Heard D.E. (editor) Analytical Techniques for Atmospheric Measurement

Подождите немного. Документ загружается.

378 Analytical Techniques for Atmospheric Measurement

standards were prepared by NPL, for all laboratories, to reduce variability associated

with the standard, and analysis was carried out in 12–14 laboratories over four years.

Agreement was generally good for high concentration standard mixtures but disagreement

increased as lower concentration, real air samples were analysed. However, some improve-

ments were achieved over the course of the project. A number of inter-comparisons

between different analytical measurements including GC-FID/ECD/MD, proton transfer

reaction mass spectrometry (PTR-MS) and open path FTIR (OP-FTIR) have been made

(Christian et al., 2004). Good agreement between GC and OP-FTIR and between GC and

PTR-MS was achieved for a number of compounds. However, differences were found

for sticky compounds (e.g. formic acid) and compounds with a low proton affinity

(e.g. formaldehyde) with OP-FTIR and PTR-MS having considerable advantages. It is

important for real samples to take into account the great difference in time resolution

between these methods. In the future, validation of techniques via inter-comparisons of

standards will increase.

Typical single column GC-FID instruments may have relative standard deviations of

the order of 0.2–0.8% for this type of measurement, with an overall level of uncertainty

at the single-figure percentage level arising mainly from gravimetric errors associated

with the preparation of standards. The frequency of calibration is dependent on appli-

cation and detector. An FID is considerably more stable over time than a quadrupole

mass spectrometer, which may require calibration every couple of samples. During an

intensive field campaign, GC-FID systems should be calibrated at least every 100 samples.

Many of the detectors used in GC analysis can give structure-related responses, allowing

a reduced number of analytes to be calibrated. This can reduce calibration time consid-

erably. For example, FIDs give a near-linear increase in response as the carbon number

increases in hydrocarbon chains. The addition of hetero-atoms can cause a deviation

from linearity which can be compensated for using the Effective Carbon Number (ECN).

The ECN relates to an FID response factor as a function of molecular structure, with

different hetero-groups causing varying deviations from linearity. The use of ECN can

be extremely important for the quantification of oxygenated materials which are usually

unstable in gas cylinders, with rapid wall losses and degradation. Carrier gas impurities

can lead to a reduction in separation efficiency with increased background signals and

damage to detectors and columns, in particular if oxygen is present. It is important to

use ultra-high purity gas, and 99.999% helium, or greater, is typically used with both

hydrocarbon and oxygen scrubbers.

8.2.6 Instrument deployment in the ﬁeld

An important aspect of atmospheric chemistry is the ability to deploy instruments outside

of the laboratory. In order to make measurements of real air samples, GCs must be easily

transported and rapidly installed. The location and accessibility of field sites can vary

greatly and the problems associated will be site dependant.

The availability of gases, cryogens and power is one of the main criteria of fieldwork.

Gases are an integral part of the GC system and gas cylinders are available in a variety

of sizes, usually between 2 and 50 L. In some cases, for example working on an aircraft,

Chromatographic Methods 379

weight is an issue and must be kept to a minimum. Cylinders can be made from

aluminium making them lighter than conventional steel cylinders. The use of commer-

cially available gas generators can reduce the number of gas cylinders required. This is

especially important in locations where gas delivery is not possible. The availability of

cryogens, such as liquid nitrogen, in remote locations makes their use limited and systems

such as Peltier-cooled PTV are more suitable. The power usage of the entire GC set-up

should be kept to a minimum and it is common to use a uninterruptible power supply

(UPS) to avoid damaging shutdown of the instrument during power disruptions. Field

instrumentation incorporating a PTV trap, sample pump, flow control and GC-FID has

a typical power usage of 2 kW (Whalley et al., 2004).

In atmospheric chemistry, the vertical profile of the atmosphere is important. GC

systems have been deployed in a number of different ways to obtain information about

the composition of the atmosphere at varying altitudes. Ground level and boundary layer

measurements can be taken from a number of platforms including conventional and

mobile laboratories, research ships and balloons. A lightweight balloon-borne instrument

has been developed to measure trace gases in the stratosphere and incorporates a series of

Carboxen absorbent tubes which are sampled at different altitudes and analysed on the

ground by a GC system (Danis et al., 2000). Long, wide bore sampling manifolds can be

used to obtain air above ground level allowing the GC to remain in a laboratory on the

ground. For example, the Cape Grim Baseline Air Pollution Station in Tasmania, Australia

(http://www.bom.gov.au/inside/cgbaps/index.shtml) has a very tall tower allowing air to

be sampled at elevations of 3, 10 and 70 m above ground level. To obtain free tropospheric

air without the use of an aircraft, requires a laboratory situated at a high altitude above

sea level. The Jungfraujoch High Altitude Research Station (http://www.ifjungo.ch/) is

situated on a mountain saddle in Switzerland at 3580 m above sea level. Evidence shows

that the Jungfraujoch observatory lies in the free troposphere throughout winter and often

in spring and autumn. A number of atmospheric measurements based on GC systems

have been made at the Jungfraujoch, including hydrocarbons, PAN and halocarbons

(Whalley et al., 2004).

The use of aircraft to obtain a vertical profiles of the atmosphere is becoming increas-

ingly popular. Aircraft allow a large spatial coverage for measurements of atmospheric

species. Flight time is limited, however, and measurements therefore only represent snap

shots in time. Measurements can be made both in situ with an onboard GC and post

analysis with whole air sampling and a ground-based GC system (Blake et al., 1996; Purvis

et al., 2003). Examples of in-flight GC systems are ORAC (Organics by Real Time Airborne

Chromatograph) (Whalley et al., 2004) and PANAK (PAN-Aldehydes-Alcohols-Ketones)

(Singh et al., 2004) which use a series of capillary columns and multiple detectors such

as ECD, PID and RGD.

Gas Chromatography is also used extensively in industry to measure stack emissions,

landfill emissions and to monitor ambient workplace air quality. In general, grab

samples are taken at the emission site using Teflon bags or evacuated containers,

which can be analysed in the laboratory. A development in emissions monitoring

are portable GCs that can be carried by hand or in a backpack. Portable GCs are

available with a variety of detectors including PID, ECD, TCD and MS. These instru-

ments have many applications including hazardous waste site investigation, indoor air

quality and emergency response, for example in the detection of chemical weapons

(http://www.inficonvocmonitoring.com/).

380 Analytical Techniques for Atmospheric Measurement

8.2.7 Examples of applications of chromatography in

atmospheric analysis

Table 8.2 shows a number of typical applications of chromatographic techniques in

atmospheric analysis. Typical sampling lengths, separation mechanisms, detectors and

sensitivity are given, as well as the temporal resolution, indicating the frequency with

which measurements can be taken. The following section highlights a number of examples

of atmospheric measurements obtained using GC.

Figure 8.3 shows a simple GC chromatogram using a WCOT column and an ECD

detector to measure PAN at the Jungfraujoch research station in Switzerland (Whalley

et al., 2004). Using a backflush facility through a pre-column, well-retained analytes were

diverted, allowing only analytes of interest to enter the analytical column (30 m long ×

053 mm i.d., 5% phenyl polysilphenylene-siloxane, 15 m film thickness). Subsequent

separation results in the isolation of PAN from the large air/water peak every 90

seconds. Calibration of the system was performed using gaseous PAN generated by the

photolysis of a mixture of NO

with excess acetone, diluted using zero air (Zellweger

et al., 2000).

The measurement of hydrocarbons requires a greater separation capacity than the PAN

GC, where only one analyte was targeted. Figure 8.4 shows the separation of C

–C

hydrocarbons obtained using a PTV injector, a long PLOT column (50 m long ×053 mm

i.d., Al

modified with Na

,10m film thickness) and an FID (Bartle & Lewis,

1996). Automation of the procedure, including the use of a H

gas generator to provide

carrier and flame gas, allowed the unmanned instrument to produce a chromatogram

Time: 0.080 minutes

0.08

0.06

0.04

Volts

0.02

0.00

0.5 1.0 2.0 3.01.5

Minutes

Peroxyacetylnitrate

2.5

270203.005, Chan A

Figure 8.3 PAN chromatogram, with peroxyacetylnitrate highlighted, taken at the Jungfraujoch research

station, Switzerland. (Fom Whalley et al., 2004, reproduced by permission of the Royal Society of

Chemistry.)

Table 8.2 Typical applications of chromatographic instruments used for gas and condensed phase atmospheric measurements

Description Separation Sample size Detectors Sensitivity RDS

Typical

resolution

Operation

Permanent

gases, CO,

GC packed

columns/molecular

sieves

1–10 mL TCD, FID,

HID

Sub-ppm (TCD) to sub-ppb (HID) Separation 1–10 min UGS

PAN GC open or packed

columns. With pre

column/back ﬂush.

1–10 mL ECD 10 ppt Separation 3 min UGS, AC

Light

hydrocarbons

-C

GC – open tubular

PLOT type

100–3000 mL FID 1–5 ppt Separation 1 hr UGS, AC

DMS GC – WCOT

typically

methylpolysiloxane

100–3000 mL FID, FPD 1–10 ppt Separation 30 min UGS

Isoprene/

terpenes

GC – WCOT

siloxane or

cyclodextrin based

if chiral

100–3000 mL FID, MS 1–10 ppt with FID 10–100 ppt

full-scan quadrupole MS. Sub-ppt

with quad MS in single ion

monitoring mode. 1–10 ppt with

TOF MS.

Separation 30 min to

1hr

UGS

CFCs GC – WCOT

typically

methylpolysiloxane

100–3000 mL ECD, MS 1–10 ppt with ECD 10–100 ppt

full-scan quadrupole MS. Sub-ppt

with quad MS in single ion

monitoring mode. 1–10 ppt with

TOF MS.

Separation 1 hr UGS

Reactive

halocarbons,

organic

nitrates

GC – WCOT

typically

methylpolysiloxane

100–3000 mL ECD, MS,

Mag sector

MS, CI-MS

1–10 ppt with ECD, sub-ppt with

quad MS in single ion monitoring

mode, sub-0.01 ppt with chemical

ionisiation, sub 0.01 ppt with

magnetic sector MS.

Separation 1 hr UGS, lab

Table 8.2 (Continued)

Description Separation Sample size Detectors Sensitivity RDS

Typical

Resolution

Operation

Aromatics GC – WCOT

typically phenyl-

methylpolysiloxane

100–3000 mL PID, MS,

FID

1–10 ppt with PID, 1–10 ppt

with FID 10–100 ppt full-scan

quadrupole MS. Sub-ppt with quad

MS in single ion monitoring mode.

1–10 ppt with TOF MS.

Separation 30 min to

1hr

UGS, AC

Carbonyls

(direct)

GC – WCOT or

PLOT

100–3000 mL FID, MS 20–100 ppt with FID, 20–200 ppt

full-scan quadrupole MS. 1–10 ppt

with quad MS in single-ion

monitoring mode. 1–10 ppt with

TOF MS.

Separation 1 hr UGS

Carbonyls

(derivatisation)

GC – WCOT

typically

methylpolysiloxane

5–100 L ECD, MS 20–200 ppt with ECD, 20–200 ppt

full-scan quadrupole MS. 1–10 ppt

with quad MS in single-ion

monitoring mode. 1–10 ppt with

TOF MS.

Sample

collection

Several

hours

AGS, lab

PAH LC – normal or

reverse phase

1–1000 m

UV, MS Sub-0.01 ppt with large sample size Sample

collection

Hours to

days

Lab

Acids LC – ion exchange

column

1–1000 m

MS ppt level Sample

collection

Hours to

days

AGS, lab

Peroxides LC – reverse phase

with derivatisation

10–100 L FL 10–100 ppt depending on sample

size

Sample

collection

15 min to

1hr

UGS, lab

RDS – rate determining step in analysis.

Operation: UGS – Unattended ground site, AGS – Attended ground site, AC – Aircraft with attendant ground-laboratory for off-site analysis of Whole Air Sample canisters.

Chromatographic Methods 383

Figure 8.4 Low molecular weight hydrocarbons including methane determined using an on-line

activated charcoal adsorbent trap in a programmed temperature vaporisation injector. Column 50 m

0.53 mm i.d. AL

/Na

PLOT (Chrompack, Netherlands) 10 m d.f. Desorption temperature was at



−1

from −20 to 400



C, and column temperature programmed from 45 to 200



C. Peaks: 1 =methane,

2 = ethane, 3 = ethene, 4 = propane, 5 = propene, 6 =2-methylpropane, 7 =ethyne, 8 = n-butane, 9 =

trans-2-butene, 10 = 1-butene, 11 =isobutene, 12 =cis-2-butene, 13 =2-methylbutane, 14 =n-pentane,

15 = 13-butadiene, 16 = pentenes, 17 = 2-methylpentane, 18 = 3-methylpentance, 19 = n-hexane,

20 =methyl hexanes and hexenes, 21 =heptane, 22 =methylcyclopentane, 23 =benzene, 24 =toluene.

(Reproduced with permission from Bartle and Lewis, 1996.)

every 45 minutes. Using the same instrumentation with a WCOT column (60 m long ×

053 mm i.d., 100% dimethyl polysiloxane, 3 m film thickness), which is more suited to

the separation of heavier weight analytes, allows the separation of C

–C

hydrocarbon

species in the atmosphere, as shown in Figure 8.5 (Lewis et al., 1996). Using a dual

column approach in one instrument facilitates the measurement of the whole range of

–C

hydrocarbons in a single analysis.

The analysis of oxidised organics in the atmosphere gives important information on the

age and processing of an air mass. Figure 8.6 shows a chromatogram of a C

–C

o-VOCs

384 Analytical Techniques for Atmospheric Measurement

0.000

Concentration 2.88 ppbV in air

10 15 20

Minutes

25 30 35 40

0.002

0.004

0.006

0.008

0.010

0.012

0.014

Volts

1. unresolved volatile material

2. hexane

19. styrene

20. o-xylene

21. nonane

22. iso propyl benzene

26. 1,3,5 trimethyl benzene

27. O-ethyl toluene

28. 1,2,4 trimethyl benzene

29. decane

31. indane

36. dodecane

35. naphthalene

34. 1,2,3,5 tetra methyl benzene

30. 1,2,3 trimethyl benzene

23. propyl benzene

24. m-ethyl toluene

25. p-ethyl toluene

3. methyl cyclopentane

4. 2,4 dimethyl pentane

5. benzene

6. 2 methyl hexane

7. 3 methyl hexane

8. 2,2,4 trimethyl hexane

9. heptane

10. methyl cyclohexane

11. 2,4 and 2,5 dimethyl hexane

12. 2,2,3 trimethyl pentane

13. toluene

14. 2 and 4 methyl heptane

15. 3 methyl heptane

16. octane

17. ethyl benzene

18. m and p xylene

32. 1,4 dimethyl 2 ethyl benzene

33. diemethyl ethyl benzenes

and undecane

Figure 8.5 Aromatic hydrocarbon species determined using an on-line Tenax™ TA adsorbent trap in a

programmed temperature vaporisation injector. Column 60 m 0.53 mm i.d. 100% methyl polysiloxane

3 m d.f. (Restek RTX-1) Desorption temperature was at 16



−1

from 0 to 220



C, and column temperature

programmed from 35 to 240



C. (Reproduced with permission from Lewis et al., 1996.)

analysis obtained at the Mace Head Atmospheric Research Station (MHARS), situated

on the West coast of Ireland (Hopkins et al., 2003). The o-VOCs were collected onto

a cooled multi-bed absorbent trap and analysed using a mixed phase capillary column

(10 m ×053mm i.d. Varian CP-LOWOX, 10 m film thickness), designed specifically

for trace oxygenate analysis with a highly polar stationary phase. Analytes were calibrated

using a combination of liquid permeation tubes and gas standards.

Chlorofluorocarbons (CFCs) have been implicated in the destruction of stratospheric

ozone, particularly in the polar regions, leading to the well-known ‘hole in the ozone

Chromatographic Methods 385

Figure 8.6 Sample chromatograms produced from a LOWOX column where 1 = Aliphatic NMHCs,

2 = benzene, 3 = toluene, 4 = acetaldehyde, 5 = methanol, 6 = acetone, 7 = ethanol, 8 = propanol.

(From Hopkins et al., 2003, reproduced by permission of The Royal Society of Chemistry.)

layer’ discovered in Antarctica. Figure 8.7 shows a chromatogram of CFCs and Halon

species found in clean marine air obtained using both a conventional ECD and an

oxygen-doped ECD. The oxygen-doped ECD can enhance the response of certain halogen

species (e.g. CH

Cl, CHF

Cl, CH

) which have poor responses using conventional

ECD detectors. The discovery of the destructive effects of CFCs has led to international

protocols limiting their use. A number of CFC replacements have been introduced and

their background concentrations have been rising steadily. A GC-magnetic sector MS

instrument has been used to measure the concentration of CFCs and Halon species,

between 1978 and 1995, at the Cape Grim Observatory in Tasmania. Samples were pre-

concentrated onto a cooled glass bead/Tenax™ trap and separated on a PLOT column

(50 m×053 mm i.d. HP Alumina PLOT (KCL)). An example chromatogram is shown in

Figure 8.8, with peaks detected in Single Ion Mode (SIM) at m/z 69, with concentrations

ranging from 80 to 0.1 ppt. Analyte identifications are given and the species detected are

long-lived greenhouse gases, with Halon-1301 being an important ozone depleter.

In marine environments, elevated levels of volatile halocarbons are found in the

atmosphere, which are emitted from marine sources such as seaweed and phytoplankton.

Many of the most important halocarbons have very unique mass spectra, making their

analysis particularly suited to GC-MS. Using the mass spectrometer in SIM, over specific

retention ranges, gives maximum sensitivity. The GC-MS chromatogram shown in

Figure 8.9 was obtained at the Mace Head station, which is situated in a coastal region.

The low concentration species have been extracted to show the total speciation achieved.

Data of this type can give important information about the air–surface partitioning

occurring within the marine boundary layer.

As described earlier, the urban atmosphere contains a highly complex mixture of

organic compounds, in both the gas and the aerosol phase. GC ×GC with FID detection

has been used to separate the C

–C

organic compounds at a roadside site in Leeds, UK

(Hamilton & Lewis, 2003). Using a primary volatility-based column (50 m ×032 mm i.d.

100% dimethylpolysiloxane, 4 m film thickness), followed by a secondary polarity-based

separation (25m×010 mm i.d. 50% phenyl polysilphenylene-siloxane, 01 m film

thickness) and a valve modulator, gives significant improvement in separation capacity.

386 Analytical Techniques for Atmospheric Measurement

Figure 8.7 CFCs and Halon species in clean marine air determined using an on-line Carbosieve micro

adsorbent trap and direct injection to capillary GC. Detection by dual ECD/oxygen doped ECD. 60 m,

0.33 mm i.d., 1 m ﬁlm DB-1 (J & W). (By courtesy of M. Bassford and P.G. Simmonds, University of

Bristol, Cantocks Place, Bristol.)

Figure 8.8 Long-lived greenhouse gases observed at the Cape Grim Observatory, Tasmania using a GC/sector MS with SIM at m/z 69. Samples were pre-concentrated

onto Tenax™/glass bead traps and separated using an Al

/KCl PLOT column. (By courtesy of D. Oram, University of East Anglia, UK.)