a reanalysis of the capacity of ptr ms to€¦ · nes methanol formic ... ulp 362, ulp 363, e10...

The Centre for Australian Weather and Climate Research A partnership between CSIRO and the Bureau of Meteorology

A reanalysis of the capacity of PTRMS to uniquely identify and quantify VOCs in global background air

Ian Galbally, Erin Dunne, Sarah Lawson, Antonio Patti with collaborative data from Dennys Angove GAW Reactive Gases Workshop Helsinki June 2010

www.cawcr.gov.au


• Measuring VOCs using PTRMS • Introduction to PTRMS as a measurement technique • Review of suitability of PTRMS for measuring VOCs in background air

• Results from Cape Grim • VOC concentrations from two recent campaigns and comparison with other remote oceanic sites

Outline

ProtonTransferReaction Mass Spectrometer (PTRMS)

Proton Transfer: Proton Transfer: H H3 3O O + + + A + A → → AH AH + + + H + H2 2O O PA (A) > PA (H PA (A) > PA (H2 2O) O) Electric field ~50 V cm Electric field ~50 V cm 1 1 Pressure ~ 2 mbar Pressure ~ 2 mbar

Detection Limit: ~10 pptv Detection Limit: ~10 pptv Response Time: 100 ms Response Time: 100 ms

Compound identification: Compound identification: Capable of detecting 100s of Capable of detecting 100s of VOCs (mass range 1 VOCs (mass range 1 512amu) 512amu) Soft ionization reduces product Soft ionization reduces product ion fragmentation ion fragmentation

Compound quantification: Compound quantification: The product ion signal is The product ion signal is proportional to the mixing ratio of proportional to the mixing ratio of that compound in the air flow. that compound in the air flow.

PTRMS Identification and quantification of atmospheric volatile organic compounds

Current hypothesis: The PTRMS can be used to accurately identify and quantify VOCs in the atmosphere.

This hypothesis can be examined in 3 parts:

1. Can the PTRMS uniquely identify atmospheric VOCs?

2. Can the PTRMS accurately quantify the concentration of atmospheric VOCs?

3. What can be learnt of atmospheric VOCs from a substantial body of real atmospheric measurements?

PTRMS Limitations

PTR PTR MS Mass resolution: 1 amu MS Mass resolution: 1 amu PTR PTR MS cannot distinguish between product ions MS cannot distinguish between product ions with the same mass! with the same mass!

Identification and quantification require information on what Identification and quantification require information on what compounds are contributing to the ion signal at each mass channe compounds are contributing to the ion signal at each mass channel l

A signal at a given m/z may contain contributions from:

2 or more compounds with the same protonated mass Fragment or cluster ions Product ions from reactions with O2 + NO + and H3O + (H2O)

Laboratory fragmentation studies currently underway of VOCs by PTRMS

Alcohols Acids Esters Other OVOCs Alkanes/Alke nes

Methanol Formic acid Methyl formate Formaldehyde Toluene

Ethanol Acetic acid Ethyl formate Acetaldehyde 2methyl 1,3 butadiene

1propanol Propanoic acid Propyl formate Acetone

2propanol Oxalic acid Butyl formate 1hydroxy propanone

Organo nitriles

1butanol Methyl acetate 2butanone Acetonitrile

2butanol Ethyl acetate Methyl

isobutyl ketone

Tert. butyl alcohol

2oxo propanal*

Ethanedial*

Hydroxy acetic Acid*

Hydroxy acetaldehyde*

* to be completed

Case Study: Mass 75 Significant ion signal at mass 75 detected in semirural air near Sydney. 20 compounds with PA>H2O that would have a protonated mass of 75.

Significant correlations (r > 0.5) were observed between mass 75 and the following masses: 33, 43, 45 , 46, 47, 61, 71

The correlations and concentrations observed for m/z 75 and 43 indicate methyl acetate is prime candidate compound for the a majority of the signal observed at m/z 75

Mass spectra studies improve Mass spectra studies improve our ability to identify our ability to identify compounds in the atmosphere compounds in the atmosphere using PTR using PTR MS MS

E = V cm E = V cm 1 1

N = molec. N = molec. cm cm 3 3

Fragmentation studies of ethanol

Mass spectra studies can be Mass spectra studies can be used to optimize the used to optimize the instrument response to specific instrument response to specific compounds compounds

All analyses are on the same humidified air sample containing ethanol

PTRMS response to ethanol is optimal at E/N = 90Td

Low response at high E/N is likely due to fragmentation to H3O + unmeasurable!

E = V cm E = V cm 1 1

N = molec. cm N = molec. cm 3 3

CET Indoor Smog Chamber – D. Angove

• Teflon lined, airfilled chamber of 18 m 3 supported by an aluminium frame and lined with highly reflective aluminium

• Provides a low contaminant environment which permits studies at low reactant concentrations.

• UVA lights are used to simulate sunlight.

• Interfaced instruments allow measurement of reactant consumption and product formation.

• Facilities are available for the sampling of gaseous and aerosol samples.

• Observations are supported by complex chemical mechanism models such as the Master Chemical Mechanism.

CET Indoor Smog Chamber – External View

Formic acid during smog chamber runs with standard and ethanol containing fuels

0 100 200 300

2

0

2

4

6

8

10 ULP 362 ULP 363 E10 360 E10 361 E5 365 E5 366

Form

ic Acid Conc. (p

pb)

Time (min)

Formic acid measured using longpath FTIR (119 m) and spectrum fitting using the formic acid lines generated from the HITRAN database and MALT version 5.5 developed by David Griffiths (UOW). The errors are outputs from the fitting procedure.

Mass 47 (expressed in units of ethanol) during smog chamber runs with standard and ethanol containing fuels.

100 0 100 0

20

40

60

80

100

120

ULP 362, ULP 363, E10 360, E10 361, E5 365, E5 366

Ethanol equiv. conc (ppb)

Time (min)

Water Injection

Fuel Injections

LIGHTS ON

Mass 47 increases in all experiments, in spite of the expected decrease in ethanol concentration – the influence of formic acid

There is indication of baseline offsets observed in all experiments with water injection. Is there either a water signal, or chamber wall desorption, at mass 47?

Aim: to determine whether PTRMS can unequivocally identify and quantify isoprene, terpenes, acetonitrile, methanol, ethanol, acetone, dimethyl sulphide, benzene and toluene in clean air.

Use mass spectral data to uniquely identify the PTRMS product ion signals of each of these VOCs in measurements from Cape Grim Tas. and rural site in WA.

e.g. acetonitrile measurements at Cape Grim. The contribution of propene must be removed from the signal at mass 42 for accurate quantification of acetonitrile.

H3O + + CH3CN → CH3CNH + + H2O m/z 42 O2 + + C3H6 → C3H6 + + O2 m/z 42

Ambient monitoring in the global atmosphere


The suitability of PTRMS to measure VOCs

VOC Protonated Mass

OK in background

air?

Issues at this mass Intercomparison studies

methanol 33 Y FTIR (Christian et al., 2004) (Karl et al., 2007)GCMS (de Gouw et al., 2003)

acetonitrile 42 Y O 2 + + propene but not significant when propene <100 ppt

GCMS (de Gouw et al., 2003)

ethanol 47 Challenging Formic acid Low sensitivity to ethanol due to fragmentation to H 3 O + (Inomata

Tanimoto, 2009b) Di methyl ether?

Saphir (Apel et al., 2008)

acetone 59 Y Propanal contributes (~10%) possibly glyoxal

GCMS (de Gouw et al., 2003) (Karl et al., 2007)

Saphir (Apel et al., 2008)

di methyl sulphide (DMS)

63 Y Hydrated acetaldehyde may contribute where [acetaldehyde] is

>> [DMS]

GCMS (de Gouw Warneke, 2007)

isoprene 69 Y (providing no biomass burning)

Furan C 5 H 8 hydrocarbons

GCMS(de Gouw et al., 2003) (Karl et al., 2007)

terpenes 81, 137 Y GCMS (de Gouw et al., 2003)


Baseline Conditions at Cape Grim


VOC concentrations (ppt) at Cape Grim and other Oceanic sites

1. Galbally, I.E., et al., Environmental Chemistry, 2007. 4(3): p. 178182. 2. Ayres, G.P. and R.W. Gillett, Journal of Sea Research, 2000. 43: p. 275286. 3.Colomb, A., et al., Environmental Chemistry, 2009. 6(1): p. 7082. 4. Williams, J., et al., Environmental Chemistry, 2010 7(2): p. 171182. 5. Warneke, C. and J.A. de Gouw, Atmospheric Environment, 2001. 35(34): p. 59235933. 6 .Singh, H.B., et al., Journal of Gephysical Research, 2004. 109. 7. Williams, J., et al., Geophysical Research Letters, 2004. 31(23): p. 5.

Protonated Mass

Most Probable Compound

Lifetime in marine

boundary layer (days)

Cape Grim 2007

Cape Grim

2006 [1]

Cape Grim 19881993

[2]

Southern Hemisphere

Mid Latitude

Oceanic [3] [4]

Tropical Oceanic [5][6][7]

33 Methanol 15 633 476 546 575890

42 Acetonitrile 471 32 25 20 111142

45 Acetaldehyde 0.9 53 nd 290 204

59 Acetone 66 61 118 127 450 466530

63 DMS 1.5 95 ~80 110 77 5089

69 Isoprene 0.1 21 14 40 66

81 monoterpenes ~0.1 25 nd 10


DMS and isoprene diurnal cycles in clean air Dec 2007

0

20

40

60

80

100

120

140

160

180

0 4 8 12 16 20 24 Hour of day

Concen

tration DM

S (ppt)

0

10

20

30

40

50

60

70

80

90

100

Concen

tration isop

rene (p

pt)

DMS Isoprene

Radon data courtesy of Wlodek


DMS diurnal cycle – past and present

From: Ayers and Gillett, Journal of Sea Research 43 (2000) 275–286

10 days in December 2007 (PTRMS)

February 1996 (GCFPD)

0

20

40

60

80

100

120

140

160

180

0 4 8 12 16 20 24 Hour of day

Con

centratio

n DMS (ppt)


Halogen chemistry?

0

20

40

60

80

100

120

140

160

180

0 4 8 12 16 20 24 Hour of day

Con

centratio

n DMS (ppt)

From: Galbally et al, Geophysical Research Letters, 27 (2000), 38413844

DMS Dec 2007 O 3 DecJan 19851997


Thank you


Ian Galbally

Phone: +61 3 9239 4428 Email: [email protected] Web: www.cawcr.gov.au

Thank you www.cawcr.gov.au

a reanalysis of the capacity of ptr ms to€¦ · nes methanol formic ... ulp 362, ulp 363, e10...

Documents