96.9 5.8% 8.1%

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Inter-Day Precision Control Charts (ISO)

• Ten determinations were made in triplicate equivalent of 50 pg/cig

• Relative recoveries obtained between 89.1 and 112.7% (median 98.6%)

K3R4F Pad Spike ISO Control

35

40

45

50

55

60

1 3 5 7 9 11

Test number

pg

/cig

2378-TCDF

2378-TCDD

12378-PCDF

23478-PCDF

12378-PCDD

123478-HxCDF

123678-HxCDF

234678-HxCDF

123478-HxCDD

123678-HxCDD

123789-HxCDD

123789-HxCDF

1234678-HpCDF

1234678-HpCDD

1234789-HpCDF

OCDD

OCDF

Nominal

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Inter-Day Precision Control Charts (HCI)

• Nine determinations were made in triplicate equivalent of 100 pg/cig

• Relative recoveries obtained between 85.3 and 104.6% (median 98.2%)

K3R4F Pad Spike HCI Control

70.0

80.0

90.0

100.0

110.0

120.0

1 2 3 4 5 6 7 8 9

Test number

pg

/cig

TCDF

TCDD

12378-PCDF

23478-PCDF

12378-PCDD

123478-HxCDF

123678-HxCDF

234678-HxCDF

123478-HxCDD

123678-HxCDD

123789-HxCDD

123789-HxCDF

1234678-HpCDF

1234678-HpCDD

1234789-HpCDF

OCDD

OCDF

Nominal

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA

Conclusion

• Dioxins were not present at detectable levels in K3R4F and CM7 (<4.3 pg TEQ/cig).

• Transfer rates from spiked K3R4F rods were 9.5% (ISO) and 23.9% (HCI).

• Measurable amounts as low as 16.8 pg TEQ/cig were obtained from rods spiked with

63 pg each congener. Scope to detect lower?

• Recoveries using Dioxin Prep system were comparable with standard methods. Matrix

interference was eliminated.

• Triple Quad MS provides a viable alternative to HRMS with comparable detection

limits.

2013

_ST

26_G

uthe

ry.p

dfS

SP

T20

13 -

Doc

umen

t not

pee

r-re

view

ed b

y C

OR

ES

TA