Journal of Chromatography A, 1217 (2010) 2685–2694

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

Review

Sorbent-based sampling methods for volatile and semi-volatile organic

compounds in air. Part 2. Sorbent selection and other aspects of optimizing
air monitoring methods
Elizabeth Woolfenden ∗
Markes International Ltd., Gwaun Elai Campus, Llantrisant, RCT, CF72 8XL, UK

a r t i c l e i n f o a b s t r a c t

Article history: Sorbent tubes/traps are widely used in combination with gas chromatographic (GC) analytical methods
Available online 14 January 2010 to monitor the vapour-phase fraction of organic compounds in air. Applications range from atmospheric
research and ambient air monitoring (indoor and outdoor) to occupational hygiene (personal exposure
Keywords: assessment) and measuring chemical emission levels. Part 1 of this paper reviewed the main sorbent-
Air monitoring based air sampling strategies including active (pumped) tube monitoring, diffusive (passive) sampling
Sorbent tubes
onto sorbent tubes/cartridges plus sorbent trapping/focusing of whole air samples that are either col-
Sorbent traps
lected in containers (such as canisters or bags) or monitored online. Options for subsequent extraction and
VOCs
SVOCs
transfer to GC(MS) analysis were also summarised and the trend to thermal desorption (TD)-based meth-
Volatile organic chemicals ods and away from solvent extraction was explained. As a result of this trend, demand for TD-compatible
Semi-volatile organic chemicals sorbents (alternatives to traditional charcoal) is growing. Part 2 of this paper therefore continues with a
Thermal desorption summary of TD-compatible sorbents, their respective advantages and limitations and considerations for
sorbent selection. Other analytical considerations for optimizing sorbent-based air monitoring methods
are also discussed together with recent technical developments and sampling accessories which have
extended the application range of sorbent trapping technology generally.
© 2010 Elsevier B.V. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2686
2. Selection of TD-compatible sorbents – factors to consider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2686
2.1. Sorbent ‘strength’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2686
2.2. Inertness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2687
2.3. Hydrophobicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2687
2.4. Artefacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2687
2.5. Temperature stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2688
2.6. Mechanical strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2688
2.7. Mesh size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2688
3. Sorbent developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2688
4. Multi-sorbent tubes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2688
5. Other practical factors to consider for sorbent-based air sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2689
5.1. Trap/tube materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2689
5.2. Sorbent retaining materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2690
5.3. Preparation of sorbent tubes/traps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2690
5.4. Storage of conditioned and sampled sorbent tubes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2690
6. Water management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2690
7. Optimisation of desorption and other analytical parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2690
8. Uncertainty and analytical quality assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2691

∗ Tel.: +44 01443 230935; fax: +44 01443 231531.

E-mail address: ewoolfenden@markes.com.

0021-9673/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.chroma.2010.01.015
2686 E. Woolfenden / J. Chromatogr. A 1217 (2010) 2685–2694

9. Extending the application range for sorbent-based ‘air’ monitoring technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2692
9.1. Rapid assessment of chemical emissions from products and materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2692
9.2. Breath sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2693
9.3. In situ sampling of VOCs in soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2693
10. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2693
Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2693
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2694

1. Introduction must be sufficiently ‘strong’ to retain target analytes during sam-

pling/concentration, but weak enough to release them efficiently
For reasons explained in part 1 of this paper (sensitivity, during the thermal desorption phase. As described in part 1, sorbent
automation, repeatability, etc.), thermal desorption (TD) methods strength is usually measured in terms of retention or breakthrough
are gradually superseding solvent extraction procedures for anal- volumes. Standard air monitoring methods [2,3] are a good source
ysis of air samples collected on sorbent tubes/traps. Whole air of validated retention and breakthrough volume information for a
monitoring (online or using containers) is already almost exclu- wide range of common sorbent/sorbate combinations and describe
sively carried out using thermal desorption to extract the target how these values can be determined experimentally. Such stan-
organics from the sorbent focusing traps and transfer them to the dards may also list ‘Safe Sampling Volumes’ (SSVs) – derived either
GC(MS) analytical system. TD is also the method of choice for most by halving the chromatographically determined retention volume
ambient air monitoring and atmospheric research studies because or by reducing the experimentally determined breakthrough vol-
of the 1000-fold sensitivity enhancement it offers when compared ume by a factor of 2/3.
with solvent extraction. However, even higher level air monitoring Reported retention volumes are susceptible to temperature and
applications such as routine industrial hygiene or fugitive emis- are typically quoted at 20 ◦ C. As a (very) approximate rule, retention
sions testing, are beginning to transfer to TD methods. The driver volumes halve for every 10 ◦ C rise in temperature. The performance
in this case is not usually sensitivity but rather automation and the (retention characteristics) of strong sorbents such as carbonized
elimination of solvents such as CS2 which present both a health molecular sieves are adversely affected by high relative humidity
& safety hazard to operators and interfere with the subsequent (>80%) as recorded in standard methods. The retention volumes of
GC(MS) analysis. hydrophobic sorbents such as carbon blacks, Tenax® TA and other
It is important to point out that solvent extraction can be car- porous polymers are much less sensitive to atmospheric humidity
ried out using standard GC(MS) instrumentation and that it offers with negligible impact reported even up to 90% RH. The competitive
some advantages versus basic TD technology in that it allows repeat effect of other organic vapours is also reported as negligible at levels
analysis – e.g. for confirmation of results or repeat analysis under up to 100 ppm [36].
different conditions. However, suitable manual and automated Flow rate has been shown to have a negligible impact on sorbent
thermal desorbers are now available from multiple commercial strength (i.e., analyte retention volumes) provided minimum and
sources and are usually compatible with any make of GC(MS). The maximum rates are observed, e.g. 10–200 ml/min for std 6.4 mm
latest TD technology also allows repeat analysis, overcoming the (1/4-in.) O.D. stainless steel tubes with 5 mm I.D. (see part 1 of this
one-shot limitation of older systems (see below). Moreover, lower paper for more information).
running costs versus solvent extraction usually mean that the addi- It is important to select the appropriate sorbent or series of sor-
tional capital investment required for TD is recovered relatively bents for the target analytes in question. If the selected sorbent is
quickly. too weak (for example Tenax® TA for n-pentane or acetone) there
One consequence of the trend away from charcoal/CS2 and will be a temptation to use bigger tubes/traps and larger quantities
towards thermal desorption is the demand for TD-compatible alter- of sorbent. Oversized tubes or traps can significantly compromise
native sorbents–charcoal is too ‘strong’ and too active to allow air monitoring methods. In the case of whole air sampling (e.g.
reliable thermal desorption of all but the most volatile and stable online or canister/bag methods) the larger the cooled focusing trap,
organic compounds. A summary of the most common sorbents used the longer it takes to desorb resulting in slower, less efficient trans-
for thermal desorption and the factors to consider when select- fer to the analytical system. This leads to broader peaks and a
ing which sorbent to use for a particular application are reviewed consequent reduction in sensitivity and resolution. Using larger
below. masses of sorbent for pumped or diffusive air sampling tubes makes
it difficult to get good blanks – wide bore tubes (e.g. 6–10 mm I.D.)
are notoriously difficult to condition stringently and are also more
2. Selection of TD-compatible sorbents – factors to consider difficult to purge leading to increased risk of analyte and/or sor-
bent oxidation. If such tubes are packed with sorbent close to the
Sorbent-packed tubes and focusing traps that are compatible sampling end, they are also prone to error due to high diffusive
with thermal desorption typically contain between 1 and 4 sorbents uptake.
arranged in order of increasing sorbent strength from the sampling Selection of sorbents of appropriate strength allows quantitative
end. There are a range of factors to consider when selecting suit- retention and release of compounds ranging from C2 hydrocarbons
able sorbents or sorbent combinations including – the strength of and freons to semi-volatiles such as PCBs, phthalates and PAHs
the sorbent–sorbate interaction, artefacts, hydrophobicity, inert- without exceeding optimized tube/trap dimensions and without
ness and mechanical strength (friability) [1]. requiring liquid cryogen coolant (see part 1 of this paper).
A wide range of weak, medium and strong commercial sorbents
2.1. Sorbent ‘strength’ are now available for air monitoring (Table 1). Generally speak-
ing vapour-phase organics should be sampled using the weakest
Analytical sensitivity and precision are largely determined compatible sorbent, i.e. one that offers a practical/useful retention
by sampling efficiency, desorption efficiency and the level of volume and quick, quantitative recovery during desorption and
interferences (see Section 2.4). The sorbent or sorbents selected analysis.
 E. Woolfenden / J. Chromatogr. A 1217 (2010) 2685–2694 2687

Table 1
Commonly used, TD-compatible sorbents and their main features.

Sorbent Strength Max. Temp. Features

◦
Quartz wool Very weak >450 C Very inert, non-water retentive
CarbographTM 2TD Weak >450 ◦ C Hydrophobic
CarbopackTM C Minimal (<0.1 ng) artefacts
Carbotrap C Friable. 40/60 mesh recommended to minimise back pressure

Tenax® TA Weak 350 ◦ C Hydrophobic

Low inherent artefacts (<1 ng)
Inert – suitable for labile components

CarbographTM 1TD Weak/medium >450 ◦ C Hydrophobic

CarbographTM B Minimal (<0.1 ng) artefacts
Carbotrap Friable. 40/60 mesh recommended to minimise back pressure

Chromosorb 102 Medium 225 ◦ C Hydrophobic

High inherent artefact levels (∼10–50 ng/component)
Inert – suitable for labile components

PoraPak Q Medium 250 ◦ C Hydrophobic

High inherent artefact levels (∼10–50 ng/component)
Inert – suitable for labile components

Chromosorb 106 Medium 225 ◦ C Hydrophobic

High inherent artefact levels (∼10–50 ng/component)
Inert – suitable for labile components

PoraPak N Medium 180 ◦ C Hydrophobic

High inherent artefact levels (∼10–50 ng/component)
Inert – suitable for labile components

HayeSep D Medium 290 ◦ C Hydrophobic

High inherent artefact levels (∼10–50 ng/component)
Inert – suitable for labile components

CarbographTM 5TD Medium/strong >450 ◦ C Hydrophobic

Minimal (<0.1 ng) artefacts
Friable. 40/60 mesh recommended to minimise back pressure

CarbopackTM X Medium/strong >450 ◦ C Hydrophobic

Minimal (<0.1 ng) artefacts
Friable. 40/60 mesh recommended to minimise back pressure

Carboxen 569 Strong >450 ◦ C Minimal (<0.1 ng) artefacts

Inert – suitable for labile compounds
Less hydrophilic than most carbonised molecular sieves

Unicarb Strong >450 ◦ C Inert; not hydrophobic

Individual artefacts below 0.1 ng
Must be conditioned slowly
Requires extensive purge to remove permanent gases

Carboxen 1003 Very strong >450 ◦ C Inert; not hydrophobic

Individual artefacts below 0.1 ng
Must be conditioned slowly
Requires extensive purge to remove permanent gases

Carbosieve SIII Very strong >450 ◦ C Minimal (<0.1 ng) artefacts

Inert – suitable for labile compounds
Significantly water retentive – do not in humid conditions

Molecular sieve 5 Å Very strong >400 ◦ C High (∼10 ng) artefacts

Significantly hydrophillic – do not use in humid conditions

Molecular sieve 13× Very strong >400 ◦ C High (∼10 ng) artefacts
Significantly hydrophillic – do not use in humid conditions

2.2. Inertness strong sorbents comprise some form of carbonised molecular sieve
and, in this case, sorbent strength can be reduced by as much as a
Some sorbents contain chemically active materials. This is espe- factor of 10 at 90% RH [36]. If a large amount of water is retained
cially true of carbon blacks, many of which derive originally from on the tube and not selectively eliminated prior to analysis, this
natural charcoals and contain trace metals. These sorbents are too can adversely affect results. Water management options for
therefore generally unsuitable for labile (reactive) species – sulphur sorbent-based air monitoring are discussed in more detail below.
compounds, terpenes, amines, etc.

2.4. Artefacts
2.3. Hydrophobicity
Sorbents vary significantly with respect to inherent artefact lev-
Most common weak- and medium strength sorbents are very els. Some porous polymers such as the Chromosorb® Century series,
hydrophobic, thus their sorbent strength is not compromised even PoraPakTM and HayeSepTM series have relatively high artefacts with
when sampling at high (>80%) relative humidity. However, most several peaks at 5–10 ng levels. The porous polymer Tenax® TA is
2688 E. Woolfenden / J. Chromatogr. A 1217 (2010) 2685–2694

better with minimum levels between 0.1 and 1 ng for well con- Graphitised carbon blacks were developed by Italian scientists
ditioned materials. Both carbon blacks and carbonised molecular in the late 70 s [4] and were introduced commercially at roughly
sieves, are excellent with respect to inherent artefacts – between the same time as the first carbonized molecular sieves (CMSs).
0.01 and 0.1 ng if well conditioned. However, carbonised molec- CMS-type sorbents were introduced as highly sorptive (retentive)
ular sieves require extended conditioning at steadily increasing alternatives to charcoal for trapping very volatile compounds but
temperatures and can continue to show a high background of inor- with reduced hydrophilicity. Both these new sorbent groups were
ganic gases for several days when new. They may also become compatible with high temperatures and had low inherent artefact
irreversibly contaminated if allowed to come into contact with levels. Their introduction enabled multi-sorbent tube combinations
compounds higher boiling than xylenes (C8 ). to be used widely and effectively for the first time allowing com-
Porous polymeric sorbents may form trace artefacts when sam- pounds ranging in volatility from vinyl chloride to n-hexadecane
pling air containing significant concentrations of reactive gases and above to be monitored simultaneously [36,5,6].
such as ozone. This effect has been reported for Tenax® TA which Work on improving sorbent strength for organic vapours while
generates trace artefacts including benzaldehyde and acetophe- at the same time minimizing water retention continues in both
none if ozone concentrations exceed 100 ppb [2]. the US and Europe and has led to the introduction of new stronger
carbon blacks (e.g. CarbopackTM X and CarbographTM 5 TD) over
2.5. Temperature stability recent years. This new generation of sorbent materials offers quan-
titative retention of compounds as volatile as 1,3-butadiene while
Most sorbents, including the porous polymer Tenax® TA, are still remaining largely hydrophobic [7].
stable up to 350 ◦ C and many of the carbon sorbents can be taken A summary of the sorbents most commonly used for air moni-
to temperatures above 400 ◦ C. However, care must be taken with toring today is listed in Table 1.
most other porous polymer sorbents–chromosorbs, HayeSeps and Ongoing sorbent research revolves around different technolo-
PoraPaks – which typically have temperature limits at or below gies such as nano-particles [8], molecularly imprinted polymers
225 ◦ C. (which can be used to selectively trap large molecules according
to their shape [9]) and sorbents impregnated with derivatising
agents [36] to selectively retain or stabilise very specific ana-
2.6. Mechanical strength
lytes. However, none of these new types of material have yet been
demonstrated to offer sufficiently robust and repeatable perfor-
Graphitised carbon blacks are extremely friable and prone to
mance characteristics for widespread application. More research is
the formation of fines. Care should be taken not to over compress
required in each case.
these sorbents during tube packing and to avoid sharp knocks once
the tubes are packed. As the carbon packing ages, the formation
of fines may increase tube impedance (back pressure) beyond the 4. Multi-sorbent tubes
limit of some pumps. Most other sorbents are mechanically strong,
although Tenax® TA can have a high percentage of fines when new If a wide volatility range of compounds is to be monitored,
and may require sieving before use. Generally speaking, recom- it is often necessary to pack a tube with more than one sor-
mended mesh sizes for sorbents in standard 4–5 mm bore sampling bent material, arranged in order of increasing strength from the
tubes range from 30 to 80 mesh (approximately 0.6–0.2 mm parti- sampling end. (Fig. 1, see also part 1 of this paper.) Note that,
cle diameter). in the case of multi-sorbent tubes and traps, it is even more
critical than normal to use backflush desorption – i.e. the flow
2.7. Mesh size of gas through the tube/trap during desorption must be the
reverse of the air/gas flow during sampling. Higher boiling ana-
Within the 30–80 mesh range specified above, sorbent parti- lytes are thus retained by and desorbed from the weaker front
cle size does not play a critical role in sorbent selection because sorbent(s) without coming into contact with the stronger sorbents
analyte retention volumes will remain constant as the particle size behind.
increases up to a limit of 5 particles across the internal diameter of Key issues to consider in relation to sorbent selection for multi-
the sorbent tube/trap [3]. sorbent tubes include:

3. Sorbent developments • The volatility range of target compounds and quantitative reten-
tion and efficient desorption of each.
The earliest sorbent-based air monitoring studies were car- • Sorbent compatibility – the temperature required for condi-
ried out using charcoal with subsequent solvent (CS2 ) extraction tioning the most thermally-stable sorbent must not exceed the
(see part 1 of this paper). However, as the advantages of ther- maximum temperature limit of any other sorbent in the tube.
mal desorption became more widely understood, use of charcoal • Stability during storage. Loosely bound analytes can migrate from
declined. The strength, hydrophilicity and reactivity of natural weak to strong sorbents within a multi-sorbent tube during stor-
charcoal simply make it incompatible with thermal desorption age. In severe cases this can lead to irreversible adsorption and
for all but the most volatile and stable organic components. Early incomplete recovery during subsequent analysis. Some specific
TD-based air monitoring methods instead relied heavily on the sorbent combinations are more prone to this – for example; a
porous polymers–Tenax® TA and other common gas–solid chro- short bed of weak sorbent (e.g. CarbopackTM C or Tenax® TA),
matographic media such as the Chromosorb Century series and backed up directly by a very strong sorbent such as a carbon
PoraPak Q, N, etc. Over time, it became clear that additional sor- molecular sieve. Migration can be reduced by extending the bed
bent options were required because Tenax® was too weak to be length of weaker sorbent or inserting a medium strength sorbent
ideal for polar solvents and species more volatile than n-hexane. between the weak and strong sorbents. Sampled multi-sorbent
The inherently high artefact levels and temperature limitations of tubes should also be stored under refrigerated conditions and
the other porous polymers also restricted their application e.g. for analyzed quickly – most standard methods recommend a maxi-
trace level monitoring and/or for use in combination with other mum storage time of 30 days. Furthermore, care must be taken,
sorbents. particularly with multi-sorbent tubes, to make sure the rear sor-
 E. Woolfenden / J. Chromatogr. A 1217 (2010) 2685–2694 2689

Fig. 1. A multi-sorbent tube.

bent does not extend beyond the heated zone of the thermal If odorous and/or reactive compounds are of interest the ideal
desorption oven or this too could compromise recovery. starting point for monitoring uncharacterized atmospheres is to
• Minimising water retention by avoiding water retentive sorbents sample using several replicates of two slightly different sampling
wherever practicable (see below). trains in parallel. The two types of sampling train should each com-
prise three inert-coated steel tubes packed with single sorbents,
Provided these issues are taken into consideration when select- connected together in series using inert, non-emitting fittings. The
ing the sorbents, active sampling onto multi-sorbent tubes is a first would use a totally inert combination of sorbents for example;
relatively straightforward procedure and offers quantitative reten- the front tube packed with Tenax® TA, the middle one packed with
tion and recovery of analytes over a uniquely wide volatility range a stronger porous polymer such as Chromosorb 106 and the third
[10] (see Fig. 3 in part 1 of this paper). Research into multi-sorbent tube packed with UniCarb or Carboxen 1003. The second sampling
tube combinations is ongoing in order to extend the application train would be similar but with an alternative carbon black medium
range even further and better define the working limits – storage strength sorbent used in the middle tube. Subsequent analysis of
times, safe sampling volume, etc. each of the separate sorbent tubes used in both types of sampling
Common sorbent combinations for sampling tubes are as fol- train would highlight any analyte losses caused by use of a carbon
lows: rather than porous polymer medium strength sorbent and would
’Universal’ tubes – there is, of course, no such thing as a universal also help identify the optimum combination of sorbents for subse-
tube. However, perhaps the most broadly applicable combination quent monitoring of the same area using single tubes packed with
of sorbents that can be packed into a single tube for pumped mon- multiple sorbents.
itoring of uncharacterised atmospheres is Tenax® TA backed up by Note that multi-sorbent tubes/traps are only really applicable
a medium strength graphitized carbon black (e.g. CarbopackTM B or to active sampling – i.e. pumped tubes, canister sampling or online
CarbographTM 1 TD) backed up in turn by a carbonized molecular air monitoring. In diffusive sampling, only the sorbent at the front
sieve (e.g. UniCarbTM or CarboxenTM 1003) (Fig. 1). Standard-sized (sampling) surface is significant. Other sorbents may be present
tubes packed with equal bed lengths of each of these sorbents offer further back in the tube, but they generally play little or no role in
quantitative retention and release of compounds ranging in volatil- the sampling process.
ity from C3 hydrocarbons to n-C26 . Note also that the guidance given in Table 1 relates to sorbent
One limitation of this sorbent combination is that the mid- air sampling tubes – pumped or diffusive (see part 1 of this paper).
dle strength carbon black sorbent is not completely inert and However, similar considerations impact sorbent selection for the
may cause degradation of labile analytes such as nitrogen- cooled sorbent focusing traps – both as deployed for whole air sam-
or sulphur-containing compounds and monoterpenes. The rear, ples (online monitoring or canister/bag analysis) and as required for
‘strong’ sorbent is also a carbon molecular sieve and prone to water refocusing analytes during tube desorption.
retention requiring dry purge or some other water management
step if used to monitor humid atmospheres (see below).
Hydrophobic tubes – another useful combination of sorbents is 5. Other practical factors to consider for sorbent-based air
a short (∼5 mm) bed of clean quartz wool, backed up by Tenax® sampling
TA, backed up again by CarbopackTM X or CarbographTM 5 TD with
the bed lengths of Tenax:carbon being roughly in the proportion 5.1. Trap/tube materials
3 or 3.5:2. This combination of sorbents is hydrophobic and offers
quantitative retention and release of analytes ranging in volatility Sorbent tubes/traps should be constructed of inert, non-
from 1,3-butadiene to n-C32 and above. Relevant application exam- outgassing and thermally-stable materials.
ples include soil gas monitoring (e.g. for studies of vapour intrusion Focusing traps are typically constructed of quartz because it is
into buildings) and monitoring wide boiling range emissions from almost perfectly inert and is a good thermal conductor thus facili-
construction products and other materials. tating rapid heating and cooling.
Tubes used for US EPA-defined ‘air toxics’ [36] are typically Glass or quartz materials can also be used for sorbent air mon-
packed with a medium strength carbon black such as CarbopackTM itoring tubes, however, they are prone to breakage and are not
B or CarbographTM 1 TD backed up by a carbon molecular sieve such generally suitable for diffusive (passive) monitoring (see part 1 of
as Carboxen 1003 in the bed length proportion 2:1 or 1:1. Such ‘air this paper). Stainless steel-based tubes are a more robust and ver-
toxics’ tubes offer quantitative retention of ethane from small vol- satile option for most field monitoring exercises and inert-coated
umes of air combined with efficient release of compounds up to versions (e.g. Silcosteel® ) provide an inert and robust option for
n-C12/14 . reactive compounds.
2690 E. Woolfenden / J. Chromatogr. A 1217 (2010) 2685–2694

5.2. Sorbent retaining materials step will be required. The sample split should be used in two stages
for optimum effect i.e. during both primary (tube) desorption and
Sorbents are typically held in place using fine mesh stainless secondary (trap) desorption.
steel or Silcosteel gauzes (stainless steel tubes) or glass frits and/or Dry purging involves passing a flow of pure dry air or nitrogen
plugs of inert quartz or glass wool (glass tubes.) The steel sorbent through the tube and/or focusing trap in the sampling direction
retaining gauzes can be very precisely and reproducibly located in prior to desorption. CMS sorbents have more affinity for organic
metal tubes (see Fig. 1) facilitating the required level of precision molecules (even highly polar compounds like light alcohols) than
for diffusive (passive) sampling and the interchange of uptake rate water. This allows dry purge conditions (temperature, dry gas
information e.g. in standard methods [36]. flow and time) to be set such that water is selectively purged to
vent without loss of the most volatile/polar compounds of interest
5.3. Preparation of sorbent tubes/traps [14,15]. It is most effective to implement dry purging in two steps
i.e. for both the sample tube and focusing trap. Both steps can be
Sorbents invariably require stringent conditioning at high tem- automated on most modern commercial TD systems.
peratures in a flow of inert gas to clean them before use. Some of It is also important to selectively eliminate water from whole air
the older porous polymer-type sorbents also require precondition- samples (online or containers) before they are transferred from the
ing, before they are used to pack tubes, because as much as 10–15% focusing device to the analytical system. This is carried out in one
of sorbent mass may be lost during the first conditioning cycle. It of two ways depending on the nature of the application. In the first
is rare for any form of solvent washing to be required, but temper- option, very volatile, non-polar compounds are compatible with in-
ature and flow conditions used for tube cleaning should invariably line permeable membrane dryers such as the Nafion® dryer. Such
be more stringent than those to be used subsequently for analyti- dryers eliminate water and polar organic compounds from the air or
cal thermal desorption. More information on this important issue gas sample stream before it reaches the focusing trap. This elimina-
is given in the literature [11]. tion of polar VOCs can be a real benefit by simplifying some online
air monitoring applications which rely on conventional FID detec-
5.4. Storage of conditioned and sampled sorbent tubes tion instead of MS. A key example is monitoring ppb level ‘ozone
precursors’ (C2 to C10 hydrocarbons) in urban air.
Conditioned and sampled tubes should be stored using long- If the lightest compounds such as C2 hydrocarbons or the most
term 1/4-in. screw caps fitted with combined PTFE ferrules as volatile freons are not of interest, it is alternatively possible to selec-
described in standard methods [36]. Sampled, single sorbent tubes tively dry purge water from the focusing device, as described above,
capped and stored in this way are reported to be stable for up to 27 whilst retaining the compounds of interest. In this case trapping
months [12] at room temperature provided the compounds con- temperature, sorbents and sampling flow are all selected to mini-
cerned are not chemically active. Multi-sorbent tubes should be mize water retention while still ensuring quantitative recovery of
analysed more quickly to minimize risk of analyte migration within the compounds of interest from the whole air sample. A focusing
the tube (see above). trap dry purge step (see above) is also implemented between sam-
pling and trap desorption/analysis. In this case, no in-line dryer is
6. Water management required and polar compounds can be measured.

GC(MS) technology is notoriously sensitive to water interfer-

7. Optimisation of desorption and other analytical
ence. Adverse effects include baseline anomalies and unpredictable
parameters
quenching of the detector response (signal) for compounds which
co-elute with the water. High water levels can also reduce the work-
As discussed above, thermal desorption is rapidly displacing
ing life of key system components such as capillary columns and
solvent extraction as the method of choice for air monitoring gen-
MS detectors.
erally. Relevant TD applications include ambient air monitoring,
Sorbent tubes (pumped or diffusive) offer many water manage-
indoor (and in-vehicle) air quality assessment, monitoring personal
ment options. Wherever possible, the most effective of these is to
exposure at work (occupational/industrial hygiene), industrial
avoid collecting water in the first place by using hydrophobic sor-
emissions (stack, vent, fugitive emissions), factory fence-line mon-
bents in the sampling tube. (Note that sorbent tubes must never
itoring, landfill gases, odour monitoring and atmospheric/pollution
be colder than the sampled air/gas to prevent water condensation
research.
within the tube.) The vast majority of vapour-phase organics can
Advice on optimising and validating the thermal desorption and
be quantitatively retained from reasonable volumes of air without
GC(MS) analytical process has been presented in the literature [1]
resorting to the less-hydrophobic CMS-type materials. Compounds
and is outlined in many of the international standard methods cited
as volatile as 1,3-butadiene (boiling point: −4.4 ◦ C) can be quanti-
in this paper. As specified in these standards; key requirements for
tatively retained using CarbographTM 5 TD or CarbopackTM X [7,13].
successful TD-GC(MS) analysis of the organic vapours retained on
However, if the compounds of interest include even more
sorbent tubes and traps include:
volatiles species, such as vinyl chloride or Freon® 113, it will be
necessary to include at least a short bed of strong carbon molecular
sieve at the rear (non-sampling) end of the tube. Such sorbents are • Reliable sealing of tubes before and after analysis (automated
prone to some water retention and water management measures systems).
will be required particularly when such tubes are used to sample • Automated and stringent leak testing.
humid atmospheres. The main options in this case are use of sample • Pre-purge of oxygen to vent (in desorption direction) to prevent
splitting (convenient when monitoring high level air contaminants sorbent or analyte oxidation.
such as work place atmospheres or stack or landfill emissions) or • Optional internal standard addition (in sampling direction).
dry purging. • Optional dry purge (in sampling direction).
With respect to sample splitting: If the mass of target analytes • Complete primary (tube) desorption (if applicable) – i.e. >99%
collected is such that a split ratio of 50:1 or more can be set with- extraction efficiency.
out compromising detection limits for the lowest concentration of • Quantitative (re)focusing of analytes using a cryogen-free sorbent
interest, then it is unlikely that any additional water management trap.
 E. Woolfenden / J. Chromatogr. A 1217 (2010) 2685–2694 2691

• Fast secondary (trap) desorption with >99% extraction efficiency Thermal desorption-based air monitoring methods were also
and rapid transfer/injection of desorbed analytes into the GC(MS) historically prone to some limitations including restricted facili-
analyser. ties for internal standard introduction and the inherent ‘one-shot’
nature of the technique which could make it difficult to repeat
The linearity of thermal desorption – GC(MS) methods should be and confirm results. As early as 1981 tube desorption (TD) sys-
the same as can be achieved using GC(MS) systems configured with tems began to be introduced which addressed the ‘one-shot’
conventional liquid inlets. The precision of TD methods is typically limitation to some extent by stringently leak testing every sam-
limited to 1–2% by the manual introduction of external gas or liq- ple prior to analysis to minimize risk of error – essentially TD
uid phase standards to sorbent tubes during calibration. However, requires the carrier gas flow path to be broken into each time
this is insignificant relative to the overall variability air monitoring a new tube is analysed and desorbed. Risk of undetected leaks
methods – typically quoted at 15–30%. would bring all TD-GC/MS data into question especially if sample
analyses could not be repeated. This explains why leak test-
8. Uncertainty and analytical quality assurance ing is specified as a mandatory requirement in standard TD
methods.
All air monitoring methods are multi-step processes resulting in As TD technology improved over the years, facilities for auto-
relatively high uncertainty. Variability can creep in at every stage matic introduction of gas-phase internal standards (IS), first onto
– during sample collection, during transport/storage, via contam- the rear of sorbent tubes and then onto the sampling end became
ination, because of sorptive losses (sink effects) in the sampler widely available. Some systems now even offer the ability to load
(particularly in canisters or bags), during the TD-GC/MS calibration internal standard onto blank as well as sampled tubes allowing pre-
process and during desorption and analysis. spiked sorbent tubes to be used for field monitoring exercises as a
Solvent extraction based air monitoring methods are also prone check on the entire process including tube transport, storage, field
to specific and significant additional limitations. These include vari- sampling and TD-GC/MS analysis. IS addition is also a standard fea-
able recovery, low sensitivity, analytical interference by the solvent ture of most automated thermal desorption systems for canisters,
and the number of manual steps required (see part 1 of this paper). bags or online air/gas streams.

Fig. 2. (a) Sequence of chromatograms showing re-analysis of re-collected phthalate mixture using an automated ULTRA 50:50TM thermal desorber from Markes International
Ltd., UK. (b) Plot showing the theoretical (lines) and measured (points) peak areas for multiple re-collection and repeat analyses of the sample shown in a.
2692 E. Woolfenden / J. Chromatogr. A 1217 (2010) 2685–2694

Fig. 3. RFID tags attached to thermal desorption sorbent tubes.

Another more recent innovation is the ability to quantitatively

re-collect any primary (tube) or secondary (trap) desorption split
flow into a single conditioned sorbent tube for repeat analysis.
This overcomes the one-shot limitation of traditional TD systems,
but more importantly provides a convenient means of validating
analyte recovery through the thermal desorber. Whereas the ear-
liest standard methods for TD specified validation of recovery by
comparing the TD calibration curve with that for a conventional
liquid injection under the same conditions, re-collection allows a
sequence of repeat analyses to be carried out on a single sample
(see Fig. 2a and b). Any loss of one or more analytes can be readily
identified from the relative responses to other compounds in the Fig. 4. Child’s plastic toy inside a micro-chamber prior to analysis. (a) Schematic of
mix or by a deviation in the result from that predicted from the split a single micro-chamber as used for bulk emission testing.

ratio. This approach has now been adopted into some of the most
recent thermal desorption standards [36].
9.1. Rapid assessment of chemical emissions from products and
Electronic labeling has recently been introduced for sorbent
materials
tubes (Fig. 3) and other air samplers – canisters, bags, etc. – which
make it much easier to track the history and performance of indi-
Recent legislative activity – e.g. under REACH [17], the Cali-
vidual samplers. The development is based on RFID technology
fornia ‘formaldehyde rule’ [18] plus the European Construction
encapsulated in TD-compatible tags which can be permanently
Products Directive (CPD) [19] and its successor the Construction
attached to the same tube (or canister) throughout its life. The tags
Products Regulation (CPR) [20] – is driving increased evaluation
can be used to record field/sample data (dates, sampling start/end
of vapour-phase chemical emissions from products and materi-
times, sampled volumes, pressures, flows, etc.), project information
als [21,22]. Many of the new regulations demand both third-party
and, most importantly, information on the history of the sampler
certification of chemical emission levels by accredited labora-
itself–sorbents, packing dates, number of thermal cycles, etc. The
tories using reference methods plus ongoing demonstration of
tags can be used for specific projects–only remaining attached to a
product conformity via ‘factory production control’. Reference pro-
given sampler throughout a field monitoring exercise before being
cedures for product emission testing require sample materials
removed and redeployed elsewhere. Alternatively (and more com-
to be placed in test chambers or cells to simulate the indoor
monly) tags are attached to a particular sorbent tube throughout
environment followed by sample collection on sorbent tubes and
its life offering users ready confirmation of the sorbents inside and
subsequent TD-GC(MS) analysis [36,23,24]. These procedures are
a much improved means of monitoring the performance and valid-
effective and increasingly well validated [25] but time consuming
ity of that sampler as it ages. Depending on how the technology is
and expensive. Each test takes between 3 and 28 days. While fea-
implemented on the TD-GC(MS) system key details such as back
sible for certification, faster and simpler tools and procedures are
pressure, number of leak test failures, number of thermal cycles
required for in-house emissions screening as part of routine quality
and even artefact levels can be stored and tracked on the tag itself
control.
and/or in an associated database thus greatly improving quality
Micro-chamber technology (Fig. 4) has recently been developed
assurance [16].
to address this need and allow chemical emissions (bulk or surface-
only) to be screened in minutes or hours rather than days. This
9. Extending the application range for sorbent-based ‘air’ new approach has been shown to correlate with reference methods
monitoring technology [26,27] and is undergoing rapid standardisation [28,29]. Most com-
mercial systems can be heated to moderate temperatures and are
A number of specialist sampling accessories have been intro- configured with multiple (typically 4 or 6) micro-chambers allow-
duced in recent years to interface difficult or unusual sample ing multiple samples to be tested in parallel. Both sorbent tubes
matrices to standard air monitoring technology. Three key exam- and DNPH cartridges (for monitoring formaldehyde) are usually
ples – for materials emissions testing, human breath and soil – are accommodated.
described briefly below. All may be interfaced to sorbent tubes or In the future micro-chambers could provide a versatile interface
alternative online air monitoring technology incorporating one or for many natural and man-made materials allowing the power of
two sorbent focusing traps. sorbent tubes/traps and TD-GCMS air monitoring technology to be
 E. Woolfenden / J. Chromatogr. A 1217 (2010) 2685–2694 2693

Fig. 6. Soil probes arranged in a grid pattern around an industrial site allow low-
Fig. 5. Bio-VOC breath sampler (Markes International Ltd., UK) developed by UK cost mapping of underground contamination. Inset shows example VOC-MoleTM soil
HSL. probe.

applied to vapour profiling an almost universal range of real world true picture of the level and spread of underground contamination
solid or liquid samples. to be built up.

9.2. Breath sampling

10. Summary
A number of breath samplers have been developed over recent
Sorbent tubes/traps form the basis of most monitoring tech-
years to interface to sorbent tubes [30,31]. The example shown
nology for organic vapours in air and a wealth of supporting
in Fig. 5 is a disposable device developed by the UK Health
information is now available in the literature and in the form of
and Safety Laboratory (HSL) for non-invasive studies of biologi-
standard methods to those who are new to the field.
cal exposure to chemicals at work, i.e. as a means of assessing
New developments, primarily in associated sampling tech-
the total body burden from all routes of exposure – ingestion
nology, are also extending the application range away from
and skin absorption as well as inhalation [32]. It collects ∼100 ml
conventional air monitoring and into more exotic fields such as
samples of end-tidal air which are subsequently transferred to sor-
odour & emission profiling and disease diagnosis via volatiles in
bent tubes and analysed by conventional thermal desorption –
exhaled breath.
GC(MS).
Other applications for breath testing include monitoring long-
term environmental exposure – particularly for individuals living Trademarks
near local emission sources (e.g. above a dry cleaning shop) or for
those living in areas with highly chlorinated water. Breath sampling Tedlar® , Freon® and Nafion® are registered trademarks of E.I.Du
is also under extensive research as a potential clinical diagnostic Pont de Nemours Co, USA.
tool [33]. Tenax® TA is a registered trademark of Buchem bv, Netherlands.
Chromosorb® is a registered trademark of Celite Corporation,
9.3. In situ sampling of VOCs in soil USA.
PoraPakTM is a trademark of Waters Corporation, USA.
Soil probes fitted with sorbent tubes were initially developed in CarbographTM is a trademark of LARA s.r.l., Italy.
the petrochemical industry [34] for rapid screening of large indus- UniCarbTM , UNITYTM , ULTRA 50:50TM , Bio-VOCTM and VOC-
trial sites. Fig. 6 shows the type of concentration map obtained MoleTM are trademarks of Markes International Ltd.
from a large scale screening operation of a brownfield site and the FLEC® is a registered trademark of CHEMATEC, Denmark.
inset shows an example soil probe. Diffusive samplers are most Silcosteel® is a registered trademark of Restek Corporation, USA.
commonly used in soil probes [34,35] because they are easy and HayeSepTM D is a trademark of Hayes Separations Inc., USA.
cost-effective to deploy in large numbers. In situ monitoring of soil CarbotrapTM X and CarbopackTM B are trademarks of
contamination involves minimal disturbance of the land allowing a Sigma–Aldrich Biotechnology LP and Sigma–Aldrich Co.
2694 E. Woolfenden / J. Chromatogr. A 1217 (2010) 2685–2694

References [20] EC Construction Products Regulation (CPR). Revised text adopted May 2009.
For more information contact Europe and Standards Branch, Communities
[1] E.A. Woolfenden, J. Air Waste Manage. Assoc. 47 (January) (1997) 20. and Local Government, UK (construction.products@communities.gsi.gov.uk) or
[2] P. Ciccioli, E. Brancaleoni, A. Cecinato, C. Di Palo, A. Brachetti, A. Liberti, J. Chro- equivalent European agencies.
matogr. 351 (1986) 433. [21] AgBB/DIBt, 2000 Health-related evaluation procedure for volatile organic
[3] J.A. Rijks, C.A. Cramers, P. Boeck, Chromatographia 8 (1975) 482. compound emissions from building products, DIBt-Mitteilungen 1/2001,
[4] P. Ciccioli, G. Bertoni, E. Brancaleoni, R. Fratarcangeli, F. Bruner, J. Chromatogr. pp. 3–12 (last updated 2008) http://www.umweltbundesamt.de/building-
126 (1976) 757. products/agbb.htm.
[5] W.A. McClenny, K.D. Oliver, H.H. Jacumin Jr., E. HunterDaughtry Jr., J. Environ. [22] Collaborative for High Performance Schools, Reference Specifications for
Monit. 4 (2002) 695. Energy and Resource Efficiency, Section 01350, Special Environmental Require-
[6] P. Ciccioli, E. Bracaleoni, A. Ceccinato, R. Sparapini, M. Frattoni, J. Chromatogr. ments, 2004 (http://www.chps.net/manual/index.htm).
643 (1993) 55. [23] ISO EN 16000-10. Indoor air. Part 10: determination of the emission of VOCs
[7] UK Health and Safety Laboratory Methods for the determination of haz- from construction products and furnishings – emission test cell, 2006.
ardous substances (MDHS) 53/2. 1,3-butadiene in air. Laboratory method using [24] ISO EN 16000-11. Indoor air. Part 11: determination of the emission of VOCs
pumped samplers, thermal desorption and gas chromatography, 2003. from construction products and furnishings – sampling, storage of samples and
[8] N. Watson, M. Bates, Evaluation of nanoparticles as a potential sorbent for preparation of test specimens, 2006.
thermal desorption. Markes International Internal Report (available from [25] O. Wilke, W. Horn, K. Wiegner, O. Jann, W. Bremser, D. Brodner, S. Kalus, R.
enquiries@markes.com), 2009. Juritsch, C. Till, Investigations for the improvement of the measurement of
[9] G. Bunte, J. Hürttlen, H. Pontius, K. Hartlieb, H. Krause, Anal. Chim. Acta 591 VOCs from floor coverings within the health related evaluation of construction
(2007) 49. products. BAM: German Federal Institute for Materials. Report ref: ZP-52-5-
[10] M.D. Wright, N.T. Plant, R.H. Brown, Storage trial of TO-14 VOC mix on 20.49.1-1251/07, March, 2009.
Supelco/Perkin-Elmer thermal desorption tubes. UK Health and Safety Labo- [26] T. Schripp, B. Nachtwey, J. Toelke, T. Salthammer, E. Uhde, M. Wensing, M.
ratory Report: IACS/98/02, 1999. Bahadir, Anal. Bioanal. Chem. (January) (2007).
[11] D. Helmig, J. Chromatogr. A 732 (1996) 414. [27] M. Pharaoh, G. Williams, Final report: work on the correlation between the VDA
[12] F. Lindquist, H. Bakkeren, Stability of chlorinated hydrocarbons on Tenax. 276 test and micro-chamber testing. Warwick Manufacturing Gp., University
CEC commissioned report from TNO, the Netherlands, Report No. R90/268, of Warwick PARD Extension Report, October, 2009.
1990. [28] ISO WD 12219-3, Indoor air – road vehicles. Part 3: screening of chemical
[13] K. Gostlow, N.T. Plant, M.D. Wright, Evaluation of carbon sorbents for the deter- emissions from car interior trim components. Micro-scale chamber method.
mination of 1,3-butadiene in air by thermal desorption. Health and Safety [29] ASTM WK 22044 Standard practice for rapid screening of VOC emissions from
Laboratory Report, 2002, OMS/2002/13. products using micro-scale chambers (draft work item within D22.05).
[14] D. Helmig, L. Vierling, Anal. Chem. 67 (23) (1995) 4380. [30] M. Phillips, Sci. Am. (July) (1992) 74.
[15] A. Tipler, Proceedings of the Air and Waste Management Assoc. Conf.: Mea- [31] D. Dyne, J. Cocker, H.K. Wilson, Sci. Tot. Environ. 199 (1997) 83.
surement of Toxic and Related Air Pollutants (May) (1994). [32] H.K. Wilson, A.C. Monster, Occup. Environ. Med. 56 (1999) 753.
[16] E.A. Woolfenden, Diffusive Monitor. (February) (2008). [33] M. Phillips, R. Cataneo, B.A. Ditkoff, P. Fisher, J. Greenberg, R. Gunawardena, C.S.
[17] ECHA: Guidance on requirements for substances in articles, May 2008 Kwon, O. Tietje, C. Wong, Prediction of breast cancer using volatile biomark-
(http://echa.europa.eu/reach en.asp). ers in the breath. Springer, doi:10.1007/s10549-006-9176-1, Breast Cancer
[18] California Air Resources Board: airborne toxic control measure to reduce research and Treatment, 2006.
formaldehyde emissions from composite wood products. California code of [34] K.J. Saunders, Diffusive Monitor. 1 (September) (1988).
regulations. 93120 to 93120.12 (implemented January 2009). [35] J.E. Odencrantz, H. O’Neill, J. Remediation (Spring) (2009) 71.
[19] EC Construction Products Directive (CPD). 1989/106/EEC, 1989. [36] E. Woolfenden, J. Chromatogr. A 1217 (2010) 2674.