GC-MS Fingerprinting Method  Pros & Cons of the Current Fingerprinting Method  Why Switching to a 7200 Agilent High Resolution MS Instrument?

‘Next Generation Fingerprinting’

A GC-HR-TOF-MS Method to Semi-Quantify Constituents in Aerosols and Aerosol Fractions

E. Dossin, A. Monge, E. Martin, P. Pospisil, A. Knorr, M.C. Bentley, P.A. Guy

Metabolomics & Analytical Chemistry Gr., Philip Morris Products SA, 2000 Neuchatel, Switzerland Outline

1. Description of the existing GC-MS fingerprinting method  Pros & cons of the current fingerprinting method  Why switching to a 7200 Agilent high resolution MS instrument?

2. How to tackle semi-quantification of the smoke constituents with the help of chemoinformatics tools  Curation step for smoke constituents to be monitored  Analysis of reference standards  Retention time prediction model (QSPR approach)  Selection of appropriate internal standards (clustering approach) 3. Semi-quantification from calibration curve of reference standards  Linearity  Assessment of silylation

4. What about other smoke constituents?

5. Conclusion

Page: 2 Existing Fingerprinting Method

• Aerosol sample generated from a smoking machine (ISO) – Whole smoke – Gas Vapor Phase (GVP) sbPBS – Total Particulate Matter (TPM)

• Compounds list

• GC columns (HP6890 GC) – DB-624: HS-SPME-GC-MS Volatile Chemicals – DB-FFAP: GC-MS Non Polar Chemicals – DB-5-MS: HT-GC-MS Polar Chemicals (TMS)

• Detection (MSD5973 MS) – Electron ionization mode

– Full scan (low resolution) Semi-quantification (d6-phenol)

Page: 3 Description of the Former GC-MS Fingerprinting Method (HP6890 GC - MSD5973 MS)

Cell culture Medium plus Cigarette DB-624 column

● Headspace(HS)-SPME-GC-MS Smoke MEK

high volatile compounds MEK

benzenebenzene

toluene toluene

acetone acetone

methylfuran methylfuran

- -

2 2

acetaldehyde acetaldehyde acrylonitrile acrylonitrile e.g.: up to 51 chemicals acetaldehyde toluene Cigarette Smoke Total Particulate Abundance DB-FFAP column ● GC-MS Fingerprint (Standard Method) 5e+07 Matter

4e+07

acid acid

nicotine nicotine triacetine nonpolar compounds Abundance triacetine

5e+073e+07

glycerine glycerine

neophytadieneneophytadiene

4e+072e+07 palmitic

palmitic

acid acid

nicotine nicotine

hydroquinonehydroquinone

triacetinetriacetine aceticacid

aceticacid limonene

3e+071e+07 limonene

phenol phenol

glycerine glycerine toluene

up to toluene neophytadiene 0 neophytadiene

2e+07 palmitic e.g.: 25.00 palmitic 10.00 30.00 35.00

Time--> 5.00 15.00 20.00 hydroquinone

toluene pyrene hydroquinone

aceticaceticacid acid limonene

1e+07 limonene 420 chemicals phenol

5e+07 phenol toluene toluene carbohydrates higher mass compounds 0 Time--4e+07> 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Cigarette Smoke Total Particulate DB-5-MSsolanesol column 3e+07 5e+07 steroids, waxes ● High Temperature(HT)-GC-MS Matter carbohydrates tocopherolshigher mass compounds 2e+07 derivatized, polar compounds 4e+07 1e+07 solanesol 3e+07 steroids, waxes

0 tocopherols 2e+075.00 10.00 15.00 20.00 25.00 30.00 35.00

1e+07 up to e.g.: 0 5.00 10.00 15.00 20.00 25.00 30.00 35.00 phenol (±)-α-tocopherol 155 chemicals A total of ~607 smoke constituents are monitored using 3 distinct GC columns but not all were unambiguously identified

Page: 4 Smoke Constituents Targeted in the Previous Method aliph. Hydrocarbons aliphatic hydrocarbon-isom.4 propylbenzene methyl-fluorene anethol benzoic acid isovaleramide-isom.2 aliph. Hydrocarbons aliphatic hydrocarbon-isom.4 propylbenzene methyl-fluorene anethol benzoic acid isovaleramide-isom.2 1,3-butadiene1,3-butadiene unsaturated hydrocarbon-isom.4 unsaturatedC3-alkyl.benzene-isom.1 hydrocarbon-isom.4phenanthrene methylanisole-isom.1C3-alkyl.benzene-isom.12-furanecarboxolicphenanthrene acid 4-methyl-pentaneamide methylanisole-isom.1 2-furanecarboxolic acid 4-methyl-pentaneamide isoprene unsaturated hydrocarbon-isom.5 C3-alkyl.benzene-isom.2 fluoranthene alcoholes (without glycerine) phenylacetic acid picolinamide isoprene1,3,5-heptatriene 4,4'-dimethylbiphenyl unsaturatedC3-alkyl.benzene-isom.3 hydrocarbon-isom.5pyrene 2-methoxyethanolC3-alkyl.benzene-isom.2phenylpropionicfluoranthene acid glykolamide alcoholes (without glycerine) phenylacetic acid picolinamide 1,3,5-cycloheptatriene aliphatic hydrocarbon-isom.5 styrene o-heterocycles 2-methoxypropanol niacine phenylacetamide 1,3,5-heptatrienedimethyl-methylenecyclohexene unsaturated hydrocarbon-isom.6 4,4'-dimethylbiphenylC3-alkyl.benzene-isom.4 furan propylenglykoleC3-alkyl.benzene-isom.3long-chainedpyrene acids (>=C8) N-methyl-nicotinamide 2-methoxyethanol phenylpropionic acid glykolamide terpene (type:limonene) unsaturated hydrocarbon-isom.7 2-propenylbenzene 2,5-dimethylfuran 2-furanmethanole caprylic acid (C8:0) pyrrole-2-carboxamide 1,3,5-cycloheptatriene2,7-dimethyl-1,6-octadiene aliphatic hydrocarbon-isom.6 aliphaticC4-alkyl.benzene-isom.1 hydrocarbon-isom.52,4-dimethylfuran beta-citronellolstyrene decanoic acid(C10:0) niacinamide 2-methoxypropanol niacine phenylacetamide 2,6-dimethyl-2,6-octadiene unsaturated diterpen C3-alkyl.benzene-isom.5 2,3,5-trimethylfuran benzylalcohol dodecanoic o-heterocyclesacid(C12:0) hexadecaneamide p-menthene (hydrocarbon-terpene) unsaturated hydrocarbon (ca. C25) C4-alkyl.benzene-isom.2 vinylfurane beta-ionol myristic acid (C14:0) O-methyl-N-methylcarbamate dimethyl-methylenecyclohexene1-undecene unsaturated hydrocarbon-isom.8 unsaturatedC4-alkyl.benzene-isom.3 hydrocarbon-isom.62-vinyl-5-methylfurane farnesol-structure1C3-alkyl.benzene-isom.4myristoleic acid(C14:1)furan amines propylenglykole long-chained acids (>=C8) N-methyl-nicotinamide trimethylheptatriene unsaturated hydrocarbon-isom.9 C4-alkyl.benzene-isom.4 benzofuran glycerine-monoacetate pentadecanoic acid (C15:0) dimethylaminoacetonitrile terpenemethyl-methyleneoctadiene (type:limonene)heptacosane (C27H56) unsaturatedmethyl-styrene-isom.1 hydrocarbon-isom.7methylbenzofuran glycerine2-propenylbenzeneBroad palmiticchemical acid2,5-dimethylfuran (C16:0) diversityN,N-dimethyl-2-aminoethanol 2-furanmethanole caprylic acid (C8:0) pyrrole-2-carboxamide beta-myrcene aliphatic hydrocarbon-isom.7 C3-alkyl.benzene-isom.6 1,3-dihydroisobenzofurane farnesol-isom.3 heptadecanoic acid (C17:0) urotropine 2,7-dimethyl-1,6-octadienelimonene 2-methyloctacosane(C29H60) aliphaticC4-alkyl.benzene-isom.5 hydrocarbon-isom.62,3-dihydrobenzofurane geranyllinaloolC4-alkyl.benzene-isom.1stearic acid 2,4-dimethylfuran(C18:0) not identified amine 1 beta-citronellol decanoic acid(C10:0) niacinamide dodecane aliphatic hydrocarbon-isom.8 methyl-styrene-isom.2 steroids derivatives cis-phytol categorizedoleic acid in (C18:1) chemicalnot identified aminefamilies 2 2,6-dimethyl-2,6-octadienedodecene aliphatic hydrocarbon-isom.9 unsaturatedethinylbenzene diterpenstigmasterolacetate trans-phytolC3-alkyl.benzene-isom.5linolic acid(C18:2)2,3,5-trimethylfurannot identified amine 3 benzylalcohol dodecanoic acid(C12:0) hexadecaneamide 2,7-dimethyl-1,3,6-octatriene aliphatic hydrocarbon (ca. C30H62) C4-alkyl.benzene-isom.6 beta-sitosterolacetate farnesol structure2 linolenic acid… (C18:3) but halogenated hydrocarbons p-menthenetridecane (hydrocarbon-terpene)triacontane(C30H62) unsaturatedpropenylbenzene hydrocarbonesters + ethers (without (ca. triacetine) C25) isomentholC4-alkyl.benzene-isom.2nitriles vinylfuranechloromethane beta-ionol myristic acid (C14:0) O-methyl-N-methylcarbamate aliphatic hydrocarbon-isom.1 2-methyltriacontane(C31H64) propenyltoluene-isom.1 3-butenoic acid methylester phytuberol smoke constituentsacrylnitrile sulfur were compounds not 1-undecene1,3,8-p-menthatriene unsaturated hydrocarbon-isom.11unsaturatedpropenyltoluene-isom.2 hydrocarbon-isom.82-propenylformiate sugars,C4-alkyl.benzene-isom.3 sum: isopropylnitrile2-vinyl-5-methylfuranemethional ((3-methylthio)-propanal) farnesol-structure1 myristoleic acid(C14:1) aliphatic hydrocarbon-isom.2 unsaturated hydrocarbon-isom.12 indene isoamylbutyrate dianhydro-glucopyranose acetonitrile methanethiol amines aliphatic hydrocarbon-isom.3 hentriacontane(C31H64) methyl-indene-isom.1 glycoldiacetate 3,4-anhydro-d-galactosan wellpropionenitrile definedaldehydes trimethylheptatrienecopaene aliphatic hydrocarbon-isom.10 unsaturatedmethyl-indene-isom.2 hydrocarbon-isom.9geranylacetate levoglucosanC4-alkyl.benzene-isom.42-methylbutanenitrilebenzofurancrotonaldehyde glycerine-monoacetate pentadecanoic acid (C15:0) dimethylaminoacetonitrile pentadecane aliphatic hydrocarbon (C32H66) methyl-indene-isom.3 benzylacetate short-ch. + ar. Acids 3-methylbutanenitrile 3-methyl-2-butenal methyl-methyleneoctadiene1-pentadecene aliphatic hydrocarbon-isom.11 heptacosanedimethylindene-isom.1 (C27H56)dimethylsuccinate aceticmethyl-styrene-isom.1 acid crotonnitrilemethylbenzofuran2-hexenal glycerine palmitic acid (C16:0) N,N-dimethyl-2-aminoethanol 1-methylcyclooctene 2-methyldotriacontane(C33H68) azulene triacetine formic acid 4-methyl-pentanenitrile 2,4-hexadienal beta-myrceneunsaturated hydrocarbon-isom.1 aliphatic hydrocarbon (C33H68) aliphaticdimethylindene-isom.2 hydrocarbon-isom.7methylpalmitate propionicC3-alkyl.benzene-isom.6 acid 2-ethylideneamino-propionnitrile1,3-dihydroisobenzofuranefurfural farnesol-isom.3 heptadecanoic acid (C17:0) urotropine 1-hexadecene unsaturated hydrocarbon-isom.13 ethylindene glycerin-diacetate 2-methyl-propionic acid benzonitrile 1-(2-furanyl)-ethanone limonenesesquiterpene(unsaturated KW) beta-caryophyllene 2-methyloctacosane(C29H60)naphthalene ethylpalmitate butyricC4-alkyl.benzene-isom.5 acid tolunitrile-isom.12,3-dihydrobenzofuranebenzaldehyde geranyllinalool stearic acid (C18:0) not identified amine 1 heptadecane alpha-caryophyllene trimethylindene methylstearate acrylic acid tolunitrile-isom.2 methylimidazole-carboxaldehyde dodecanefarnesene arom. Hydrocarbons aliphatic2-methylnaphthalene hydrocarbon-isom.86-acetoxy-2,7,11-cembratriene 2-/3-methylbutyricmethyl-styrene-isom.2 acid nicotinonitrile 5-methyl-2-furfural cis-phytol oleic acid (C18:1) not identified amine 2 unsaturated hydrocarbon-isom.2 toluene 1-methylnaphthalene methyl-pyroglutamate isocrotonic acid benzylnitrilesteroids derivativestolualdehyde delta-cadinene ethylbenzene dimethylnaphthalene benzyl benzoate crotonic acid amides cinnamaldehyde dodecene3,7,11,15-tetramethyl-2-hexadecene xylene-isom.1 aliphaticbiphenyl hydrocarbon-isom.9muskalactone 3-methyl-valericethinylbenzene acid acetamide stigmasterolacetatep-ethylbenzaldehyde trans-phytol linolic acid(C18:2) not identified amine 3 neophytadiene xylene-isom.2 methyl-biphenyl methyl-salicylate capronic acid (C6:0) propionamide acetaldehyde 2,7-dimethyl-1,3,6-octatrieneunsaturated hydrocarbon-isom.3 isopropylbenzene aliphatictrimethylnaphthalene hydrocarbon1,1-dimethyl-2-phenylethylbutyrate (ca. C30H62) 2-methyl-crotonicC4-alkyl.benzene-isom.6 acid isovaleramide-isom.1beta-sitosterolacetatetrimethoxy-acetophenone(Ga farnesol structure2 linolenic acid (C18:3) halogenated hydrocarbons tridecane1,4-eicosadiene xylene-isom.3 triacontane(C30H62)fluorene triethylcitrate sorbicpropenylbenzene acid acrylamide esters + ethers (without triacetine) isomenthol nitriles chloromethane aliphatic hydrocarbon-isom.1 2-methyltriacontane(C31H64) propenyltoluene-isom.1 3-butenoic acid methylester phytuberol acrylnitrile sulfur compounds 1,3,8-p-menthatrieneA time consumingunsaturated curation hydrocarbon-isom.11 of smoke constituentspropenyltoluene-isom.2 was required2-propenylformiate to register sugars, sum: isopropylnitrile methional ((3-methylthio)-propanal) aliphatic hydrocarbon-isom.2every compoundunsaturated as Unique hydrocarbon-isom.12 Compoundindene Spectral Databaseisoamylbutyrate (UCSD) dianhydro-glucopyranose acetonitrile methanethiol aliphatic hydrocarbon-isom.3 hentriacontane(C31H64) methyl-indene-isom.1 glycoldiacetate 3,4-anhydro-d-galactosan propionenitrile Martin et al. Building an R&D chemical registration system, J. Cheminformatics 2012, 4:11. aldehydes copaene aliphatic hydrocarbon-isom.10 methyl-indene-isom.2 geranylacetate levoglucosan 2-methylbutanenitrile crotonaldehyde pentadecanePage: 5 aliphatic hydrocarbon (C32H66) methyl-indene-isom.3 benzylacetate short-ch. + ar. Acids 3-methylbutanenitrile 3-methyl-2-butenal 1-pentadecene aliphatic hydrocarbon-isom.11 dimethylindene-isom.1 dimethylsuccinate acetic acid crotonnitrile 2-hexenal 1-methylcyclooctene 2-methyldotriacontane(C33H68) azulene triacetine formic acid 4-methyl-pentanenitrile 2,4-hexadienal unsaturated hydrocarbon-isom.1 aliphatic hydrocarbon (C33H68) dimethylindene-isom.2 methylpalmitate propionic acid 2-ethylideneamino-propionnitrile furfural 1-hexadecene unsaturated hydrocarbon-isom.13 ethylindene glycerin-diacetate 2-methyl-propionic acid benzonitrile 1-(2-furanyl)-ethanone sesquiterpene(unsaturated KW) beta-caryophyllene naphthalene ethylpalmitate butyric acid tolunitrile-isom.1 benzaldehyde heptadecane alpha-caryophyllene trimethylindene methylstearate acrylic acid tolunitrile-isom.2 methylimidazole-carboxaldehyde farnesene arom. Hydrocarbons 2-methylnaphthalene 6-acetoxy-2,7,11-cembratriene 2-/3-methylbutyric acid nicotinonitrile 5-methyl-2-furfural unsaturated hydrocarbon-isom.2 toluene 1-methylnaphthalene methyl-pyroglutamate isocrotonic acid benzylnitrile tolualdehyde delta-cadinene ethylbenzene dimethylnaphthalene benzyl benzoate crotonic acid amides cinnamaldehyde 3,7,11,15-tetramethyl-2-hexadecene xylene-isom.1 biphenyl muskalactone 3-methyl-valeric acid acetamide p-ethylbenzaldehyde neophytadiene xylene-isom.2 methyl-biphenyl methyl-salicylate capronic acid (C6:0) propionamide acetaldehyde unsaturated hydrocarbon-isom.3 isopropylbenzene trimethylnaphthalene 1,1-dimethyl-2-phenylethylbutyrate 2-methyl-crotonic acid isovaleramide-isom.1 trimethoxy-acetophenone(Ga 1,4-eicosadiene xylene-isom.3 fluorene triethylcitrate sorbic acid acrylamide Next Generation Fingerprinting: Practical Considerations

• Smoke constituents (SC) to monitor: – Unique identified compounds – List of chemicals & chemical compounds identified by FDA* as Harmful & Potential Harmful Constituents (HPHCs) in tobacco products & tobacco smoke (n=72 out of 93) – Reported compounds in tobacco plant and tobacco smoke – Flavor compounds

• GC (7890A) – DB-624: HS-SPME-GC-MS Volatile chemicals – DB-5-MS: GC-MS Non-polar and Polar chemicals (TMS) – DB-FFAP: HT-GC-MS

• MS detection (7200 Agilent Q-TOF) – Electron ionization mode – Full scan - Low resolution  High Resolution, Mass Accuracy Semi-quantification (several ISs)

*FDA: Federal Register / Vol. 77, No. 64 / Tuesday, April 3, 2012 / Notices, 20034-20037

Page: 6 The Impact of use of High Resolution MS Instrument

benzoic acid TMS; C9 H11 O2 Si; 9.15145: + FBF Spectrum 4 +EI EIC(179.0523) with ± 0.09amu 130222-3R4F001.D % x10 179.0515 SNR (9.15 min) =38.7 [C9H11O2Si]+ 100 500ppm 2 S/N=39 4.6 ppm 90 Mass accuracy 80 0 Isotopic pattern 4 +EI EIC(179.0523) with ± 0.04amu 130222-3R4F001.D 70 x10 SNR (9.15 min) =83.2 ion abundance 200ppm 60 S/N=83 ion spacing 2 50

40 2.9 ppm 0 4 +EI EIC(179.0523) with ± 0.0035amu 130222-3R4F001.D 30 x10 180.0539 SNR (9.15 min) =244.2 [C9H11O2Si]+ 15.3 ppm 20ppm 20 2 S/N=244 181.0486 10 [C9H11O2Si]+

0 0 179 179.5 180 180.5 181 181.5 182 8.6 8.8 9 9.2 9.4 9.6 9.8 Counts (%) vs. Mass-to-Charge (m/z) Counts vs. Acquisition Time (min) Sensitivity  Selectivity  Confidence level 

Page: 7 Analysis of Reference Compounds

Reference standards ordered (n~600 + 63 ISs) – Chemical registration within UCSD  unique PMI code – GC-EI-HR-MS (DB-5) with silylation (BSTFA) n=360 standards – GC-EI-HR-MS (DB-5) without silylation n=182 standards – HS-GC-EI-HR-MS (DB-624) n= 60 standards – Assignment of EI HR mass spectrum – reference chemical – Uploading of HR EI mass spectra in UCSD (unique SpecID) – Building MS library specific to experimental conditions • Absolute Retention Time • Retention Index • EI mass spectrum • Mol file,…

Page: 8 Building Statistical Prediction Model of Retention Time from Reference compounds on a DB-5-MS GC column

QSPR (structure-based) modeling was used to predict RT of compounds (Dragon)

35 Without derivatization

• Training set: 30 n=116 reference compounds y = 0.9813x + 0.0864 R² = 0.9765

10 chemical descriptors 25 r2 = 0.986

20 predicted • Test set: predicted

n=56 reference compounds 15 2 r = 0.979 (min) RT

RT (min) RT (min) 10 • Total external Validation set: n=346 reference compounds 5 r2= 0.9765 0 Median dev. = -8 sec 0 5 10 15 20 25 30 35 RT (min) Min = -3.3 min RT (min) experimental Max = +3.2 min experimental

Very good prediction model was obtained from a set of broadly diverse reference compounds. Same model was used to predict over 900 components.

Page: 9 Clustering approaches to select appropriate internal standards

• Fingerprinting method should be high throughput  No calibration curve  Concentration estimated from an adequate IS (spiking addition)

• Clustering approach was used to select appropriate internal standards (Accelrys Pipeline Pilot 8.5)

• Adequate selection of IS is based on:  Similar physicochemical properties (i.e., logP, number of rotational bonds, polar surface area, volatility…)

• Smoke constituents were clustered into chemical families so that  At least one isotopically labeled IS (cluster center) was purchased for each cluster

Page: 10 Example: Cluster of Long Chain Fatty Acids

13 Stearic acid- C18

Selection of isotopically labeled internal standard was made from Central Euclidian distance value & commercial availability

Page: 11 Clustering approach to improve semi-quantification

• Selection of 1 or 2 internal standards (IS) per cluster : 63 IS were purchased (from which 19 IS are present in the FDA list) • Determination of semi-quantitative compound concentrations in relation to their respective internal standard • Compensation for compound degradation issues (stability, partial silylation,…) was best by selecting adequate IS from the same cluster

Page: 12 GC-EI-HR-MS of Internal Standards (n=63) Naphtalene-d Toluene-d 8 Benzene-d6 8 KI=1175 DB-624 Butanone-d 5

Furan-d 4 Quinoline-d KI=492 7 Benzonitrile-d5 KI=945 n=21

Quinoline-d7 DB-5-MS

Benzonitrile-d 5 Naphtalene-d KI=945 8 Benzo[a]pyrene-d2 KI=1175 KI=2800

n=19

Quinoline-d7 DB-5-MS TMS

Naphtalene-d8 Benzonitrile-d5 KI=1175 KI=945

n=24

Page: 13 Good coverage of retention times Outline

1. Description of the existing GC-MS fingerprinting method  Pros & cons of the current fingerprinting method  Why switching to a 7200 Agilent high resolution MS instrument?

4. What about other smoke constituents?

5. Conclusion

Page: 14 Quantitative Aspect: Importance of Qualifier ions

Phenylpropionic acid-TMS

-5.3ppm

C H Si C2H7OSi 8 8

O O -3.5ppm

C11H15O2Si C7H7 C3H9Si -4.1ppm -1.5ppm C H O Si C11H13OSi 12 18 2

Picolinamide-TMS

C2H7OSi

NH Si N -6.3ppm -10.5ppm O C8H11NSi C6H4N2 C8H11N2OSi

Ions highlighted in yellow were extracted for the semi-quantification

Page: 15 Case Study: Cresol Isomers o-Cresol

Mass Accuracy o-cresol Mass Accuracy o-cresol-d8 RT shift Library Match Score 0.11% Average: 78%-5.2ppm -4.9ppm -2.8ppm -4.8ppm -4.4ppm +0.6ppm

R2=0.9936 Si O IS

IS IS IS IS

D D D

Si O-cresol D O IS D D D

Page: 16 Case Study: Cresol Isomers m-Cresol

Mass Accuracy RT shift Library Match Score m-Cresol 0.11% Average: 70% -4.8ppm -4.6ppm -8.2ppm o-Cresol-d8

R2=0.9892 Si O

D D D

Si D O

D D D

Very good chromatography separation Reliable automatic peak integration Page: 17 Case Study: Cresol Isomers p-Cresol Mass Accuracy RT shift Library Match Score p-cresol 0.11% Average: 57%-5.1ppm -4.6ppm - 3.2ppm o-cresol-d8

m/z 165.0730 Quant. R2=0.9905 Si O

D D D

Si D O

D D D

Very good chromatography separation Reliable automatic peak integration using MassHunter software

Page: 18 Key Consideration for Semi-Quantification

• Need to have a good chromatographic separation to resolve: – Isomers issue – Peak deconvolution • Selection of characteristic for Quantifier & Qualifier ions • Robustness due to usage of both Retention Times & KI • Assessment of linearity: – Using a 5-points calibration curves [0.29 – 4.8 ng/µL] – Analysis with and without silylation – Average of R2=0.9694 (n=71 compounds) – R2 < 0.9 only for 3 standards (due to sensitivity) After silylation

Page: 19 Silylation Efficiency – example of o-cresol-d8

6.13 Before derivatization o-Cresol-d8

After derivatization

o-Cresol-d8 No trace of underivatized o-Cresol

7.359

o-Cresol-d8-TMS

Less fragmentation occurred after TMS derivatization Higher sensitivity

Page: 20 Sample of reference cigarette (3R4F TPM) GC-EI-HR-MS after TMS derivatization

DB-5-MS after silylation

Stigmasterol

The csv file is used for an automatic and systematic search of smoke constituents Fast data processing

Page: 21 Sample Stability After Chemical Derivatization (Silylation)

8 x10 +EI TIC Scan 130124 3R4F derivatized DB5-001.D 2.4 10µL 3R4F TPM extract (ethanolic) 2.2 2 evaporated under nitrogen 1.8 1.6 3R4F Add 100 µL of IS mix + RI-markers 1.4 1.2 Add 800 µL CH2Cl2 / Acetone 80:20 v/v 1 Add 50 µL of BSTFA and 50 µL of pyridine 0.8 0.6

Incubate @ 80°C for different time period 0.4

0.2

0 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Counts vs. Acquisition Time (min)

Samples stability was monitored over a period of approximately 15 hours. The analyzed smoke constituents do not show any significant differences.

After derivatization, aerosol components are stable over time Appropriate selection of IS needed

Page: 22 What about other Smoke Constituents?

Updating the Fingerprinting Method with New Compounds

Page: 23 Extraction of Silylated compounds can be of great help to identify additional smoke constituents

EIC of m/z 73.0468

EIC of m/z 75.0261

Si OH

Mass accuracy AND predicted RT are of great interest to postulate new compounds

Page: 24 Analytical Workflow for Unknowns: Advantage of Full Scan / High Resolution analysis Chromatogram

Integrate

2 Deconvolute 1 Extract

Spectra Target list 3 Library Match Spectral Library Target Match

Rtprediction

Match Hit Non-Target (NIST, MA, IS. Pat., Hit RT prediction)

Further 4 investigations

Page: 25 Conclusion

 A GC-EI-HR-MS method has been established to semi-quantify smoke constituents in aerosol and aerosol fractions.  The method has been simplified with:  Two GC columns instead of three,  Removal of Solid Phase MicroExtraction (SPME)  We have demonstrated the great advantage of high resolution _ mass accuracy:  Improved S/N  Specificity and selectivity  Enhanced sensitivity  Improved semi-quantification with:  Clustering approach to select adequate internal standards (n=63)  Compensate for compound degradation issues  Very good automatic peak integration (MassHunter software)  Very good linearity: averaged R2 = 0.9694 (n=71) • Monitoring of compounds can be expanded thanks to:  Full scan analysis at high mass accuracy (< 10 ppm)

 Very good prediction model of retention time (median RTshift = -8 sec, n=346)

Page: 26 Acknowledgements

Eric Dossin Philippe Guy Arno Knorr Martin Almstetter

Mark Bentley Pavel Pospisil Elyette Martin Aurelien Monge

Audrey Debrick Stefania Della Gatta Quentin Dutertre Daniel Arndt

Thank you for your attention. Page: 27