The Diffusive

Monitor
April 2008
Issue 16
Health and Safety Executive
Committee of Analytical
Requirements - Working Group 5
Inside this Issue
1
By way of introduction
1
Current News
4
NO
2
and SO
2
in ambient
air by membrane-closed
Palmes tube
7
Sample tube tagging
Enhanced tracking for
thermal desorption
10
Controlling GC carrier flow
through thermal desorb-
tion transfer line
15
Uptake of occupational
BTX on Carbograph TD1
tube sampler







Conferences..p.17
Registering your interest
in Diffusive Monitor. p. 19 e Monitor. p. 19


By way of Introduction By way of Introduction
Welcome to the sixteenth edition of The
Diffusive Monitor, which is a free
publication of the Health and Safety
Executive CAR Committee (Committee
of Analytical Requirements), Working
Group 5. This working group is
concerned with workplace, indoor and
environmental applications of diffusive
sampling for assessing air quality.
Welcome to the sixteenth edition of The
Diffusive Monitor, which is a free
publication of the Health and Safety
Executive CAR Committee (Committee
of Analytical Requirements), Working
Group 5. This working group is
concerned with workplace, indoor and
environmental applications of diffusive
sampling for assessing air quality.
The newsletter was started in May 1988
as a consequence of the Diffusive
Sampling Symposium held in
Luxembourg the previous year and was
originally published about once a year.
In recent years the frequency has
reduced for reasons described later in
Current News. It contains articles on
diffusive monitoring techniques and
applications, and is a useful source of
information on European and
international standardisation in this area
and of sampling rate data. Contributions
are mostly from members of the
Working Group, which has an
international membership.
The newsletter was started in May 1988
as a consequence of the Diffusive
Sampling Symposium held in
Luxembourg the previous year and was
originally published about once a year.
In recent years the frequency has
reduced for reasons described later in
Current News. It contains articles on
diffusive monitoring techniques and
applications, and is a useful source of
information on European and
international standardisation in this area
and of sampling rate data. Contributions
are mostly from members of the
Working Group, which has an
international membership.
Contributions to the newsletter are not,
however, intended to be exclusively
from CAR/WG 5, and any reader is
welcome to submit an item for
consideration. The only limitations are
that articles should have some diffusive
sampling application and should not be
too obviously commercial.
Contributions to the newsletter are not,
however, intended to be exclusively
from CAR/WG 5, and any reader is
welcome to submit an item for
consideration. The only limitations are
that articles should have some diffusive
sampling application and should not be
too obviously commercial.
The newsletter has a world circulation
of some 200 people, all of whom have
specifically requested the publication,
so if you wish to contribute articles, you
can be assured of a wide and receptive
audience. Articles are not peer-
reviewed, but are subject to the Editors
discretion. A Word template for authors
is recommended and is available from
the Editor on request.
The newsletter has a world circulation
of some 200 people, all of whom have
specifically requested the publication,
so if you wish to contribute articles, you
can be assured of a wide and receptive
audience. Articles are not peer-
reviewed, but are subject to the Editors
discretion. A Word template for authors
is recommended and is available from
the Editor on request.
Copies of this newsletter and previous
issues back to no.12 (July 2001) may be
downloaded from the Health and Safety
Laboratory website at
http://www.hsl.gov.uk/publications/diffuse-
Copies of this newsletter and previous
issues back to no.12 (July 2001) may be
downloaded from the Health and Safety
Laboratory website at
http://www.hsl.gov.uk/publications/diffuse-
monitor.htm
No registration is necessary to
download a copy. However, if you want
to be placed on a list to be notified when
a new issue is published contact the
Editor (but see below for a change of
Editor in 2008).
Copyright statement: Issues 1-11 of
The Diffusive Monitor are available on
request to the Editor. The early issues
can be supplied by e-mail as PDF files
on condition that they are for in-house
use, private study and not for
distribution on a website or by other
means. These particular copyright
restrictions do not apply to issues 12
and later, but in all cases the source
should be acknowledged if quoting.
Current News
First some personal news. After 35
years with HSL and its predecessors I
will be retiring in May 2008 to pastures
new. My successor as Editor and
secretary of CAR/WG5 will be Neil
Plant (neil.plant@hsl.gov.uk ) of HSL
who has been closely involved in the
practical side of workplace and ambient
monitoring campaigns for over 10
years.
Committee of Analytical
Requirements (CAR) on the back
burner
Since the last issue in February 2006 the
minutes of recent CAR meetings have
been placed on the HSL website at
http://www.hsl.gov.uk/publications/car.htm
the last one being held at HSL Buxton
on 16 May 2006. However, a meeting
scheduled for October 2006 was
1 . . . . . . . . . . . . . . . . . . . . . . .
April 2008
postponed. The secretary of CAR has
taken into account the response of
members to further meetings in 2007,
the fact that only one working group is
active (WG5) and it was agreed in HSL
that further proceedings of the main
CAR committee would be by
correspondence only. The decision to
end face to face CAR meetings was
taken with reluctance but reflects what
is happening in specialist measurement
forums elsewhere. CAR/WG5 continues
unaffected with Dr Kevin Saunders as
chairman.
CEN air quality standards and
European news
Issue 15 of The Diffusive Monitor
described the progress of various work
items in expert working groups (WGs)
reporting to Technical Committees of
the Comit Europen de Normalisation
(CEN) up to the end of 2005. CEN is a
legal association, the members of which
are the national standards bodies of EU
member states and associate countries,
supported by a management centre in
Brussels. The programmes of TC137
(workplace exposure) and TC264 (air
quality) derive mainly from the
requirements of the Chemical Agents
Directive (CAD)(98/24/EC) and the
Ambient Air Directive (96/62/EC)
respectively [1,2]. TC137 prepares
standards for the protection of workers
against hazardous substances and
biological agents in workplace
atmospheres. Its work excludes the
proposal of limit values which are
established by separate expert
committees. TC264 has a similar role
for ambient atmospheres.
An ad hoc group of TC137/WG2
completed its report on recommended
methods for measuring priority
chemical substances in workplace air
[3,4] and an on-line summary became
available in 2006 [5]. Following its
publication the convenor of the ad hoc
group Dr Dietmar Breuer of
Berufsgenossenschaftliches Institut fr
Arbeitsschutz, St Augustin Germany
(BGIA) proposed an international
database of occupational airborne limit
values to be maintained by BGIA. Since
first going on line in late 2006 this
database has now expanded in March
2008 to include the limit values of
Austria, Belgium, Canada (Qubec),
Denmark, European Union, France,
Germany (AGS), Germany (DFG),
Hungary, Italy, Japan, Spain, Sweden,
Switzerland, Netherlands, USA (OSHA
PEL) and United Kingdom [6]. Ireland
and Poland may be added in the near
future. For copyright reasons the
ACGIH TLVs have to be excluded.
Between these regulatory authorities
about 1100 substances are listed,
therefore coverage is more extensive
than that of any single country.
However, for legal purposes some
caution is necessary. The BGIA
database is useful for comparing limit
values of one country with another, but
has indicative status. Wherever possible
there are linked citations that point to
the national source material (EH40
Table 1 in the case of the UK). The
national source material will always be
the official controlled version.
The revised 'General requirements for
the performance of procedures' standard
EN 482 [7] was published in 2006.
Numerical performance requirements
(eg. relative expanded uncertainty
30 % at 0.5 2 times limit value) are
unchanged from the 1994 version. One
of the changes has been the removal of
a statistical confusion over method
'bias'. In accord with the Guide to the
Expression of Uncertainty in
Measurement (GUM) bias is only part
of the uncertainty budget if it is
unknown [8]. In EN 482:1994, because
of the way bias was presented as part of
'overall uncertainty', consistent and
correctable bias was not clearly
distinguished from unknown and
uncorrectable bias. Also EN 482:2006
can now refer to tests in daughter
standards such as prEN 838 [9] and
prEN 1076 [10] that were yet to be
decided in 1994. Publication of the
revised EN 838 (diffusive sampling)
and EN 1076 (active sampling)
standards is expected not later than
2009. For prEN 838 the working group
has decided to exclude direct-reading
stain length samplers, which really
should have their own different
performance requirements, but reagent-
impregnated systems are included. The
statistical treatment of uncertainty will
be quite different to EN 838:1995, being
partly based on GUM and partly on the
Nordtest Handbook [11].
In the ambient air quality field
TC246/WG11 (diffusive sampling) is
waiting for funding questions to be
resolved before progressing further with
a diffusive EN standard for NOx among
others. In TC264/WG12 (Reference
methods for SO
2
/NOx/O
3
/CO) the
publication of corrections to existing
standards has been delayed for similar
reasons.
ISO air quality standards
The International Organisation for
Standardisation (ISO) has a similar
management structure to CEN.
Workplace, ambient and indoor air
quality aspects in ISO are covered by
sub-committees of TC146.
In TC146/ SC2 (Workplace
atmospheres) a revision of the diffusive
sampling protocol ISO 16107 was
published in 2007 [12]. This standard is
complementary in many ways to EN
838 and there are no conflicting aspects.
While ISO 16107 contains no
performance requirements as such,
examples of performance tests are
described in some detail. The revised
isocyanate by liquid chromatography
standard ISO 16702 using
methoxyphenypiperazine reagent, was
published in December 2007 [13]. An
isocyanate standard using
anthracenylmethylpiperazine (MAP)
reagent is at the draft stage and has been
balloted in ISO [14]. A guide for
selection of isocyanate methods was
published in April 2006 as a Technical
Report ISO/TR 17737 [15]. Since the
last issue of The Diffusive Monitor the
four generic ISO standards for
measurement of volatile organic
compounds (VOCs) by solvent and
thermal desorption have been up for
systematic review and have all been
renewed [16-19].
In TC/146/SC4 (General aspects) the
ISO 9169 standard that defines
performance tests for automatic
measuring systems was published in
July 2006 [20] and the ISO 20988 air
quality guide to estimating
measurement uncertainty was published
2 . . . . . . . . . . . . . . . . . . . . . . . .
The Diffusive Monitor, 16
3 . . . . . . . . . . . . . . . . . . . . . . . .

in July 2007 [21]. Annex C.7 of ISO
20988 includes a worked example of
uncertainty estimation in diffusive
sampling of NO
2
compared with a
reference method.
In TC146/SC6 (Indoor air) the VOC
measurement strategy standard ISO
16000-5 was published in 2007 [22].
ISO 16000-2 (formaldehyde sampling
strategy) and ISO 16000-4
(formaldehyde by diffusive sampling)
were circulated for systematic review
[23, 24].
UK Methods for the Determination of
Hazardous Substances (MDHS)
For a list of titles and revision history in
the UK MDHS series see
http://www.hsl.gov.uk/publications/mdhs_list.htm
All MDHS titles currently in print are
available for download from the HSE
site at
http://www.hse.gov.uk/pubns/mdhs/index.htm.
For contractual reasons we have not
migrated old out-of-print MDHS titles
to the HSE website as promised in the
last CAR minutes, however, they will
be available by Email as searchable
PDF files by application to the Editor.
The only restriction is that they are for
in-house study and not for onward
distribution or uploading to a website.
The criterion for inclusion in the
"available" out of print group is that the
titles are not actually withdrawn as
technically deficient. In a few cases they
are cross-referenced by titles still in
print.
Proficiency testing news
A training DVD aimed at WASP
participants is available for purchase
from HSL via the link below. If you
have Windows media player there is a
link on the WASP information page to a
short video extract from the DVD.
http://www.hsl.gov.uk/proficiency-
testing/wasp.htm#dvd
The topics covered include:
metals on filters by ICP-AES;
VOCs on charcoal by solvent
desorption and gas
chromatography;
VOCs on Tenax by thermal
desorption and gas
chromatography;
isocyanate derivatives on glass
fibre filters using liquid
chromatography.
Other news from CAR/WG5 members
Downloadable English language reports
are available for two major studies of
low-level pollutants, the first being the
Flemish Indoor Exposure Study (2005-
2007) Amongst the substances
measured were benzene, toluene, xylene
formaldehyde, particulate matter and
nitrogen dioxide with the objective of
determining the indoor/outdoor relation.
Recently more English language
versions of the latest German
Environmental Survey for Children
(2003/06 - GerES IV) have been issued
and are freely downloadable. Blood and
urine monitoring results from GerES IV
are also obtainable via the above link
although the biomonitoring study did
not involve VOC markers or other
volatiles, but rather heavy metals and
persistent organic pollutants.
References
1. Council Directive 98/24/EC on the
protection of the health and safety of
workers from the risks related to chemical
agents at work (1998).
2. Council Directive 96/62/EC on ambient air
quality assessment and management (1996).
3. Comit Europen de Normalisation (CEN):
Project BC/CEN/ENTR/000/2002-16
Analytical Methods for Chemical Agents
Final Report, Sankt Augustin, Germany, 27
June 2005. Brussels: CEN, 2005.
4. D. Breuer et al (2006). Journal of
Occupational and Environmental Hygiene,
3: D126D136.
5. http://www.hvbg.de/e/bia/gestis/analytical_
methods/
6. http://www.hvbg.de/e/bia/gestis/limit_values
/index.html
7. EN 482:2006 Workplace atmospheres -
General requirements for the performance of
procedures for the measurement of chemical
agents.
8. Guide to the Expression of Uncertainty in
Measurement (ISO, 1995, equivalent to EN
13005:1998 equivalent to BS PD 6461 Part
3:1995).
9. prEN 838 Workplace atmospheres
Diffusive samplers for the determination of
gases and vapours Requirements and test
methods.
10. prEN 1076 Workplace atmospheres
Pumped sorbent tubes for the determination
of gases and vapours Requirements and
test methods.
11. Practical Handbook for Calculation of
Uncertainty Budgets for Accredited
Environmental Laboratories, Technical
Report No. 537, February 2003, Nordtest
project 1589-02,
http://www.nordicinnovation.net/nordtestfile
r/tec537.pdf
12. ISO 16107:2007 Workplace air quality
Protocol for evaluating the performance of
diffusive samplers.
13. ISO 16702:2007 Workplace air quality -
Determination of total organic isocyanate
groups in air using 1-(2-
methoxyphenyl)piperazine and liquid
chromatography.
14. ISO/DIS 17735 Workplace atmospheres --
Determination of total isocyanate groups in
air using 1-(9-anthracenylmethyl)piperazine
(MAP) reagent and liquid chromatography.
15. ISO/TR 17737 Workplace air quality
Guide for the selection of isocyanate
measuring methods.
16. ISO 16200-1:2001 Workplace air quality
Sampling and analysis of volatile organic
compounds by solvent desorption/capillary
gas chromatography Part 1: Pumped
sampling method.
17. ISO 16200-2:2000 Workplace air quality
Sampling and analysis of volatile organic
compounds by solvent desorption/capillary
gas chromatography Part 2: Diffusive
sampling method.
18. ISO 16017-1:2000 Workplace air quality
Sampling and analysis of volatile organic
compounds in ambient air, indoor air and
workplace air by sorbent tube/thermal
desorption/capillary gas chromatography
Part 1: Pumped sampling.
19. ISO 16017-2:2003 Workplace air quality
Sampling and analysis of volatile organic
compounds in ambient air, indoor air and
workplace air by sorbent tube/thermal
desorption/capillary gas chromatography
Part 2: Diffusive sampling.
20. ISO 9169:2006 Air quality - Definition and
determination of performance characteristics
of an automatic measuring system.
21. ISO 20988:2007 Air quality - Guidelines for
estimating measurement uncertainty.
22. ISO 16000-5:2007 Indoor air - Part 5: :
Sampling strategy for volatile organic
compounds (VOCs).
23. ISO 16000-2:2004 Indoor air Part 2:
Sampling strategy for formaldehyde.
24. ISO 16000-4:2004 Indoor air Part 4:
Determination of formaldehyde -- Diffusive
sampling method.
April 2008
Stainless steel meshes
coated with TEA
Teflon filter to stop
aerosols and stabilize
diffusion path
Coloured cap
accomodating
the meshes
Acrylic plastic tube
Cap to be
removed during
sampling and replaced
by a filter holder
Determination of NO
2
and SO
2
by ion chromatography in ambient air by use of
membrane closed Palmes tube
Daniela Buzica, Michel Gerboles, DG-JRC, Institute for Environment and Sustainability, Joint Research Centre Via E. Fermi, I-
1027 Ispra VA, Italy E-mail: daniela.buzica@jrc.it , michel.gerboles@jrc.it
Introduction
Several authors have investigated the diffusive sampler
method for measuring NO
2
and SO
2
in ambient air. The two
pollutants are either independently [1,2,3,4] or
simultaneously [5,6,7] analysed using the diffusive sampler.
Different types of diffusive samplers can be used e.g badge
type, radial type or open-ended longitudinal diffusion tubes.
The Palmes tube, coated with triethanolamine (TEA), allows
the simultaneous determination of NO
2
and SO
2
. However,
field measurements showed that the open-ended diffusion
tube is affected by a strong artefact on the SO
2
determination
arising from sulphate particulate matter. To avoid this
interference, a Teflon membrane has been introduced at the
open end of the Palmes tube able to prevent contamination
coming from the particulate matter. However the
introduction of the membrane creates an additive resistance
to the diffusion of molecules to the absorbent and therefore
modifies the uptake rate of the Palmes tube [8]. The
evaluation of NO
2
membrane-closed Palmes diffusion tubes
(MCPTs) is already presented elsewhere [9]. Hereafter, an
evaluation of the modified Palmes sampler with a membrane
for the determination of SO
2
is presented.
Figure 1 Description of a Palmes tube
Principle
NO
2
and SO
2
are collected by molecular diffusion along the
acrylic tube to the TEA where it is retained for subsequent
measurement. The first Ficks law describes diffusive
transport and allows after integration to determine the
airborne concentration using an equation [10] of the
following form:
t U
m m
C
b

= (1)
where: U sampling rate, ng ppb
-1
min
-1
;
m mass of the pollutant, ng;
m
b
mass of the pollutant in the blank, ng;
C concentration of the pollutant, ppb;
t time of exposure, minutes.
Materials and methods
The modified Palmes sampler was described by Gerboles et
al. [9]. The method of preparation of MCPTs is to clean
tubes (Gradko DIF100), membranes (XDIF500F) and caps
(XDIFCAP-001, XDIFCAP-003 and XDIFCAP-011) in an
orbital shaker using Millipore water and changing the water
every half hour for 3 hours. All samplers are then placed in


an oven, at 45
0
C until they are completely dry. The stainless
steel mesh discs (XDISC) are cleaned in an ultrasonic bath,
at 60
0
C for 5 hours, changing the water every half an hour.
Then, they are placed in an oven, flushed with nitrogen, at
125
0
C until they are completely dry. Three clean and dry
discs are placed in the coloured cap (see Figure 1). 40 l of a
10% v/v solution of TEA (Merck n8379) with 0.3 % of non
ionic detergent (Brij-35, Merck n 1.01894) in deionised
water is spread all over the mesh using a micropipette. A
tube is placed immediately on the coloured cap while the
other end is sealed immediately with a membrane for
immediate use. It is advised to check that the membrane is
correctly placed to make sure that the pollutants diffuse only
through the membrane.
NO
2
can be analysed using colorimetric method [8]. Even
though this method gives good results it is still time
consuming and since it is destructive it only allows
determination of one pollutant at a time. Conversely, using
ion chromatography (IC) [11], it is possible to
simultaneously determine both pollutants.
The tubes are analysed by adding 5 ml of MilliQ water
directly into the Palmes tube and then stirring them up with
an orbital shaker creating a strong vortex for 5 minutes. The
5 ml solution is then transferred into an IC cleaned vial and
10 l of 30 % hydrogen peroxide (H
2
O
2
) is added to ensure
complete oxidation of SO
3
2-
to SO
4
2-
. The vial is then closed
4 . . . . . . . . . . . . . . . . . . . . . . . .
The Diffusive Monitor, 16
5 . . . . . . . . . . . . . . . . . . . . . . . .

with a cleaned Dionex cap. The ion chromatographic
analyses were performed using a Dionex ICS 1000 with
suppressed conductivity detection and an auto-sampler
(Dionex AS 40). The chromatographic system consisted of a
column IonPac AS12A (2 mm x 250 mm) with
carbonate/bicarbonate isocratic elution (flow of 0.38
ml/min), an injection loop of 25 l and an anion self
regenerating suppressor AMMS2 mm with sulphuric acid.
For NO
2
and SO
2
, a 6-point calibration, in the range of 0 to
0.3 g ml
-1
was plotted prior to the analysis. Sulphate and
nitrite were identified by retention time and quantified by
peak area. Although the sample vials and caps are washed
by the producer before packing, it is recommended that they
are rinsed with deionized water before filling to remove any
traces of dust (see [12]). This is also necessary in order to
limit the value and variability of the blank. It is important
that once H
2
O
2
is injected in the vial, the IC analysis is
performed as fast as possible as oxidation of nitrite to nitrate
generally takes place in the vial. The variability of the 5-ml
of MilliQ water injected in the tubes was controlled to a
relative standard deviation (RSD) of 0.1 %. It is
recommended that the analyst wears gloves.
Table 1 Blank values stored under different conditions and limit of detection
of SO
2
membrane closed Palmes tube.

Laboratory experiments
NO
2

The study of Gerboles et al. [9] focused on NO
2

determination using the membrane closed Palmes tube. The
limit of detection was calculated as the three times standard
deviation of the blank tubes and was found to be 1.4 g m
-3

week.
The uptake rate (U) in ng ppb
-1
min
-1
is calculated using the
following equation 2, where RH is the relative humidity in
%, T is the temperature in C, w is the wind speed in m s
-1
:
( )
( )
2
10 . 64 . 0 10 18 . 3 28 . 1 00130 . 0 86 . 0 134
2
10 18 . 3 28 . 1 00130 . 0 86 . 0 134
5 10 2
5
t
m
w T t RH
w T t RH
U

+ + +
+
+ + +
=
(2)

SO
2

In order to evaluate the limit of detection for SO
2
, blank
tubes were kept in different conditions (field, room and
fridge) for different periods of time (Table 1). All this was
performed during a series of campaigns [13] which took
place in Martigue (F) between June 2001 and May 2002.
Observing the results for SO
2
, one notices that laboratory
blanks stored in fridge are much lower than blanks stored in
field and at room (in the dark) conditions. Fridge blanks can
be a good indicator of how well prepared the Palmes tubes
are. Even though the blanks samplers stored under field
conditions have a higher value, they are considered the most
characteristic blank values and it is recommended to use the
field blanks of each measurement campaign.
The limit of detection (LoD) refers to the lowest amount of
analyte that is detectable using the method and is determined
by using the analysis of blanks MCPT. The LoD is
calculated as 3 times the standard deviation of the blank
tubes.
The nominal uptake rate for SO
2
for membrane-closed
Palmes tube was determined by exposing the samplers to
different conditions of concentration, relative humidity,
temperature, wind velocity and duration. The results are
presented in Table 2. The averaged uptake rate was 0.00214
ng ppb
-1
min
-1
for the first four experiments in which SO
2
,
temperature and relative humidity were kept constant. The
uptake rate was found independent of the exposure time
between 3 days and 2 weeks. In the 5
th
experiment which
was similar to the first one except that NO
2
was added, the
uptake rate was found to be 0.00213 ng ppb
-1
min
-1
. This
value confirms the uptake rate of the first experiments and
does not show any evidence of NO
2
interference on the
measurement of SO
2
with the MCPTs. Finally, the last two
experiments, at high relative humidity, low temperature and
SO
2
concentration gave an uptake rate of approximately
0.002075 ng ppb
-1
min
-1
which confirmed within 3 % that an

No of days field, g room, g fridge, g
63 0.022 (n=6) 0.024 (n=9) 0.015 (n=9)
62 0.010 (n=4) 0.045 (n=10) 0.013 (n=10)
50 0.033 (n=4) 0.041 (n=10) 0.017 (n=10)
51 0.063 (n=4) 0.087 (n=9) 0.016 (n=10)
average, g 0.031 (n=18) 0.032 (n=38) 0.015 (n=39)
st.dev, g 0.026 (n=18) 0.021 (n=38) 0.009 (n=39)
LoD,
g m
-3
week
-1
9.3 7.2 3.3
Table 2 Uptake rate in the chamber according to the conditions of
exposure. The quoted values are standard deviations.

Test SO
2
ppb
NO
2
ppb
RH
%
T
C
w,
m.s
-1

Time,
days
Uptake rate,
ng.ppb
-1
min
-1

1 51.5 - 38 24 0.2 3 0.00210 0.00013
(n=15)
2 50.5 - 45 23 0.2 7 0.00203 0.00006
(n=5)
3 50.5 - 45 23 0.2 10 0.00224 0.00014
(n=4)
4 50.5 - 42 23 0.2 14 0.00216 0.00009
(n=4)
5 48.4 41.4 45 22 0.3 3 0.00213 0.00014
(n=5)
6 22.9 - 80 15 0.5 3 0.00208 0.00011
(n=6)
7 18.2 - 80 15 0.5 14 0.00207 0.00005
(n=7)
6 . . . . . . . . . . . . . . . . . . . . . . . .
April 2008
average value of 0.00214 ng ppb
-1
min
-1
can be used for the
SO
2
uptake rate.
Field experiments
In the framework of the AIRPECO project [14], 94
duplicate pairs of membrane-closed Palmes tubes were
exposed during a measuring campaign over the city of
Ljubljana (Slovenia) in February 2004. The samplers were
installed in protective boxes placed on lamp-posts at a height
of
plers, the NO
2
repeatability was 5.5 g m
-3
.
PTs are installed in the
e reference methods.
References
of 3 m.
The repeatability based on the duplicates has been evaluated
for both NO
2
and SO
2
. For NO
2
, the repeatability was 9 g
m
-3
(standard deviation of 3.2 g m
-3
) while for SO
2
it was 8
g m
-3
(standard deviation of 2.7 g m
-3
). During another
measuring campaign in June 2004 including 10 pairs
duplicate sam
Conclusion
The work reported here highlights the possiblity of
determining both NO
2
and SO
2
in one run by ion
chromatography using the MCPT. While for NO
2
the uptake
rate is already known, the uptake rate for SO
2
was found to
be independent of the conditions of exposure with a value of
0.00214 ng ppb
-1
min
-1
. However, it is necessary to perform
some field tests in which the MC
vicinity of th

[1] Lin, J.M., Lin, T.S., A diffusive sampler for the ion-chromatog
measurement of sulfur dioxide in ambient air, Toxicol
raphic
g
ry
er,

oekens, E., Keppens, V., Laboratory and field validation of a combined

ogical and
Environmental Chemistry 39 (3-4), 1993, 229 236.
[2] Tang, H., Brassard, B., Brassard, R., Peake, E., A new passive samplin
system for monitoring SO2 in the atmosphere, Field analytical chemist
and technology, 1 (5), 1997, 307 314.
[3] Cruz, L.P.S., Campos, V.P., Silva A.M.C., Tavares, T.M., A field
evaluation of a SO2 passive sampler in tropical industrial and urban air,
Atmospheric Environment 38, 2004, 6425 6429.
[4] Buzica, D., Gerboles, M., Amantini, L., Prez Ballesta, P., De Saeg
E., Modelling of the uptake rate of nitrogen dioxide Palmes diffusive
sampler based on the effect of environmental parameters, Journal of
Environmental Monitoring, 2005, 7, 169 174.
[5] Plaisance, H., Sagnier, I., Saison, J.Y., Galloo, J.C, Guillermo, R.,
Performances and application of a passive sampling method for
simultaneous determination of nitrogen dioxide and sulfur dioxide in
ambient air. Environmental Monitoring and Assessment 79, 2002, 301
315.
[6] Kasper Giebl, A., Puxbaum, H., Deposition of particulate matter in
diffusion tube samplers for the determination of NO2 and SO2, Technical
Note, Atmospheric Environment 33, 1999, 1323 1326.
[7] Swaans, W., Goelen, E., De Fr, R., Damen, E., Van Avermaet, P.,
R


NO2 SO2 Radiello passive sampler, Journal of Environmental
Monitoring, 2007, 9, 1231 1240.
[8] Passive Samplers for Nitrogen Dioxide, Agence de lEnvironnement
et de la matrise de lEnergie, ADEME ditions, Rf. : 4414, Paris, 2002.
[9] Gerboles, M., Buzica, D., Amantini, L., Modification of the Palmes
diffusion tube and semi-empirical modelling of the uptake rate for
monitoring nitrogen dioxide, Atmospheric Environment 39, 2005, 2579
2592.
[10] European Committee for Standardization, Ambient air quality
Diffusive samplers for the determination of concentrations of gases and
vapours. Requirements and test methods, EN 13528:2002.
[11] Miller, D.P., Ion chromatographic analysis of Palmes tubes for nitrite,
Atmospheric Environment 1984, 18, 891-892.
[12] Dionex, P/N 053891-16B, AS40 Automated sampler operators
manual
[13] Buzica, D., Gerboles, M., Evaluation of the Palmes tube sampler with
membrane for the simultaneous determination of nitrogen dioxide and
sulphur dioxide. Measurement campaign in the industrial area of Martigue
(F), 2002, Technical Note No. I.02.110.
[14] Field, R.A., Gerboles, M., Perez-Ballesta, P., Nikolova, I., Baeza-
Caracena, A., Buzica, D., Connolly, R., Cao, N., Amantini, L., Lagler, F.,
Stilianakis, N., Forcina, V., De Saeger, E., Air quality, Human exposure
and Health impact assessment of air pollution in Ljubljana, Slovenia, 2005,
EUR 21649.



The Diffusive Monitor, 16
7 . . . . . . . . . . . . . . . . . . . . . . . .

Tube-tagging Enhanced tracking of sample and tube-related information for thermal
desorption
Liz Woolfenden, Markes International Ltd, Llantrisant, UK, enquiries@markes.com
Wishful thinking?
Wouldnt it be good if there was a fail-safe way of linking
field sampling information with the relevant sorbent tube
without relying on a chain of different people reading and
recording the individual number etched onto each tube
without making any mistakes?
Wouldnt it also be good if TD users and field sampling
personnel could immediately identify the sorbent-
combination in the 6 tubes that have been rolling around on
the bench all week and when they were last used?
Wouldnt it be even better if there was a way of
automatically tracking a thermal desorption tube throughout
its entire life recording what it is packed with, when it was
packed, how many times its been used and all the details
associated with its performance?
This paper describes one possible answer to some of these
questions
Historically, associating information with thermal desorption
(TD) tubes has relied on manually reading and recording of
tube serial numbers. Bar code technology has proved
difficult to apply to TD tubes because the high temperatures
required limit the lifetime of bar code labels. Bar codes
etched onto curved tube surfaces also get increasingly
difficult to read electronically - especially after extensive
handling. Another limitation of bar codes is that they cant
be programmed to record sample or tube specific
information.
A new RFID-tag based technology has recently been
introduced for TD tubes which could overcome some of
these limitations and offer a real step forward in sample
tracking and analytical quality control for TD-GC(MS)
users. The tags are re-usable, read/write programmable
RFID devices which can be attached to standard sorbent
tubes (metal or glass) and may be applied in two ways:
Transit tagging used for tracking samples within a lab
and in transit between lab and field during air
monitoring projects. Available to all TD users
Tube tagging used both for sample tracking and to
monitor the history of each individual sample tube
throughout its life. Requires tag-compatible
instrumentation.
The two modes of operation are illustrated in Figure 1.
Background
There are significant challenges in developing re-usable
RFID tag technology for TD tubes not least the high
temperatures required for analysis. RFID tags are destroyed
at temperatures above 140C and the associated read/write
devices dont work through the metal walls of most tubes.
Tags must also be unobtrusive, resistant to environmental
factors humidity, high particulate levels, etc. and still
allow a tube to be capped for long-term storage.
Developed by Markes, the new RFID TubeTAGs* seem to
have overcome these difficulties and provide a robust,
permanent and programmable tube labelling solution. They
attach to the non-sampling ends of ordinary -inch (6.4
mm) or 6 mm O.D. TD tubes and comprise a compact
RFID-chip assembly mounted on a special tube clip. The
RFID chip itself is embedded in a protective, high
temperature, low-emission polymer to reduce the effect of
temperature and protect it from environmental factors see
figure 2. The clips are designed such that tube tags cant be
attached to or removed from a TD tube without using a
special tool.
* Patent number: US 6,446,515 B2

Data entry in field

Write project-specific
information to tag in
the lab
Write tube specific
data to tag in the
laboratory








Figure 1 Using RFID tags for transit tagging and tube tagging.
Clear tube and sample data. Tag removed from
tube for re-use on another tube
Write sample start &
end information to
tag in the field
Read tube and sample
information from tag in
the laboratory
Tube-history updated.
Sample specific data cleared.
Tagged tube ready for re-use
Transit-tagging
Tube-tagging
April 2008
8 . . . . . . . . . . . . . . . . . . . . . . . .
Tube tags in operation
When used for transit tagging, individual RFID tags are
attached to every tube in a batch and programmed with
relevant information tube ID number, sorbent type, project
code, etc. prior to dispatch. Once that batch of tagged
tubes reaches the field, additional sample collection details
such as monitoring location, sampling method, sampling
start- and end-times, etc. can be entered onto each tag. When
the tubes are received back into the laboratory after
monitoring, all of the stored information can be readily
downloaded from the tags and entered in the laboratorys
information management system. The tags are then removed
from the sampled tubes using the special tool as they are
placed into the automated thermal desorber for analysis.
Figure 2 Close up of tag assemblies and tagged tubes

From the moment a tube is tagged and programmed prior to
dispatch, in the relative calm of the lab environment (!!), no
manual re-entry of tube ID #, sorbent packing or project
number is required. Write-access to primary fields like these
can be disabled by the system administrator if required.
Subsequent reading and entering of other information onto
the tags in the field, for example monitoring location,
sampling method and sample start & end times, simply
allow multiple opportunities for users to confirm the tube ID
number programmed into the tag before dispatch.
Tags that have been removed from a batch of tubes just
before analysis, can be cleared of information relating to the
last monitoring exercise and re-applied to the next batch of
tubes going out for field sampling. Relevant new tube ID
numbers, sorbent details and project information can be
entered onto the tags by the system administrator and the
whole cycle repeated. In this way, one RFID tag can be
shared between several sampling tubes and costs can be kept
down to around 25 cents per tube per monitoring cycle.
When used for tube tagging, a given tag is linked to a
specific sorbent tube throughout its life or at least until that
tube is re-packed (typically 200 or more sampling/analysis
cycles.) This allows the history of that tube to be recorded
and tracked. In this case, a tag is assigned to a tube as soon
as it has been packed and conditioned and the tube ID
number, date of packing and combination of sorbents are
entered only once. Each time a permanently-tagged tube is
about to be sent to the field, project information can be
entered onto the tag in the lab before dispatch. As described
above, sampling information (pumped or diffusive, flow rate
/duration of exposure, start time, etc) can then be entered
onto the tag in the field. An example of the type of tube and
sample data that can be recorded is shown in Figure 3 and a
typical field-portable tag read/write system is shown in
Figure 4.
Operation in tube-tagging mode requires the use of tag-
compatible TD instrumentation. Once the tagged tubes are
received back into the laboratory they are placed into the
tag-compatible automated TD system (see example in figure
5) which automatically reads the recorded tube and enters
the relevant sample information into the automation
sequence. Post run, the desorber can also write to the tags
incrementing the number of thermal cycles, changing tube
status (e.g. from sampled to desorbed) and clearing the
sample collection information. Analytical anomalies such as
leak test failures or unusually high back pressure can also be
recorded on the tag if required.
Tags used in tube-tag mode i.e. permanently attached to
the same tube also last indefinitely. Tests have shown
them to be compatible with over a thousand thermal cycles
even under extreme desorption conditions e.g. 400C for 30
minutes. As above, this means that tagging costs are
minimal Less than $0.25 per thermal cycle.
Figure 3 Tube and sample parameters recorded on tube tags

Tube conditioning in tube tag mode
The process of desorbing TD tubes is usually sufficient to
condition them. In other words, no additional cleaning is
necessary in most cases and analysed tubes can be re-used
straight away. However, there are instances where
additional, post-analysis conditioning is recommended for
example if tubes have been stored for extended periods (> 30
days) or if the specific monitoring protocol requires the

The Diffusive Monitor, 16
9 . . . . . . . . . . . . . . . . . . . . . . . .

confirmation of tube blank levels before they can be used for
field sampling.

Figure 4 Example of portable device for programming tags
in the field or laboratory.

If additional tube cleaning is required it can be carried out
either using the TD-GC(MS) system or by using separate
off-line multi-tube conditioning rigs. The advantages of
using tag-compatible TD analytical equipment for tube
conditioning is that the number of thermal cycles can be
automatically incremented and that a blank profile can be
obtained automatically as part of the conditioning process.
However, if ever / whenever multi-tube off-line equipment
is preferred for cost effective conditioning of an entire batch
of tubes, tags can be readily removed from the tubes using
the special tool and re-attached to the same tubes post-
conditioning. The number of thermal cycles can be manually
incremented as each tag is re-attached to its specific
conditioned tube.
Data output and information storage
Users of tag-ready TD instrumentation record the status of
every tagged tube whenever that tube is desorbed allowing
the information to be recorded as part of the sequence report.
Moreover, a comma separated variable (CSV) file is created
every time a tube tag is read or written to whether using
the field portable tag-scribe device or via the desorber. This
allows all tube- and sample-related data relevant to that tube
to be simply and easily entered into a database and accessed
as and when required. Subsequent interrogation of that
database could then be used to determine for example; when
that tube or batch of tubes needs repacking or whether one
or more tubes have a history of leak test failures.
Summary
RFID tube tags such as those described have the potential to
greatly enhance the analytical quality assurance of air
monitoring studies and TD-GC(MS) applications generally.
This is only the start. Future developments should allow
tube-tags to be linked to TD methods allowing the analytical
system to generate its own automatic sequence for tubes
loaded randomly into it. Tube tags also offer the potential
for intelligent interaction with GC(MS) data processing
systems. In the future, this should allow key analytical
factors such as background levels or key artifacts to be
linked with specific tubes and tracked over the lifetime of
the tube.
Figure 5 Example of tag-compatible autosampler.



10 . . . . . . . . . . . . . . . . . . . . . . . .
April 2008
Controlling GC carrier gas flow rate through a thermal desorption system transfer line
Andrew Tipler
Senior Scientist, GC Applications and Technology Group, PerkinElmer LAS, 710 Bridgeport Avenue, Shelton, Connecticut,
06611, USA.
1. Introduction
Thermal desorption has become a popular technique for the
extraction, concentration and injection of sample vapors
collected onto an adsorbent tube into a gas chromatograph
for separation, identification and quantification. Figures 1
and 2 illustrate the main steps involved in a typical 2-stage
thermal desorption analysis.
GC
Detector
Optional
inlet split Desorb flow
Cooled
trap
Carrier gas in
Heated
sample
tube
Anal ytical column
GC
Detector
Optional
inlet split Desorb flow
Cooled
trap
Carrier gas in
Heated
sample
tube
Anal ytical column

Figure 1 First step in a 2-stage thermal desorption analysis
the primary (tube) desorption.

Carrier gas in
GC
Detector
Anal ytical col umn
Optional outlet split
heated trap
Carrier gas in
GC
Detector
Anal ytical col umn
Optional outlet split
heated trap

Figure 2 Second step in a 2-stage thermal desorption
analysis the secondary (trap) desorption.

As can be seen in Figures 1 and 2, the sequence of
operations involved in extracting and transferring the sample
vapors into the column may have a dramatic effect on the
gas about to enter the GC column there may be significant
changes in temperature, gas flows and gas pressures.
Throughout this whole process, we are trying to regulate the
flow of carrier gas along the GC column from the thermal
desorption system.
The flow rate of carrier gas through the column is also
significantly affected by the column temperature as the gas
increases in temperature, it will become more viscous and, if
we are using a pressure controller to supply the gas, the flow
rate through the column will drop.
Modern GCs employ electronic systems to regulate carrier
gas supplies and users are now very familiar with concept of
constant carrier gas flow control through the GC column.
Such systems will provide better column efficiency and will
eliminate changes in response or background in a flow-
sensitive detector such as a mass spectrometer.
This article describes systems and algorithms specifically
developed to overcome effects on the carrier gas just
described and to provide a constant flow rate of gas
controlled from the thermal desorption system, through a
transfer line, through a column and into a detector.
2. Project requirements
This project was initiated to develop electronic
programmable pneumatic control (PPC) systems capable of
providing a level of flow control and performance not
possible with manual pneumatics systems. However, it was
also important that the stability and flexibility in using the
manual pneumatic systems was not lost. Table 1 lists some
of the key requirements for this project.

Table 1 Key requirements for PPC systems on a thermal
desorption system.
All control should come from the thermal desorption system
There should be no need for any additional external hardware (e.g.
injectors or pneumatic controllers on the GC)
It should work with any GC (this implies the use of a flexible transfer
line)
There should be no transcription of column temperature programs
between the GC and the ATD
There is no need for the GC and ATD to communicate digitally with
each other
The flow rate will always track the current column temperature
automatically without any additional input
The system will provide stable and precise operation over a wide
range of flows and pressures


3. Electronic carrier gas control from a GC
To set the flow rate of gas through a GC column, we dont
normally control the flow rate directly. To precisely control
a flow rate of 1mL/min through a capillary column is not
easy and we may wish to open split vents that will increase
the required flow rate by a factor of over 500 times. Also the
slightest leak will represent a very significant lost portion of
the gas that should be flowing through the column.
For these reasons, we normally apply the gas pressure that is
expected to deliver a required flow rate through the column.
This approach makes the flow control through the column
largely insensitive to changes in split flow rates and leaks.
The Diffusive Monitor, 16
11 . . . . . . . . . . .
. . . . . . . . . . . . .

As a GC column is heated, the viscosity of the carrier gas
flowing through it increases. In such cases the flow rate
through the column will decrease with increasing
temperature. For most applications this will not have a
detrimental effect on analytical results but with others, for
instance when a mass spectrometer is being used for
detection, the changing flow rate may have a dramatic effect
on detector performance.
Most modern gas chromatographs are equipped with
electronic programmable pneumatic controls (PPC). These
are able to compensate for the changing viscosity during a
temperature program by increasing the inlet pressure at a
rate calculated to maintain a constant flow rate through the
column. To maintain a constant flow rate, the controlling
system must have knowledge of the column temperature at
all times and be able to calculate the gas viscosity at that
temperature and make the appropriate adjustments to the
applied pressure. The viscosity versus temperature
relationships are well documented for all the popular carrier
gases used in GC and the Hagen-Poiseuille relationship
given in Equation 1 is used by the GC control systems to
perform these calculations.

( )

=
o
o i c
o
P L
P P d
F
256
2 2 4
Equation 1.

Where:
F
o
is the flow rate at the column outlet
d
c
is the internal diameter of the column
L is the length of the column
P
i
is the carrier gas pressure at the column inlet
P
o
is the carrier gas pressure at the column outlet
is the viscosity of the carrier gas at the column
temperature


With a given column that is temperature programmed under
isobaric conditions, the only variable that will alter will be
the viscosity, . Inspection of Equation 1 indicates that, as
the viscosity increases an appropriate increase in the inlet
pressure, P
i
, can be applied to keep the column outlet flow
rate, F
o
, at a constant setting.
The oven temperature is known because it is controlled by
the GC. The viscosity of the carrier gas can be derived from
this temperature. If the column dimensions are entered into
the system, then a specific flow rate may be controlled using
Equation 1.
4. Electronic carrier gas flow control through a transfer
line from a thermal desorption system
The situation becomes more complicated when the carrier
gas pressure is controlled on a system remote to the GC such
as a thermal desorption system.
Figure 3 summarizes the effects on the carrier shown in
Figures 1 and 2 between where it exits the pressure regulator
on the thermal desorption system and where it enters the
column in the GC.

A. Regulator connected via transfer l ine to column
B. Regulator connected via trap and transfer li ne to column
C. Regulator connected via trap, split and transfer line to column
pressure
regulator
secondary
trap
split
transfer
line
column
B
B
B
detector
A
A
A
ATD GC

Figure 3 The various routes carrier gas can take between
the pressure regulator on the thermal desorption system and
the column inlet on the GC.

Both the secondary trap and the transfer line represent
restrictions to gas flow and so the gas pressure delivered to
the column inlet indicated by point [B] will be less than the
pressure set by the pressure regulator at point [A].
Furthermore, with the trap inline, the pressure drop across it
will increase as the trap temperature increases (the gas
viscosity increases with temperature) and as the split flow
rate increases. The pressure drop will also change across the
transfer line if its temperature or the temperature of the GC
column is changed.
All this leads to the fact that the classic flow control
equation given in Equation1 cannot be used in this situation
and some other approach must be used.
For the system to be effective, we must be able to control the
pressure at point [B] shown in Figure 3. One significant
difference between a PPC system and a mechanical pressure
regulator is that the pressure sensing device may be remote
from the control valve. Figure 4 shows how a distributed
pressure control system could be applied to the worst-case
scenario shown in Figure 3C.


. . . . . . . . . . .
. . . . . . . . . . . . .
April 2008
12
C. Pressure regul ated at col umn i nlet
B
A
C T
V
A. Pressure regul ated pri or to trap
B
A
C T
V
B. Pressure regul ated between trap and transfer li ne
B
A
C T
V

Figure 4 Distributed PPC systems. (T = pressure
transducer, C = control system and V = control valve).

Figure 4A shows the PPC equivalent of a mechanical
pressure regulator. The control system (C) adjusts the
control valve (V) until the required pressure is seen at the
pressure transducer (T). This configuration would perform in
a very similar manner to a mechanical pressure regulator.
Figure 4C looks as if it would provide the ideal solution
the pressure would be regulated directly at the column inlet.
This means that Equation 1 could be used to provide carrier
gas flow control capabilities through the GC column.
Unfortunately this configuration also provides some
practical challenges. The first of these is that the pressure
transducer would need to be mounted on the GC this then
makes the installation instrument specific. The major
problem, however, would be that the transducer would now
be very remote from the control valve and so the time
constant of such a system would be very long which could
lead to instability (oscillation) or poor response to changes
in upstream impedance or flows.
A variant on the configuration given in Figure 4C is to use
an additional (and independent) pressure regulator on the
GC which would be connected to a T-piece or inlet system
at the interface between the transfer line and the column.
This was not considered as it would require additional
hardware and expense and would restrict the choice of GCs
that could be used. The additional carrier gas would also
dilute the sample entering the column.
The best solution appears to lie with Figure 4B. Tight and
stable control of the gas pressure as it enters the transfer line
is achievable and the system responds well as the trap is
brought in and out of the gas stream and changes are made
to the trap temperature and split flow rate. Also, all the
controlling hardware is now mounted within the thermal
desorption system facilitating operation with any GC.

The main issue with Figure 4B is that we still cannot use
Equation 1 to control the flow rate of carrier gas through the
GC column we still have a transfer line of (usually)
different temperature and geometry to pass through first.
This matter is resolved by regarding the transfer line and the
GC column as being two columns in series as shown in
Figure 5.

P P
i i
P P
x x
P P
o o
Transfer line Transfer line GC Column GC Column
( )
x t t
x i t
t
P L
P P d
F


=

256
2 2 4
( )
x c c
o x c
i
P L
P P d
F


=

256
2 2 4
F F
i i
F F
t t
F F
o o
P P
i i
P P
x x
P P
o o
Transfer line Transfer line GC Column GC Column
( )
x t t
x i t
t
P L
P P d
F


=

256
2 2 4
( )
x c c
o x c
i
P L
P P d
F


=

256
2 2 4
F F
i i
F F
t t
F F
o o

Figure 5 Viewing the transfer line and GC column as
serially connected columns.

In Figure 5, we have associated a different form of Equation
1 to each of the two columns .

Where:
F
t
is the flow rate at the transfer line outlet
d
t
is the internal diameter of the transfer line
L
t
is the length of the transfer line
P
i
is the carrier gas pressure at the transfer line inlet
P
x
is the carrier gas pressure at the transfer line outlet
and GC column inlet

t
is the viscosity of the carrier gas at the transfer line
temperature

F
i
is the flow rate at the GC column inlet
d
c
is the internal diameter of the GC column
L
c
is the length of the GC column
P
o
is the carrier gas pressure at the GC column outlet

c
is the viscosity of the carrier gas at the GC column
temperature


Because the flow rate out of the transfer line, F
t
, and the
flow rate into the GC column, F
i
, will be the same (once
corrected for temperature), the two equations shown in
Figure 5 may be solved simultaneously to produce a
relationship between the outlet flow, F
o
, and the applied
conditions to both the transfer line and the GC column.
Equation 2 gives the final relationship to describe how the
column output flow rate may be related to the applied
conditions applied to a column and transfer line of differing
temperatures and geometries connected in series.


The Diffusive Monitor, 16
13 . . . . . . . . . . .
. . . . . . . . . . . . .

( )

=
4 4
2 2
256
c
c c c
t
t t t
o i
o
c
o
d
L T
d
L T
p p
p
T
F

----- Equation 2

Where:
F
o
is the flow rate at the column outlet
T
t
is the transfer line absolute temperature
T
c
is the column absolute temperature

This approach needs information on the geometry of both
the column and the transfer line. This is easily addressed on
the thermal desorption instrument by the user entering both
their geometries as inputs into the control system.
The temperature of the transfer line and the applied pressure
are known as they are controlled from the thermal
desorption system so the only parameter not known is the
temperature of the GC column in the GC oven. To address
this need, the transfer line has a thermocouple threaded
through it as shown in Figures 6 and 7.

Gas Chromatograph Thermal Desorber
Detector
GC
column
GC oven
Heated
transfer line
tubing
PPC
pressure
regulator
Signal cable
from
thermocouple
Thermocouple
T4
Gas Chromatograph Thermal Desorber
Detector
GC
column
GC oven
Heated
transfer line
tubing
PPC
pressure
regulator
Signal cable
from
thermocouple
Thermocouple
T4 T4

Figure 6 Using a thermocouple inside the transfer line to
monitor GC column temperature.

The temperature sensor may be calibrated using the GC
column oven at one or more set-points to enable either a
single point or multi-point temperature calibration.

Thermocouple
GC Column
Thermocouple
GC Column

Figure 7 Thermocouple positioned against GC column
inside GC oven.

5. Examples of PPC system operation
To evaluate the efficacy of the new control algorithm, a
series of tests was conducted using helium carrier gas doped
with ~0.5% of methane using the apparatus shown in Figure
8. This small concentration of methane was not expected to
change the behavior of the helium during these experiments.

Methane
Helium
PR1
PR2
MFC1
MFC2
BPR1
(100psig)
2mL/min
400mL/min
Transfer
line
GC
Flame
Ionization
Detector
ATD
P P
Methane
Helium
PR1
PR2
MFC1
MFC2
BPR1
(100psig)
2mL/min
400mL/min
Transfer
line
GC
Flame
Ionization
Detector
ATD
P P

Figure 8 System for delivering a stream of helium carrier
gas doped with 0.5% methane to a thermal desorption
system.

The back-pressure regulator (BPR1) in Figure 8 ensures that
the upstream flows were unaffected by the gas demands on
the thermal desorption system (for example, as split vents
were opened) and so gas with a constant composition was
consistently applied to the instrument. The flame ionization
detector is a very linear mass-flow sensitive detector and is
very sensitive to methane. Consequently, the output signal
from the detector was directly proportional to the flow of
doped carrier gas flowing through it as shown by the
calibration plot given in Figure 9.

. . . . . . . . . . .
. . . . . . . . . . . . .
April 2008
14
0
100
200
300
400
500
600
700
800
900
1000
0 2 4 6
Measured Flow Rate (mL/min)
F
I
D

S
i
g
n
a
l

(
m
V
)
8
r
2
=0.9996
0
100
200
300
400
500
600
700
800
900
1000
0 2 4 6
Measured Flow Rate (mL/min)
F
I
D

S
i
g
n
a
l

(
m
V
)
8
r
2
=0.9996
0
100
200
300
400
500
600
700
800
900
1000
0 2 4 6
Measured Flow Rate (mL/min)
F
I
D

S
i
g
n
a
l

(
m
V
)
8
r
2
=0.9996

Figure 9 Calibration plot of FID output signal versus flow
rate of methane doped helium carrier gas produced using the
apparatus shown in Figure 8.

1mL/min Flow Control
7.2psig Pressure Control
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
min
mV
1mL/min Flow Control
7.2psig Pressure Control
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
min
mV

Figure 10 Comparison between constant flow and pressure
controls on a 15m x 0.25mm column programmed from
40C for 1 minute, then 10C/min to 300C and held for 10
minutes. The transfer line was 1.8m x 0.28mm and held at
300C. The set flow rate of 1mL/min of the doped helium
had an initial pressure of 7.2 psig this was used for the
constant pressure test. The test system was a PerkinElmer
TurboMatrix 650 ATD and Clarus 500 GC.


This method of measuring gas flow is particularly suited to
this experiment as it allows the low flow rate of gas exiting
from the column to be measured directly and under the
conditions used for chromatography.

Figures 10 and 11 show comparisons between constant flow
control using the new PPC algorithm and constant pressure
control for two temperature programs.

1mL/min Flow Control
7.2psig Pressure Control
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0
min
mV
1mL/min Flow Control
7.2psig Pressure Control
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0
min
mV

Figure 11 Same conditions as Figure 10 but with a
programming rate of 40C/min.

As can be seen from Figures 10 and 11, the new PPC control
algorithm gave a very acceptable performance in the
constant flow control mode. The deviation was less than 2%
throughout the whole temperature range applied to the
column oven.

6. Conclusions
A mathematical function has been developed that describes
the relationship between applied pressure and outlet flow
rate from a GC column connected to a thermal desorption
system via a transfer line.
This function has been integrated into a programmable
pneumatic control system to provide the effective control of
a set flow rate of carrier gas through a GC column.
The PPC hardware has been implemented in such a way so
that changes in trap impedance or split flow rate do not
perturb the PPC control stability.
The system developed should be applicable to a wide range
of analytical methods and should function with any GC.

The Diffusive Monitor, 16
15 . . . . . . . . . . .
. . . . . . . . . . . . .

Diffusive uptake rates of aromatic hydrocarbons on Carbograph 1TD at workplace
concentrations using a thermal desorption tube sampler
Neil Plant, Glen McConnachie, Kate Shrivastava and Mike Wright
Health and Safety Laboratory, Harpur Hill, Buxton, SK17 9JN UK, neil.plant@hsl.gov.uk

Introduction
Graphitized carbons suitable for thermal desorption from
tube samplers have been available for many years. HSL and
others have proved their utility in the diffusive sampling of
ambient air for up to four weeks. However, for diffusive
tube sampling in the workplace over 0.5 8 hours there is a
shortage of published data for graphitized carbons. What
follows is an account of diffusive sampling/thermal
desorption methods applied to workplace air. Different
choices and compromises will apply to ambient air
sampling. Workplace validations, such as those published in
MDHS 80 [1], were mostly of porous polymers that were
consistent from batch to batch and which did not generally
catalyse thermal decomposition. The earliest carbonaceous
sorbents used in thermal desorption were of variable quality
and unsuitable for semi-volatile or thermally labile
substances. This was not surprising since they were not
primarily intended for sample recovery by thermal
desorption. The situation improved when chromatography
suppliers started to make carbon sorbents with a variety of
closely controlled properties. By this time so much work had
gone into measuring diffusive uptake rates on porous
polymers that there was little enthusiasm for duplicating the
validations. We will briefly mention the pros and cons of
polymers versus carbons here. Out of one sorbent study
during 1994-95 came the nomination of Chromosorb 106 as
the best compromise when compared with graphitized
carbon and carbon molecular sieves [2]. The test substances
were selected for a wide range of properties and possibly the
outcome of a repeated study with more sorbents using the
original criteria would have been the same. Nevertheless, a
medium strength graphitized carbon is a good choice for
compliance monitoring of substances with low limit values
or in diffusive sampling for short exposure times. Artefact
levels are much lower than those of porous polymers. HSL
originally chose Carbograph 1TD for thermal desorption
because its performance with thermally labile substances
was better than some other carbon sorbents. Carbopack B is
very similar for sorbent strength and wherever historical
diffusive uptake rates exist for both sorbents there seems to
be no significant difference. Stability of aromatic
hydrocarbons at high temperatures is not at issue here.
However, we have had more general experience with
Carbograph 1TD than Carbopack B.

Experimental
Test atmospheres of benzene, toluene, m-xylene and 1,3,5-
trimethylbenzene (TMB) as mixed vapour (1-100 ppm each
component) were generated by a syringe infusion pump
(Harvard model 22) injecting at a known rate into 30 l/min
(nominal) dilution air via a heated block and glass exposure
chamber. The reference concentrations and uptake rates
were determined by active and diffusive sampling on
stainless steel tubes, 89 mm x 6.4 mm od., 5.0 mm id.,
packed with 300 mg Carbograph 1TD (Markes International
Ltd), followed by thermal desorption and gas
chromatography with flame ionization detection (Markes
Unity/Agilent 6890 and PerkinElmer Turbomatrix 650/PE
Clarus). Within the exposure chamber diffusive tubes were
mounted on a plate rotating at about 80 rpm. The effective
air velocity experienced by the diffusive tubes was estimated
at about 50 cm/s. For determining reference concentrations
active sampling conditions were 20 ml/min for 30 minutes.
A number of sequential active samples (up to 12) were taken
to cover exposure periods up to 480 minutes. Flow rates
were controlled to within 0.5 % by electronic mass flow
devices (Brooks 5850S, 0-100 ml/min) traceable by
calibration to national standards. Equilibration time and
stability of the test atmosphere was recorded by a total
hydrocarbon monitor (3000HM, Signal Instruments).
Calibration of the gas chromatographs was by liquid spiking
of thermal desorption tubes from methanol solutions
prepared gravimetrically. The delivery volume of a
microlitre syringe for liquid spiking (5 l, SGE Ltd.) was
determined by the gravimetric method of ISO 8655-6 using
a small weighing vessel with lid. [3]
Results
Diffusive uptake rates calculated as ng/ppm/min are given in
Table 1. It was estimated that the combined expanded
uncertainty of each mean value in Table 1 is not greater than
5 % for 30-120 exposure and 7 % for 480 minutes exposure
(at 95 % confidence). The effect of exposure time appears to
be anomalous at 480 minutes exposure time, where uptake
rates are significantly lower than those at 30-120 minutes.
We do not believe that the use of mixed vapour atmospheres
in place of single substances has any bearing. There is at the
moment no good explanation other than some unknown bias
in the measurement of the reference concentration or the
amounts on the diffusive tubes that happened only on the
tube sequences from the 480 minute exposure.
. . . . . . . . . . .
. . . . . . . . . . . . .
April 2008
16
Table 1 Diffusive uptake rates (ng/ppm/min) for selected
aromatic compounds on Carbograph 1TD, PerkinElmer type
tube sampler, effect of sampling time and concentration,
mean values from 6-8 samplers, typical combined expanded
uncertainty 5 % (30-120 mins), 7 % (480 mins) at 95 %
confidence.
Exposure time (mins) 30 60 120 480
Concentration
range ppm

1-3 2.02 2.14 2.08 1.36
Benzene
100 2.01 - - 1.51
1-3 2.12 2.30 2.22 1.69
Toluene
100 2.14 - - 1.71
1-3 2.11 2.28 2.22 1.78
m-Xylene
100 2.23 - - 1.82
1-3 2.38 2.34 2.39 1.80
1,3,5-TMB
100 2.18 - - 1.90
We regard the results at 480 minutes as provisional and to be
confirmed or otherwise by a repeat determination. The
greatest confidence is assigned to exposure times of 30-120
minutes. Table 2 lists the mean uptakes rates over sampling
times 30 -120 minutes and concentrations 1 ppm -100 ppm,
compared with theoretical (ideal) uptake rates estimated
from diffusion coefficients taken from Lugg [4]. The
experimental data of Lugg was temperature corrected from
25C to 20C. For this estimation we have assumed that the
Area/Length ratio of the tube sampler air gap is 0.121 cm.





Table 2 Estimated theoretical/ideal uptake rates compared
with mean experimental uptake rates on Carbograph 1TD,
averaged over the sampling conditions of Table 1
(excluding 480 mins), combined expanded uncertainty at
95 % confidence .
D
20
U
d
(ideal) U
d
(exp.)
cm
2
/s ng/ppm/min
Benzene 0.0902 2.13 2.06 0.08
Toluene 0.0822 2.29 2.20 0.09
m-Xylene 0.0666 2.14 2.21 0.09
1,3,5-TMB 0.0641 2.32 2.32 0.09

Conclusions
Over the sampling times 30 480 minutes there was some
evidence that at the longest time, corresponding to a full
shift, the uptake rate was significantly reduced. However,
the reduction was a little more than was expected and needs
further confirmation, particularly for benzene. There was no
evidence of a concentration effect in the range 1-100 ppm
and this would simplify the estimation of concentrations
when not using a single mean value for uptake rate, but
correcting for known bias as a function of sampling time.

References
1. Methods for the Determination of Hazardous Substances, MDHS 80,
Volatile organic compounds in air: Laboratory method using diffusive
solid sorbent tubes, thermal desorption and gas chromatography,
Health and Safety Executive, August 1995, ISBN 0-7176-0913-8.
2. R H Brown, What is the best sorbent for pumped sampling - thermal
desorption of volatile organic compounds? Experience with the EC
sorbents project, Analyst, 1996, 121, 1171-1175.
3. BS EN ISO 8655-6:2002 Piston-operated volumentric apparatus
Part 2: piston pipettes.
4. G A Lugg, Diffusion coefficients of some organic and other vapours
in air. Analytical Chemistry, 1968, 40, 1072-1077.




The Diffusive Monitor, 16
17 . . . . . . . . . . .
. . . . . . . . . . . . .



Indoor Air 2008
The 11th International Conference on Indoor Air Quality and Climate


The 11
th
congress is being organised by the Technical University of Denmark. At the the time of writing the Copenhagen venue
had not been announced.

17-22 August 2008, Copenhagen, Denmark

The series of Indoor Air and Climate conferences started in August 1978 in Denmark. The 11th congress in 2008 celebrates the
30 year anniversary of the inaugural conference by revisiting Copenhagen. The 11th Indoor Air conference will be a
multidisciplinary event involving participants from medicine, engineering, architecture and related fields. The conference will
cover all aspects of Indoor Air Quality and Climate and the effects on human health, comfort and productivity. The conference
will address a variety of indoor environments - residential, office, school, industrial, commercial and transport.

Topics:
Indoor environmental exposure assessment in buildings and vehicles;
Risk Characterization in the indoor environment;
Control and Regulatory options;
Socio-economic context of management of the indoor environment.

The published deadline for abstracts has expired

For further information see the conference website home page http://www.indoorair2008.org/


. . . . . . . . . . .
. . . . . . . . . . . . .
April 2008
18



16th International Conference on
Modelling, Monitoring and Management
of Air Pollution
Organised by the Wessex Institute of Technology, UK;
sponsored by The ASCE UK International Group and
WIT Transactions on Ecology and the Environment.
22 - 24 September, 2008, Skiathos,
Greece


Topics:
Air pollution modelling
Air quality management
Urban air management
Emission studies
Monitoring and measuring
Global and regional studies
Aerosols and particles
Climate change and air pollution
Atmospheric chemistry
Indoor air pollution
Environmental health effects
Remote sensing
Policy studies
Air Pollution Effects on Ecosystems

For further information see the conference website home
page http://www.wessex.ac.uk/conferences/2008/air08/



29th Triennial Congress of the
International Commission on
Occupational Health (ICOH2009)
International organisations participating are ILO and WHO.
22-27 March 2009, Cape Town
International Convention Centre,
South Africa
The Scientific Program has been posted on the website, and
the Call for Abstracts has been issued, with an end date of 21
July 2008 for receipt of Abstracts. Early bird registration has
expired (30 April, 2008). Grants are available for some
presenters from developing nations. Two page brochures
containing the scientific sessions and other key details can be
downloaded from the website for printing and distribution.
Among the 160 listed topics in the Scientific Program are:
Toxicology
Industrial hygiene
Indoor air quality
For further information see the conference website home
page http://www.icoh2009.co.za

The Diffusive Monitor, 16
19 . . . . . . . . . . .
. . . . . . . . . . . . .


THE DIFFUSIVE MONITOR

is available on the HSL web site at: http://www.hsl.gov.uk/publications/diffuse-monitor.htm

If you wish to add your name to the electronic mailing list or delete your name from the list, or have any other
comments, E-mail the Editor at Neil.Plant@hsl.gov.uk, or send the reply slip below (or a photocopy) with details of
your name and address, tel., fax and E-mail to: N T Plant, Health and Safety Laboratory, Harpur Hill, Buxton,
SK17 9JN, UK. fax +44 1298 218570..

THE DIFFUSIVE MONITOR

Response form for fax

... Please add my name to the electronic mailing list

... Please delete my name from the mailing list

E-mail...

Other comments....




Correct address as follows:

Name and Title....

Organisation....

Address....

Post code....

Country....

Telephone....

FAX....

SIGNED....


Return to: N T Plant, Health & Safety Laboratory, Harpur Hill, Buxton SK17 9JN UK, fax +44 1298 218570, e:mail
Neil.Plant@hsl.gov.uk


Data Protection Act: The above details will be collated on a computer database for the purpose of distributing The Diffusive Monitor only.