The Diffusive Monitor

April 2008Issue 16

Health and Safety Executive Committee of Analytical Requirements - Working Group 5

Inside this Issue

1 By way of introduction

1 Current News

4 NO2 and SO2 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 Editor’s 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 Editor’s 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 Comité Européen 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 für 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 (Québec), 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 SO2/NOx/O3/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 NO2 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. Comité Européen 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: D126–D136.

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/nordtestfiler/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 meshescoated with TEA

Teflon filter to stopaerosols and stabilize

diffusion path

Coloured cap accomodatingthe meshes

Acrylic plastic tube

Cap to be removed during

sampling and replaced by a filter holder

Determination of NO2 and SO2 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 NO2 and SO2 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 NO2 and SO2. However, field measurements showed that the open-ended diffusion tube is affected by a strong artefact on the SO2 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 NO2 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 SO2 is presented.

Figure 1 Description of a Palmes tube

Principle

NO2 and SO2 are collected by molecular diffusion along the acrylic tube to the TEA where it is retained for subsequent measurement. The first Fick’s law describes diffusive transport and allows after integration to determine the airborne concentration using an equation [10] of the following form:

tUmmC b−

= (1)

where: U – sampling rate, ng ppb-1 min-1; m – mass of the pollutant, ng; mb – 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 0C until they are completely dry. The stainless steel mesh discs (XDISC) are cleaned in an ultrasonic bath, at 60 0C for 5 hours, changing the water every half an hour. Then, they are placed in an oven, flushed with nitrogen, at 125 0C until they are completely dry. Three clean and dry discs are placed in the coloured cap (see Figure 1). 40 µl of a 10% v/v solution of TEA (Merck nº8379) with 0.3 % of non ionic detergent (Brij-35, Merck nº 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. NO2 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 µl of 30 % hydrogen peroxide (H2O2) is added to ensure complete oxidation of SO3

2- to SO42-. 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 µl and an anion self regenerating suppressor AMMS–2 mm with sulphuric acid. For NO2 and SO2, a 6-point calibration, in the range of 0 to 0.3 µg 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 H2O2 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 SO2 membrane closed Palmes tube.

Laboratory experiments

NO2

The study of Gerboles et al. [9] focused on NO2 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 µg m-3 week.

The uptake rate (U) in ng ppb-1min-1 is calculated using the following equation 2, where RH is the relative humidity in %, T is the temperature in ºC, w is the wind speed in m s-1:

( )

( )

2

10.64.01018.328.100130.086.0134

21018.328.100130.086.0134

5102

5

tmwTtRH

wTtRHU

−−

−

−++−+

+++−+

= (2)

SO2

In order to evaluate the limit of detection for SO2, 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 SO2, 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 SO2 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-1min-1 for the first four experiments in which SO2, 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 5th experiment which was similar to the first one except that NO2 was added, the uptake rate was found to be 0.00213 ng ppb-1min-1. This value confirms the uptake rate of the first experiments and does not show any evidence of NO2 interference on the measurement of SO2 with the MCPTs. Finally, the last two experiments, at high relative humidity, low temperature and SO2 concentration gave an uptake rate of approximately 0.002075 ng ppb-1min-1 which confirmed within 3 % that an

No of days field, µg room, µg fridge, µg

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, µg 0.031 (n=18) 0.032 (n=38) 0.015 (n=39)

st.dev, µg 0.026 (n=18) 0.021 (n=38) 0.009 (n=39) LoD,

µg 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 SO2

ppb

NO2

ppb

RH

%

T

°C

w,

m.s-1

Time,

days

Uptake rate,

ng.ppb-1min-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-1min-1 can be used for the SO2 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 NO2 repeatability was 5.5 µg 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 NO2 and SO2. For NO2, the repeatability was 9 µg m-3 (standard deviation of 3.2 µg m-3) while for SO2 it was 8 µg m-3 (standard deviation of 2.7 µg 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 NO2 and SO2 in one run by ion chromatography using the MCPT. While for NO2 the uptake rate is already known, the uptake rate for SO2 was found to be independent of the conditions of exposure with a value of 0.00214 ng ppb-1min-1. However, it is necessary to perform some field tests in which the MCvicinity of th

[1] Lin, J.M., Lin, T.S., A diffusive sampler for the ion-chromatogmeasurement 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 samplinsystem for monitoring SO2 in the atmosphere, Field analytical chemistand 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., Pérez Ballesta, P., De SaegE., Modelling of the uptake rate of nitrogen dioxide Palmes diffusivesampler 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 Fré, 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 l’Environnement et de la maîtrise de l’Energie, ADEME Éditions, Réf. : 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 operator’s 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?

Wouldn’t 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?

Wouldn’t 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?

Wouldn’t 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 can’t 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 140ºC and the associated read/write devices don’t 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 ¼-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 can’t 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 laboratory’s 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. 400ºC 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.

GCDetector

Optional‘inlet’ split ‘Desorb flow’

Cooledtrap

Carrier gas in

Heatedsample

tube Analytical column

GCDetector

Optional‘inlet’ split ‘Desorb flow’

Cooledtrap

Carrier gas in

Heatedsample

tube Analytical column

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

Carrier gas in

GCDetector

Analytical column

Optional ‘outlet’ split

heated trap

Carrier gas in

GCDetector

Analytical column

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 don’t 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.

( )

ηπ

⋅⋅⋅−⋅⋅

=o

oico PL

PPdF256

224

………………… Equation 1.

Where: Fo is the flow rate at the column outlet dc is the internal diameter of the column L is the length of the column Pi is the carrier gas pressure at the column inlet Po is the carrier gas pressure at the column outlet η 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, η. Inspection of Equation 1 indicates that, as the viscosity increases an appropriate increase in the inlet pressure, Pi, can be applied to keep the column outlet flow rate, Fo, 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 line to column

B. Regulator connected via trap and transfer line to column

C. Regulator connected via trap, split and transfer line to column

pressure regulator

secondary trap

splittransfer

line

column

B

B

B

detectorA

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 regulated at column inlet

BA

C T

V

A. Pressure regulated prior to trap

BA

C T

V

B. Pressure regulated between trap and transfer line

BA

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.

PPii PPxx PPoo

Transfer lineTransfer line GC ColumnGC Column

( )xtt

xitt PL

PPdF

⋅⋅⋅−⋅⋅

=η

π256

224 ( )xcc

oxci PL

PPdF

⋅⋅⋅−⋅⋅

=η

π256

224

FFiiFFtt FFoo

PPii PPxx PPoo

Transfer lineTransfer line GC ColumnGC Column

( )xtt

xitt PL

PPdF

⋅⋅⋅−⋅⋅

=η

π256

224 ( )xcc

oxci PL

PPdF

⋅⋅⋅−⋅⋅

=η

π256

224

FFiiFFtt FFoo

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:

Ft is the flow rate at the transfer line outlet dt is the internal diameter of the transfer line Lt is the length of the transfer line Pi is the carrier gas pressure at the transfer line inlet Px is the carrier gas pressure at the transfer line outlet

and GC column inlet ηt is the viscosity of the carrier gas at the transfer line

temperature Fi is the flow rate at the GC column inlet dc is the internal diameter of the GC column Lc is the length of the GC column Po is the carrier gas pressure at the GC column outlet ηc is the viscosity of the carrier gas at the GC column

temperature Because the flow rate out of the transfer line, Ft, and the flow rate into the GC column, Fi, 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, Fo, 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 . . . . . . . . . . . . . . . . . . . . . . . .

( )⎟⎟⎠

⎞⎜⎜⎝

⎛ ⋅⋅+⎟⎟

⎠

⎞⎜⎜⎝

⎛ ⋅⋅

−⋅

⋅⋅

=

44

22

256

c

ccc

t

ttt

oi

o

co

dLT

dLT

ppp

TFηη

π ----- Equation 2

Where: Fo is the flow rate at the column outlet Tt is the transfer line absolute temperature Tc 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 ChromatographThermal Desorber

Detector

GC column

GC oven

Heated transfer line

tubing

PPC pressureregulator

Signal cable from

thermocouple

Thermocouple

T4

Gas ChromatographThermal Desorber

Detector

GC column

GC oven

Heated transfer line

tubing

PPC pressureregulator

Signal cable from

thermocouple

Thermocouple

T4T4

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

Transferline

GC

Flame Ionization Detector

ATD

ΔΔPP

Methane

Helium

PR1

PR2

MFC1

MFC2

BPR1(100psig)2mL/min

400mL/min

Transferline

GC

Flame Ionization Detector

ATD

ΔΔPP

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 6Measured Flow Rate (mL/min)

FID

Sig

nal (

mV)

8

r2=0.9996

0

100

200

300

400

500

600

700

800

900

1000

0 2 4 6Measured Flow Rate (mL/min)

FID

Sig

nal (

mV)

8

r2=0.9996

0

100

200

300

400

500

600

700

800

900

1000

0 2 4 6Measured Flow Rate (mL/min)

FID

Sig

nal (

mV)

8

r2=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 36min

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 34 360 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 340 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36min

mV

Figure 10 Comparison between constant flow and pressure controls on a 15m x 0.25mm column programmed from 40°C for 1 minute, then 10°C/min to 300°C and held for 10 minutes. The transfer line was 1.8m x 0.28mm and held at 300°C. 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 16.00.0 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 40°C/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 μl, 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 25°C to 20°C. 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 .

D20 Ud (ideal) Ud (exp.) cm2/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 11th 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.