Performance of methods for measurement of exposure to inorganic acids in workplace air

Abstract

BGIA has organised round robins for the analysis of samples of inorganic acids in workplace air for a number of years. Test samples of the volatile acids HCl and HNO₃ are collected from a standard atmosphere and samples of the non-volatile acids H₃PO₄ and H₂SO₄ are prepared by spiking filters with acid solution. The last two round robins have also covered the sampling of volatile acids, with up to 15 “active” participants able to visit the test facility in Dresden and take samples themselves. For other “passive” participants, BGIA takes samples from the same atmosphere. The acid concentrations generated lie between 0,1 and 1 times the German limit values for HCl and HNO₃. The results for the last round robin showed no significant difference between the performance of the “active” and “passive” participants. The participant means were in good agreement with the theoretical concentrations and the quality control measurements. For “active” participants RSDs were between 7% and 14% and for all participants between 8% and 16%. The round robin for the non-volatile acids showed similar results. The participant means were again in excellent agreement with the quality control measurements and RSDs were between 12% and 15%. The BGIA round robins have demonstrated the proficiency of laboratories measuring exposure to inorganic acids in air. However, concerns remain about the performance of published methods. It has shown that the sampling efficiency of sorbent tubes falls off with increasing particle size and hence silica gel tube methods may give low results for acid mists. Another issue with silica gel tubes is that a substantial proportion of the sample can be collected on the glass wool plugs that retain the sorbent. This can be up to 50% for HCl and 100% for HNO₃. Hence, low results may be obtained if the glass wool plugs are discarded. Similarly, methods for volatile inorganic acids that use a prefilter to remove particulates usually overlook the fact that the acids can react with coparticulate matter on the prefilter. Low recoveries in the range 30%–50% have been found when sampling HCl through filters loaded with potential interferents. Finally, particulate salts interfere with filter sampling methods for non-volatile inorganic acids. A two-part International Standard is in preparation for inorganic acids by ion chromatography and the issues discussed above are being taken into consideration during its development.

---

Dietmar Breuer

Dietmar Breuer was born in Germany, in 1957. He received his PhD in chemistry from University of Paderborn in Germany in 1988. Since 1988 he has been head of the section chemistry II at BGIA. His current research interests are: workplace air, indoor air, organic vapours, volatile organic compounds, inorganic acids, hydroxides, complex organic analytes (e.g. metal working fluids, bitumen), gas chromatography, ion chromatography, infrared spectrometry and proficiency testing.

Introduction

Hydrochloric acid (HCl), nitric acid (HNO₃), sulfuric acid (H₂SO₄) or phosphoric acid (H₃PO₄) are basic industrial chemicals. They are used in millions of tons a year in processes such as extraction of rock phosphates, metal processing, pickling and electroplating and as reagents in chemical processes. The physical state of the acids in workplace air differs from liquid aerosols for non-volatile acids such as H₂SO₄ or H₃PO₄, to mists and vapours for volatile acids like HCl or HNO₃.

Acid mists and vapours are highly corrosive and irritate the eyes and the mucous membranes of the nose, pharynx and respiratory tract, even at low concentrations. The acids have occupational exposure limit values between 0.1–1 mg m⁻³ (H₂SO₄), 1 mg m⁻³ (H₃PO₄), 2–8 mg m⁻³ (HCl) and 5 mg m⁻³ (HNO₃) in various European countries.

Procedures for sampling inorganic acids in workplace air can be divided into two types: non-volatile acids are typically sampled as inhalable aerosols on filters;^1,2 whilst volatile acids are collected on sorbent tubes or alkali-impregnated filters.^2–4 Analysis is usually carried out by ion chromatography.

The Berufsgenossenschaftliches Institut für Arbeitsschutz (BGIA) has been organising round robins for hazardous substances in air since 1989, with round robins for inorganic acids having been conducted annually 1999. During the first five years of inorganic acids round robins, samples of H₂SO₄/H₃PO₄ on quartz fibre filters and HCl/HNO₃ on silica gel tubes were prepared and supplied to participating laboratories by BGIA. However, since 2004 it has been possible for participants to sample the volatile acids themselves at a test gas facility in Dresden.⁵

BGIA round robin results are evaluated in accordance with Annex A of ISO Guide 43.⁶ The participant mean is used as the reference value and the permissible variability is 10% of the reference value. Results with a z-score of <2 are deemed to be satisfactory. BGIA round robins have worldwide participation and the BGIA has regular contact with the organisers of several other international proficiency testing schemes, including the UK’s ‘Workplace Analysis Scheme for Proficiency’ (WASP) and the USA’s ‘Proficiency Analytical Testing’ (PAT) scheme.

The BGIA round robins have demonstrated the proficiency of laboratories involved in measuring exposure to inorganic acids in air. However, there remain concerns about the reliability of exposure data for inorganic acids obtained using published methods. Firstly, some methods overlook the need to collect the appropriate size fraction of airborne particles. Secondly, silica gel tubes, which are commonly used to sample inorganic acids, can collect a substantial proportion of the sample on the glass fibre plug and filter that retain the front section of sorbent in place, rather than on the sorbent itself. If this is overlooked and the plug and filter are discarded it can lead to errors. And thirdly, measurements of exposure to inorganic acids in air can suffer from interferences. In particular positive interferences from inorganic acid salts can occur if the method used is not designed to avoid them; and negative interferences can occur if inorganic acids react with co-sampled particulate matter.

The first part of this paper describes the BGIA inorganic acids round robin and gives an overview of the performance of participating laboratories in 2004 and 2005; whilst the second part of the paper describes on-going concerns about the reliability of exposure data for inorganic acids obtained using published methods, based on the results of laboratory studies carried out independently of the round robin.

Round robin exercises for inorganic acids

Volatile acids

Sampling facility. The volatile inorganic acid round robin exercises have to date been carried out at the BGIA test gas facility in Dresden. However, this has recently been relocated to Sankt Augustin, so future round robin exercises will be take place there. The test gas facility was designed for different types of sampling exercises and, in addition to the sampling exercises for inorganic acids, round robins for organic solvents have been organised twice a year.⁷

Generation of the test gas for volatile acid gases. A Hovalcal® vaporizer was used to generate the test atmosphere. The principal components of the Hovacal® are a computer-controlled analytical balance, a reservoir of dilute acid solution, a mass flow controller for adjustment of the secondary air flow (0.3 m³ h⁻¹) and a peristaltic pump for delivering the acid solution to the evaporator. The evaporator is heated to 150 °C and acid solution and secondary gas flow are introduced in parallel into a centrifugal chamber. Here the solution is vapourised and the gas/acid vapour mixture is injected directly into a heated section (150 °C) of the test gas facility. The technical data are:

Length	15 m (11 m effective)
Internal diameter	5 cm
Air flow	1–4 m^3 h^−1
Humidity	∼5–95% (rel.)
Sample ports	45 (30 for participants)
Participants	15
Diffusion chamber	25 l (for ∼20 diffusive sampler)

Fig. 1 illustrates the experimental design for generating an inorganic acid test gas.

Fig. 1 Experimental design of the test gas facility.

The round robin exercise for volatile inorganic acids involved 3 sampling rounds in which different test gas atmospheres were generated at target concentrations that were calculated based on the current German limit values (LV) for HCl (8 mg m⁻³) and HNO₃ (5.2 mg m⁻³) and the requirements for the minimum working range (0.1 × LV to 2 × LV) of a measurement method according to EN 482.⁸

An electrochemical HCl sensor was used to measure the total acid concentration in the test gas and check that it remained constant over the sampling period. After a stabilisation period of about 1 hour the total acid concentration was constant and the sampling period of 2.5 hours was started.

Control sampling of the test gas. Up to 14 of the sampling ports can be used for control samples and 1 is used for measurement of humidity and temperature. For this round robin, 10 control samples were collected in each sampling round at ports distributed evenly over the test gas facility. The sampling and analytical conditions were as follows:

Sampling tubes	ORBO 53™ silica gel tubes
Air flow	250 ml min^−1
Sampling time	2 h
Humidity	∼50%
Extraction	10 ml extraction solution
Extraction solution	3.1 mmol l^−1 Na2CO3
	0.35 mmol l^−1 NaHCO3
Ion chromatograph	Dionex ICS 2000
Eluent	Automated eluent generation
Column	AS 14A
Flow	1.5 ml min^−1
Injection volume	50 μl
Detection	Conductivity

Table 1 gives the results of the control samples from the 2005 sampling exercise. All results compare well with the calculated target concentration and the low RSDs suggest that the homogeneity of the test gas atmosphere was good. The average results of the control sample were used as reference values for evaluation of the round robin.

Table 1 Results of control samples

Round		Number of samples	Reference concentration^a /mg m^−3	RSD (%)	Calculated target concentration/mg m^−3
a Mean value of control samples.
2005/1	HCl	10	1.11	2.3	1.05
2005/1	HNO3	10	0.99	2.7	1.05
2005/2	HCl	10	6.52	0.9	6.64
2005/2	HNO3	10	5.58	1.8	5.84
2005/3	HCl	10	3.21	1.7	3.23
2005/3	HNO3	10	11.2	0.8	11.5

---

Round robin samples. Participants were able to choose between taking a set of test samples collected by BGIA (“passive participation”) and taking the samples themselves at the test gas facility.

For the “passive” participants, BGIA collected the samples on ORBO™ 53 silica gel tubes at the same time as the “active” participants carried out their own sampling of the same test gas. For this reason, results submitted by both “active” and “passive” participants could be evaluated together. In the 2005 round robin, 25 participants received a sample set containing two blank tubes, one opened unloaded tube and three sample tubes (one for each concentration). The sampling conditions were the same as described in the previous chapter for control samples.

For the “active” participants, there was no limit imposed on the number of samples that could be taken over the sampling period of 2.5 hours. All “active” participants took several samples and they mostly used ORBO 53 silica gel tubes. However, some participants also used fritted wash bottles with an alkaline sampling solution and impregnated filters.

Round robin results. The round robin had 31 participants from 7 European countries and the United States.

The analytical methods used by the participants were very similar. 23 participants used chemically suppressed ion chromatography and 8 participants used electronically suppressed ion chromatography with conductivity detection. Depending upon the chromatographic technique employed, the extraction medium used was mostly either alkaline sodium carbonate solution or pure water, but 3 laboratories used eluent buffer solution for extraction. Nearly all participants used an ultrasonic bath.

In the 2004 round robin, the target concentrations, results for the control samples and participant results showed good comparability.⁵

In the 2005 round robin, the second and third sampling rounds once again gave results with good comparability, as shown in the general data in Table 2 and in the single results for round 3 illustrated in Figs. 2 and 3. The results were normally distributed and with a few exceptions all participants fulfilled the requirements and gave satisfactory results with z-scores of <2. For HCl, as in 2004, the results of the “active” participants exhibited a tighter distribution and gave a better fit with the target concentration than those of participants who received the sample set from BGIA. On-site sampling has the advantage that participants can take as many samples as they want and then average or pre-select the results they report. A “passive” participant has only one sample with which to work, although this can be analysed several times. Contrary to this explanation and to the outcome of the 2004 round robin, the 2005 HNO₃ results of the “active” participants were slightly worse than those of the “passive” participants. It is hoped that the results of future round robins will determine if the apparent difference between “active” and “passive” participants is random or systematic.

Fig. 2 Results HCl third round (“active” participants: □, “passive” participants: ◆, outlier: ○).

Fig. 3 Results HNO₃ third round (“active” participants: □, “passive” participants: ◆, outlier: ○).

Table 2 Summary of participant results

Round	Acid	Number of samples	Average concentration for all participants/mg m^−3	RSD (%)	Number of samples	Average concentration for “active” participants/mg m^−3	RSD (%)
2005/1	HCl	31	0.98	15.8	7	0.96	Not calculated
2005/1	HNO3	31	0.99	11.6	7	1.10	Not calculated
2005/2	HCl	31	6.50	12.1	7	6.40	5.2
2005/2	HNO3	31	5.25	10.8	7	5.30	15.3
2005/3	HCl	31	3.32	8.7	7	3.25	3.2
2005/3	HNO3	31	10.6	10.0	7	10.4	12.1

---

However, the results of the first sampling round in 2005 were unsatisfactory. A first glance at the summary given in Table 2 suggests that the results for round 1 are also fine, but on closer inspection it can be seen that for both acids one third of the participants reported a significantly low recovery (see Figs. 4 and 5). Three of the “active” participants belong to this group and the average result given in Table 2 is calculated from only 4 results, so it was impossible to calculate further statistical data. For HNO₃ the results appear to be split into two groups with a step change between them; whereas for HCl there is also an apparent discontinuity, but the lower values are distributed over a larger range. Notwithstanding this, for both acids about 20 participants reported results that were in a good agreement with the reference value. It is not obvious why the results showed this behaviour. The lowest concentration of both acids was generated in the first sampling round and it is possible that the observed performance characteristics might be related to this in some way. One possible explanation for the low recovery is that the acids were partially adsorbed onto the glass fibre plug that retains the sorbent and that some participants analysed this along with the sorbent and some did not. It might be expected that this effect would be more pronounced at lower concentrations, which could explain why it was only apparent in the first sampling round. Specific investigations are planned in order to clarify this question.

Fig. 4 Results HCl first round (“active” participants: □, “passive” participants: ◆).

Fig. 5 Results HNO₃ first round (“active” participants: □, “passive” participants: ◆).

Non-volatile acids

Round robin samples. Test samples were prepared at BGIA by spiking an aqueous solution containing dilute H₂SO₄ and H₃PO₄ directly onto a quartz fibre filter using a microlitre syringe. The round robin exercise for non-volatile inorganic acids involved three different sample concentrations that were calculated based on the 2004 German limit values (LV) for H₂SO₄ (0.1–1 mg m⁻³) and H₃PO₄ (1.0 mg m⁻³) and the requirements for the minimum working range (0.1 × LV to 2 × LV) of a measurement method according to EN 482.⁸ The participants were asked to calculate results based on a 420 l air sample volume.

In the 2005 round robin, 26 participants received a sample set containing 3 blank filters and 3 sample filters (one for each concentration). For quality control purposes, every third filter was analysed under the same conditions given earlier for volatile acid control samples. The results for the non-volatile acid control samples are given in the first three columns of Table 3.

Table 3 Summary of results for non-volatile acids round robin

Round	Acid	Number of control-samples	Calculated target concentration/mg m^−3	RSD (%)	Number of participants	Average concentration/mg m^−3	RSD (%)
2005/1	H2SO4	12	0.050	1.3	26	0.051	15.0
2005/1	H3PO4	12	0.084	1.2	26	0.083	15.3
2005/2	H2SO4	12	0.30	1.0	26	0.30	13.6
2005/2	H3PO4	12	0.30	0.7	26	0.30	12.2
2005/3	H2SO4	12	0.72	0.9	26	0.71	12.6
2005/3	H3PO4	12	0.93	1.0	26	0.91	12.1

---

Round robin results. 25 of the 26 laboratories that participated in the 2005 round robin for non-volatile acids also participated in the round robin for volatile acids and they normally used the same sample extraction and analytical method for both types of sample. In case of the non-volatile acids, 19 participants used chemically suppressed ion chromatography and 7 participants used electronically suppressed ion chromatography with conductivity detection. No influence of the analytical technique was observed.

In 2005, there was excellent agreement between the participant mean and the calculated target concentration for all three rounds, as shown in Table 3. The results were normally distributed with RSDs between 12% and 15%. About 80% of the participants fulfilled the requirements and gave satisfactory results with z-scores < 2. Unlike for the volatile acids, the performance of laboratories participating in the 2005 round robin for non-volatile acids was very similar to that previously observed.

Concerns about methods for inorganic acids

Size fraction of airborne particles collected

When sampling any aerosol, including inorganic acid mists, it is important to collect the appropriate size fraction of airborne particles to which an exposure limit applies. This is clearly the case for non-volatile acids, such as H₂SO₄, but it is also relevant for volatile inorganic acids, such as HCl, if these could be present in the air as mist, in addition to vapour.

The size fraction of airborne particles to which a particular exposure limit applies can vary from country to country and can be implicit rather than explicit. However, exposure limits generally apply to the inhalable fraction of airborne particles¹⁰ unless it is explicitly stated to the contrary. It is therefore this size fraction that should be collected when sampling acid mists. The situation is complicated somewhat for HCl by a lack of clarity in the limit-setting process, in that whilst most countries have set a limit value for hydrogen chloride rather than hydrochloric acid, occupational hygienists generally take the view that the limit value for HCl applies to both mist and vapour.

Filter sampling methods such as the German DFG method for inorganic acid mists¹ and the French Métropol 009 method for mineral acids² either sample the inhalable fraction of inorganic acid mists, or they could do so if used with an appropriate sampler. However, silica gel sorbent tube methods, such as the widely used NIOSH 7903 method,⁴ significantly under-sample with respect to the inhalable convention because the sampling efficiency of sorbent tubes falls off with increasing particle size. This is illustrated by the results of some experiments carried out at HSL to compare the sampling efficiency of open-face cassettes with silica gel sorbent tubes.⁹ Dust aerosols were generated and samples collected in a dust box, the amount of dust collected being determined gravimetrically. It can seen from Table 4 that, at a typical flow rate of 200 ml min⁻¹, aerosols consisting of particles with a median diameter of 15 μm are collected with an efficiency of about 33% in comparison with open-face cassettes, which in turn are known to under-sample in comparison with inhalable samplers. This strongly suggests that silica gel tube methods are likely to give low results if used to sample acid mists.

Table 4 Sampling characteristics of silica gel tubes

		Ratio to open-face samplers
Particle size/μm	Flow rate/ml min^−1	Orientation up	Orientation down
5	100	0.48	0.88
	200	0.48	0.66
	500	0.21	0.21
15	100	0.54	0.38
	200	0.41	0.26
	500	0.13	0.06
25	100	0.07	0.08
	500	0.01	0.01

---

Trapping characteristics of silica gel tubes

Non-volatile acid mists, to the extent that they are collected by sorbent tubes, are trapped on the glass fibre plug and filter that retain the silica gel in place, and the NIOSH 7903 method⁴ properly describes their desorption and analysis by IC.

However, a substantial proportion of the volatile acids that are meant to be collected on the silica gel can also in fact be trapped on the preceding glass fibre plug and filter. Although the normal procedure described in the NIOSH 7903 method⁴ is to analyse the glass fibre plug and filter together with the silica gel, the method does advise separate analysis of the glass fibre plug and filter to estimate particulate salts, when these are present. This is a concern, firstly because if this advice is followed low results will be obtained for volatile acids and high results for particulate salts, but also because it could lead users to assume that they need only analyse the silica gel when no particulate salts are present.

The extent of this problem was identified through the analysis of samples taken over the last two rounds of the BGIA round robin exercises. A high percentage of the HCl and HNO₃ was found on the glass fibre plugs and filters, decreasing with the amount of analyte collected. This is illustrated by the results shown in Figs. 6 and 7.

Fig. 6 Distribution of chloride on silica gel tubes.

Fig. 7 Distribution of nitrate on silica gel tubes.

The first figure shows that typically around 5μg of chloride is trapped on the glass fibre plug and filter. There appears only to be a slight increase with the total amount of chloride trapped, so the % chloride trapped on the plug falls of quite rapidly with the amount collected.

For nitrate the situation is rather different. At low HNO₃ concentrations, virtually 100% of nitrate is trapped on the glass fibre plug and filter. This falls off as the amount of nitrate collected increases, but it is still around 40% when the total nitrate collected is over 100 μg.

The two figures clearly demonstrate, that if the glass wool plugs are analysed separately or discarded to avoid possible interference from particulate salts, low results may be obtained for volatile acids.

Interference from particulates

Volatile acids. Most methods for volatile inorganic acids use a prefilter to remove particulate salts of the corresponding anions, when present, since these will otherwise interfere with their measurement by ion chromatography. However, such methods generally overlook the fact that the acids can react with co-sampled particulate matter on the prefilter, leading to erroneous results. For example, in galvanizing plants, reaction of HCl with ZnO may lead to low results for HCl and high results for ZnCl₂. Preliminary laboratory experiments have been carried out to investigate the extent of this problem. These involved sampling HCl vapour through filters loaded with milligram quantities of some potential interferents and the results are shown in Table 5.

Table 5 Effect of co-sampled particulates on HCl results

		HCl collected on GLA5000 prefilter (%)
Material deposited on GLA5000 prefilter	Total HCl collected/μg	Replicate	Mean	SD	RSD
No deposit	77	0.4	0.4	0.1	36
	56	0.3
	82	0.5
Fe2O3	60	0.4	0.5	0.1	23
	74	0.7
	61	0.5
Fe	61	16	16	7	45
	56	23
	69	9
ZnO	46	52	54	11	21
	42	66
	58	44
Welding fume	213	81	71	11	16
	125	59
	127	72

---

It can be seen that there was a substantial interference from iron, zinc oxide and welding fume observed, with HCl recoveries of around 85, 50 and 30%, respectively, for these three substances. However, perhaps surprisingly, the presence of iron oxide did not lead to a low HCl recovery. The results are of particular concern because of the somewhat intractable nature of the problem they show can occur. Whilst the use of a prefilter is necessary to prevent positive interference from particulates, e.g. ZnCl₂ and NH₄Cl in galvanizing works, it leads inevitably to negative interference from co-sampled particulate. Perhaps denuder technology is the answer, but at the present time the use of denuders is not really a practical proposition for personal exposure measurement.

Non-volatile acids. Measurement of non-volatile acids is also prone to interference from particulate salts of the corresponding anions. However, in this case a prefilter can clearly not be used to remove the particulate salts as the acids themselves are present in the form of an aerosol. Therefore, the only option is to attempt to separate the acids from their salts on the basis of physical or chemical properties. One approach that is used to separate H₂SO₄ from particulate sulfates involves dissolution of the acid in anhydrous propan-2-ol. This technique was developed some years ago and incorporated in the French Métropol 009 method.² However, whilst data are available to show that this approach works for relatively large amounts of H₂SO₄, in the range 350–700 μg,¹¹ there is no information about its performance at low levels. The United States National Institute of Occupational Safety and Health (NIOSH) has therefore recently commissioned research to evaluate the method and this work is on-going.

Another approach used in Germany is to estimate the relative amounts of anion present as acid and particulate by collecting a sample from the emission source (e.g. pickling bath fluid) and analysing this for anion and acid content. The results from the emission source sample can then be used to estimate the free acid concentration in the air sample.

Conclusions

The round robin exercises organised by BGIA have shown that the participating laboratories mostly have a good command of techniques for measurement of completely vaporised HCl and HNO₃ and for H₂SO₄ and H₃PO₄. However, some participants obtained low results at low concentrations of HCl and HNO₃. The reason for this is currently not clear. One possibility is the adsorption of HCl and HNO₃ vapours onto the glass fibre plug of ORBO™ 53 silica gel sorbent tubes. Future research at the test gas facility will attempt to establish whether this is the case.

A second question in measuring volatile inorganic acids is their physical state in workplace air. Volatile acids are mostly used as solutions and during the work process the acid is vapourised and/or aerosols are generated. If the acid is completely vaporised, the commonly used silica gel sorbent tube method works well. If a significant part of the acid is present as an aerosol during the sampling period, undersampling can occur when using sorbent tubes. This leads to the conclusion that sampling volatile inorganic acids in workplaces where aerosols could be present should be carried out using a combined aerosol/vapour sampling system.

Thirdly, measurement results for volatile acids in workplace air can be adversely influenced by co-sampled particulate matter. The acids are able to react with substances like zinc oxide or welding fume and this can result in a significant decreased in measured HCl and HNO₃ concentrations. More work is required in order to identify and quantify possible interferences.

References

1. Deutsche Forschungsgemeinschaft (DFG), Inorganic acid mists (H₂SO₄, H₃PO₄), in Analyses of hazardous substances in air, ed. A. Kettrup, Wiley-VCH Verlag, Weinheim, Germany, 2002, vol. 6, ISBN 3 527 27053 1.
2. Institut National de Recherche et de Sécurité (INRS), Métropol Fiche 009-Anions Minéraux, in Métrologie des polluants (Métropol), INRS, Vandoeuvre, France, 2005, www.inrs.fr.
3. Deutsche Forschungsgemeinschaft (DFG), Volatile inorganic acids (HCL, HBr, HNO₃), in Analyses of hazardous substances in air, ed. A. Kettrup, Wiley-VCH Verlag, Weinheim, Germany, 2002, vol. 6, ISBN 3 527 27053 1.
4. US National Institute of Occupational Safety and Health (NIOSH), Method 7903-Acids, Inorganic, in NIOSH Manual of Analytical Methods, DHHS (NIOSH) publication 94–113, 4th edn., 1994, http://www.cdc.gov/niosh/nmam/.
5. D. Breuer, B. Maybaum, K. Gusbeth, S. Rosenthal and M. Seifert, Ringversuche mit Probenahme zur Bestimmung von flüchtigen anorganischen Säuren (HCl und HNO₃), Gefahrstoffe Reinhaltung der Luft—Air Quality, 2005, 65, 122–127.
6. International Organisations for Standardisation (ISO), Proficiency testing by interlaboratory comparisons—Part 1: Development and operation of proficiency testing schemes, ISO/IEC guide 43-1, 2nd edn, annex A, Geneva, 1997.
7. S. Rosenthal, M. Seifert, B. Maybaum, K. Gusbeth and D. Breuer, Ringversuche mit Probenahme zur Bestimmung von Lösemitteln an der neuen Prüfgasstrecke im BGAG, Gefahrstoffe Reinhaltung der Luft—Air Quality, 2003, 63, 355–362.
8. EN 482:1994, Workplace atmospheres—General requirements for the performance of procedures for the measurement of chemical agents.
9. K. Y. K. Chung, Collection efficiencies of sorbent tubes for larger particles, HSL report IR/A/98/06. Health and Safety Laboratory, England, 1998, www.hsl.gov.uk.
10. EN 481:1993, Workplace atmospheres—Size fraction definition for measurement of airborne particles.
11. S. K. Ludlow, J. S. Lee and J. H. Nelson, A method for separating sulfuric acid from particulate sulfate salts on filters, Rocky Mountain Centre for Occupational and Environmental Health, University of Utah, Salt Lake City, 1994.

---

Footnote

† Presented at the Fifth International Symposium on Modern Principles of Air Monitoring & Biomonitoring, June 12–16 2005, Norway.

---