Document npaVObqbRRxYjpYkgMK7eapo8

MDHS Methods for the Determination of Hazardous Substances Health and Safety Laboratory HSE 101 Crystalline silica in respirable airborne dusts Direct-on-filter analyses by infrared spectroscopy and X-ray diffraction February 2005 This method updates and replaces MDHS37 (ISBN 0 7176 0208 7, last published in March 1988) MDHS51/2 (ISBN 0 7176 02559, published in March 1988) and MDHS76 (ISBN 0 7176 0634 1, published in March 1994). The principal changes are (i) discussion of Fourier Transform Infrared spectroscopy (FTIR) as well as dispersive; (ii) changes, mostly in the section on analysis by X-ray diffraction, to the discussion of the crystalline content of quartz and cristobalite as a consequence of a single explosure limit being set in terms of crystalline silica; (iii) the inclusion of data on detection limits and uncertainty. INTRODUCTION Requirements of the COSHH Regulations 1 The Control of Substances Hazardous to Health Regulations (COSHH)1 are designed to ensure that the exposure of people at work to substances which could cause health damage is either prevented or, where that is not reasonably practicable, adequately controlled. Employers are required to make an assessment of the health risk created by such work, and to prevent or control exposure to the substances involved. The COSHH Regulations also require that persons who could be exposed to substances hazardous to health receive suitable and sufficient information, instruction and training. Employers must therefore ensure that the requirements of the COSHH Regulations are fulfilled before allowing employees to undertake any procedure described in this MDHS. 2 Guidance is given in Control of substances hazardous to health. The Control of Substances Hazardous to health Regulations 2002. Approved Code of Practice and guidance.2 Health effects 4 The health effects of crystalline silica are summarised in EH59.3 The scientific evidence relating to the risks of silicosis and lung cancer are reviewed in Hazard Assessment Documents EH75/44 and EH75/55 respectively. Health and safety precautions 5 Prevention and control of exposure, emergency procedures and health surveillance are described in HSE Guidance Note EH59.3 Exposure 6 Work activities which are likely to lead to exposure to crystalline silica are listed in HSE Guidance Note EH59.3 They include primary extraction or working of rocks containing quartz, especially tunnelling and quarrying; the drilling, cutting, dry grinding, scabbling, abrasive blasting and demolition of building and other materials containing quartz or cristobalite; the handling of finely ground material in ceramics manufacture and brick making; and the fettling of metal castings and earthenware. Additional information is contained in HSE guidance6-11 relating to control of exposure to respirable silica in potteries, stonemasonry, heavy clay and refractory processes, and construction. Exposure limits 7 Regulation 7 of the COSHH Regulations1 lays down the requirements for using maximum exposure limits (MELs) and occupational exposure standards (OESs) for the purpose of achieving adequate control of worker exposure. Occurrence, properties and uses 3 The occurrence, properties and uses of crystalline silica are summarised in HSE Guidance Note EH59.3 8 Schedule 1 of the COSHH Regulations1 specifies a maximum exposure limit (MEL) of 0.3 mg/m3, 8-hour timeweighted average (TWA) reference period, for crystalline silica in respirable airborne dust. This limit is reproduced in HSE Guidance Note EH4012 and the criteria on which it was based are documented in the 1994 supplement of 1 HSE Guidance Note EH64.13 HSE believes that in most cases it should be reasonably practicable to control exposure to 0.1 mg/m3 (8-hour TWA) or less by engineering or process control. Employers should aim to ensure that workers are not exposed to respirable crystalline silica dust concentrations above this level. If exposure cannot be controlled to 0.1 mg/m3 (8-hour TWA) or below by elimination or process or engineering controls, then exposure must be controlled by provision and use of suitable respiratory protective equipment.14 airborne concentrations much lower than 0.3 mg/m3 will need to be measured.The mass of quartz/cristobalite on a typical filter is of the order of 50 |ig. 16 The upper limit of the analytical range is also governed by the total amount of dust on the filter. Experience has shown that heavy dust deposits may easily become dislodged during handling or transit of the samples and the total amount of dust on the filter is best kept below about 2 mg. 9 The MEL applies to the total amount of crystalline silica, quartz and cristobalite, and there are no longer separate exposure limits for those two forms of crystalline silica. Analytical methods 10 There may be alternative methods available for the determination of a particular analyte. With the exception of a few cases where an exposure limit is linked to a specific method (eg rubber fume or asbestos), the use of methods not included in the MDHS series is acceptable provided that they have been shown to have the accuracy and reliability appropriate to the application. 11 This method has been validated to demonstrate that it complies with BS EN 482 Workplace atmospheres General requirements for the performance of procedures for the measurement of chemical agents.15 If an alternative method is used it is necessary to carry out a similar assessment of its performance. PRINCIPLE OF METHOD 12 A sample of respirable dust is collected on a membrane filter using a respirable dust sampler.165 The filter is then placed directly into the sample beam of either an infrared spectrophotometer or an X-ray diffractometer. The mass of crystalline silica on the filter is determined from the infrared or X-ray diffraction response, calibrated against filters loaded with known amounts of standard quartz or cristobalite. Since the volume of air sampled is known, the concentration of airborne crystalline silica is readily calculated. 13 The choice of analytical technique, ie infrared spectroscopy or X-ray diffractometry, depends largely on other materials present on the filter which may interfere in the analysis. SCOPE OF METHOD 14 The method is suitable for the determination of quartz and cristobalite within the approximate range 20 |ig to 1 mg. Both infrared and X-ray diffraction responses are linear over this range. 15 For a 500 litre air sample (corresponding to sampling for approximately four hours at 2.2 litre/min), a concentration of 0.3 mg/m3 corresponds to 150 |ig of quartz/cristobalite on the filter. However, the concentration of airborne crystalline silica should be reduced to as low a value as is reasonably practicable and, accordingly, 17 Further limitations to the scope of the method are discussed separately below for measurement by infrared and by X-ray diffraction. 18 HSG17317 advises employers about how they should conduct investigations into the nature, extent and control of exposure to substances hazardous to health which are present in workplace air. The objective of air monitoring is usually to determine worker exposure, and therefore the procedures described in this method are for personal sampling in the breathing zone. The method may also be used for background or fixed location sampling, but it should be recognised that, due to aerodynamic effects, samplers designed for personal sampling do not necessarily exhibit the same collection characteristics when used for other purposes. SAMPLING EQUIPMENT Samplers 19 Higgins-Dewell cyclone samplers with 25 mm diameter filters are the type most commonly used but any of the other types of respirable dust sampler described in MDHS14/316 may be used. However, CIP10 samplers use a foam pad to collect the respirable dust and they are not suitable for the direct-on-filter method of measurement described here. Filters 20 Filters should be of a diameter suitable for use in the selected sampler. The chosen filter type should have a capture efficiency of not less than 90% and be suitable for collection of samples of crystalline silica. PVC filters (such as Gelman GLA5000) or PVC-acrylonitrile co-polymer filters (such as Gelman DM800) have suitable infrared characteristics. Silver filters are suitable for X-ray diffraction analysis but not infrared spectroscopy: they have a lower background for XRD than organic membranes, giving a better signal-to-noise ratio, but are much more expensive and may have other disadvantages related to their high X-ray absorption. Sampling pumps 21 Sampling pumps should comply with the provisions of BS EN 1232,18 have an adjustable flow rate, incorporate a flowmeter or a flow fault indicator, and be capable of maintaining the selected flow rate to within 5% of the nominal value throughout the sampling period. When worn, the pump should not impede normal work activity. 2 22 For cyclone samplers, pulsation damped flow is particularly important and an external pulsation damper must be used if the pump does not contain an integral damper. Flowmeter 23 The portable flowmeter used should be capable of measuring the appropriate flow rate to within 5%, and calibrated against a primary standard.16 Flowmeters incorporated in sampling pumps are not suitable for accurate measurement of the flow rate. However, they can be useful for monitoring the performance of samplers, provided they have adequate sensitivity. Standard reference materials 30 For both techniques, it is essential to have a standard quartz or cristobalite in which the percentage of crystalline quartz or cristobalite is known. 31 The infrared absorption response of both quartz and cristobalite is particle size dependent.19, 20 The response increases as the particle size decreases to about 1.5 pm; below about 1.5 pm, the response falls due to the presence of an amorphous surface layer. A standard reference sample for infrared analysis should therefore ideally have the same particle size distribution as the sample under consideration. LABORATORY APPARATUS Infrared spectrophotometer 24 This method requires either an FTIR or a double beam dispersive instrument capable of at least x10 ordinate expansion on absorbance spectra. 25 A suitable filter holder, eg a rotatable polariser mount, is required so that the sample can be rotated in its own plane. This will allow the effect of non-uniform sample deposition to be reduced by taking absorbance measurements at several orientations. 32 Only crystalline silica will give an XRD response, and particle size differences between the sample and the standard are much less important for that technique,21 although peak broadening and extinction effects are related to particle size.21-23 But the effect of particle size on crystalline content may be a serious source of uncertainty in another way. Smaller particles contain a lower proportion of crystalline material because the amorphous surface layer makes up a higher proportion of the total.22, 23 The material on calibration filters is sizeselected from the standard by the process of dispersal and sampling during filter preparation. As a consequence, it may not contain exactly the same proportion of crystalline X-ray powder diffractometer 26 An X-ray powder diffractometer with reflection geometry and X-ray generator are required. Diffractometers with Bragg-Brentano semi-focusing geometry with Cu- or Co- target X-ray tubes are in general use. Dust cloud generator 27 A dust cloud generator of some kind is required for the generation of atmospheres containing quartz or cristobalite dust. Figure 1 shows a simple design constructed from 3 mm thick borosilicate glass, apart from the top which is made of 5 mm thick clear perspex. A 2 mm deep groove is machined into the top which mates with the rim of the main glass chamber to form a closed system. Four symmetrically spaced holes, through each of which a cyclone sampler can be suspended, are drilled in the top. Balance 28 A micro-balance calibrated against a primary standard and capable of weighing to within 1 pg over the range 0-20 mg for the preparation of standard filters for calibration. An electrostatic eliminator for use when weighing filters. REAGENTS 29 No 'reagents' in the strict sense are required in this method since the determination is carried out directly on the collected sample without chemical or physical treatment. Figure 1: Glass chamber for the preparation of calibration standards 3 material as the untreated standard. An ideal standard for XRD would be such that its crystalline content is unaltered in the preparation of calibration filters or is modified to a known extent. In practice, no such information is available for current standards and it is necessary to assume that the crystalline content of material on calibration filters is the same as in the bulk standard. The analytical results may be subject to bias to the extent that this assumption is uncertain. using dispersive IR; pre-scanned when using FTIR), and label each sampler with a unique identification number. 40 Set the volumetric flow rate on each sampler to the appropriate value (see MDHS14/316), to an accuracy of 5 %. Collection of samples 33 A satisfactory calibration procedure is to generate an atmosphere containing the standard quartz or cristobalite and sample from this through the respirable dust sampler onto filters. 34 The US National Institute of Standards and Technology (NIST) offers a certified respirable a-quartz standard (Standard Reference Material (SRM) 1878) of known crystalline quartz content which is suitable for calibration when analysing by infrared spectroscopy or X-ray powder diffraction. Sikron F600 (HSE standard quartz A9950) is also a suitable standard and this can be obtained from the Health and Safety Laboratory. It does not differ significantly from SRM 1878 when used to calibrate for direct on-filter analysis by XRD.24 Alternatively, it is permissible to use any other documented standard or a secondary standard which has been compared with a primary standard using the same analytical method. 35 The US National Institute of Standards and Technology (NIST) offers a certified respirable a-cristobalite standard (Standard Reference Material (SRM) 1879) of known crystalline cristobalite content which is suitable for calibration when analysing by infrared spectroscopy or X-ray powder diffraction. Alternatively, it is permissible to use any other documented standard or a secondary standard which has been compared with a primary standard using the same analytical method. Drift correction sample 36 In X-ray diffraction, an aluminium plate or any other suitable stable robust material can be used as an external standard to correct for the gradual decline in X-ray tube emission. Such a standard should be fine-grained, free from marked texture and have a strong X-ray diffraction peak in roughly the same 20 range as the quartz or cristobalite peaks being used for analysis. SAMPLING 37 Personal samples for respirable crystalline silica are collected as described in MDHS14/3.16 Preparation of sampling equipment 41 Attach the sampler to the wearer, preferably on the lapel, and as close to the mouth and nose as possible.17 Cyclone samplers are not generally sensitive to orientation, but should be attached to the wearer with the grit-pot at the base. Attach the pump to a suitable belt or harness so that it causes minimum inconvenience to the wearer, and safely secure any tubing connecting the pump and sampler. 42 For each sampler, carefully record the sample identity and all relevant sampling data (see Appendix A, page 13). 43 To begin sampling, remove the protective cover (if any) from the sampler and switch on the pump. Record the time and volumetric flow rate at the beginning of the sampling period. Where the pump is fitted with an integral timer, ensure that this is reset to zero. 44 At the end of the sampling period, again record the volumetric flow rate and the time, and calculate the duration of the sampling period. If the pump is fitted with an integral timer, check that the indicated period agrees with the calculated period. Consider the sample to be invalid if the two sampling times differ by more than 5% or if the volumetric flow rate at the end of the sampling period differs from that at the beginning by more than 0.1 litre/min. 45 Retain as blanks one unused loaded sampler from each batch of ten prepared; a minimum of three blanks should always be kept. Treat these as far as possible in the same manner as those actually used for sampling, but do not draw air through them. Background (fixed position) sampling 46 In use, the personal samplers should be mounted at approximately head height, away from obstructions, fresh air inlets or strong winds.The sampling procedures are otherwise the same as for personal sampling. It is not appropriate to compare fixed point (background) samples with the exposure limit. Fixed-position samples may be useful in identifying the main source(s) of crystalline silica exposure. Comparison of airborne concentration measurements from personal and fixed point samples may give some indication of the extent to which exposure arises from local or general conditions. 38 Clean the samplers before use. Dismantle the parts that come into contact with dust (referring to the manufacturer's instructions where necessary), soak in detergent solution, rinse thoroughly with water, and allow to dry before reassembly. 39 Using clean flat-tipped tweezers, load each sampler with a filter (pre-weighed, in order to match blanks, when Sampling time 47 A long sampling time ensures a heavier deposit of dust on the filter, thus reducing measurement inaccuracies. Sampling times should therefore be as long as is reasonably practicable (preferably not less than four hours), and should be representative of the working periods of the individuals being monitored. If the dust concentration is so high that a single filter would be overloaded, several filters may be used consecutively. 48 Further advice on sampling procedure is given in Guidance Note HSG173.17 Transport 49 Remove the filter from the sampler using flat-tipped tweezers, place it in an airtight tin and close with a lid. Take particular care to prevent material being dislodged from the filter. Transport the samples to the laboratory in a container capable of preventing damage in transit, and labelled to ensure proper handling. In some types of sampler, the filter is in a cassette whose inlet and outlet openings are closed with a plastic clip for transport to the laboratory where the filter may then be removed. CALIBRATION Preparation of calibration standards 55 When the required filters have been loaded with standard, allow them and the blanks to equilibrate in the balance room and re-weigh. Use the weights of the blank filters to determine a 'blank correction' for adsorption/ desorption of moisture and apply it to the weight increases of the loaded filters to obtain the mass of standard on each. Calibration of spectrophotometer/diffractometer 56 Calibration of the infrared spectrophotometer and X-ray diffractometer are discussed separately below. ANALYSIS BY INFRARED SPECTROSCOPY Scope 57 The infrared method is suitable for the determination of quartz/cristobalite over the range 10 g.g to 1 mg on a 25 mm filter. Over this range, there is a linear relationship between infrared response and quartz/cristobalite content of the sample. 50 Take 30 filters, ideally from the same box, and place separately on a clean surface in an oven at 50C for two hours. 51 Leave overnight in sample tins with lids left slightly ajar in the balance room to equilibrate with the atmosphere. Remove each filter in turn, weigh to the nearest g.g, record the weight and return the filter to its tin. 52 Number the tins in ascending order of filter weights, and select five or six filters as reference blanks such that each standard has a reference blank filter within 200 g.g of its weight. If using a Fourier Transform InfraRed spectroscopy (FTIR), scan the clean filters and save the resulting spectra on disk so that each filter can later be used as its own reference. 53 Use the remaining filters to prepare a series of standards by sampling from an airborne cloud of standard quartz or cristobalite generated in a dust cloud chamber (Figure 1). Safety note: Adequate means must be provided to prevent the glass chamber becoming dangerously pressurised during use when compressed air is applied to the side arms. The dust cloud generator must be placed inside a fume cupboard while atmospheres of quartz or cristobalite are being produced and sampled. 54 Place about 0.1 g of the standard quartz or cristobalite in the bowl. Attach a sampling pump to each of four cyclone samplers and set their volumetric flow rates to the required value. Fix the cyclone samplers to the lid of the chamber and place it on top. Apply a jet of compressed air to the side arms of the bowl for a few seconds. Allow approximately one minute for the coarser particles to settle out from the dust cloud. Run each sampling pump for sufficient time, typically 5-20 sec, to obtain filters loaded with the required amounts of standard covering the range 20-500 g.g by trial and error. A minimum of five or six filters is required to give a reliable calibration. 58 The lower limit of detection and the accuracy of the method are dependent upon a number of parameters, for example: particle size, other constituents, type of spectrophotometer. Interferences 59 Several minerals which may occur in conjunction with quartz and cristobalite absorb infrared radiation in the region of the quartz absorbances at 800 and 780 cm-1 and the cristobalite absorbances at 800 and 620 cm-1, giving rise to positive interferences. Infrared spectra of quartz and cristobalite and the most common of these interfering phases are shown in Figure 2. Positive identification, however, can only be made by comparison with a standard reference spectrum of the mineral suspected to be an interfering phase. On-filter mixtures containing the mineral and quartz/cristobalite can be prepared using the same technique as for standard quartz/cristobalite samples, and the effect of the interference on the quartz/cristobalite absorbances assessed. Using software techniques, it is possible to subtract any interfering bands from the infrared spectrum of a sample. Re-radiation effects 60 Some dark-coloured samples (eg those containing graphite or magnetite) can act as black body radiators while in the spectrophotometer beam. In older dispersive instruments, this re-radiation effect25 could result in up to 75% attenuation of the quartz absorbances at 800 and 780 cm-1. This problem could be overcome by the use of a spectrophotometer equipped with an asynchronously double chopped optical system, or, in lower-cost single chopped instruments, by the use of germanium cuton/blocking filters in the sample and reference beams.25 The response of an FTIR instrument is unaffected by re radiation. Here, the sample is placed between the modulated source radiation and the detector; re-radiation is unmodulated and not detected. 5 Absorbance Frequency (cm-1) Frequency (cm-1) Absorbance Absorbance Absorbance Frequency (cm-) Frequency (cm-1) Frequency (cm-1) Frequency (cm-1) Figure 2: Infrared spectra of quartz, cristobalite and interfering phases 6 Concentration of respirable crystalline silica in air X-ray diffractometry 78 Calculate the airborne concentration of quartz or cristobalite (mg/m3 or its equivalent ^g/litre) as the mass on the filter (g.g) divided by the air volume (litres). Calculate the airborne concentration of respirable crystalline silica by adding the concentrations of quartz and cristobalite if both are present. ANALYSIS BY X-RAY DIFFRACTION Scope 81 The diffractometer operating conditions should be chosen: (a) to maximise intensity even if this leads to some loss of resolution; (b) to minimise errors arising from particle statistics (largest practicable divergence and receiving slits; sample spinner). An example of suitable operating conditions is given in Table 1. 79 The X-ray diffraction method is suitable for the determination of quartz and cristobalite in respirable dust samples weighing up to about 2 mg when deposited on a 25 mm diameter filter. If the layer of dust on the surface of the filter is thin enough for X-ray absorption to be negligible, the diffracted intensity is directly proportional to the mass of quartz/cristobalite on the filter. At high dust loadings the layer of dust is thicker, the observed diffracted intensity is reduced by absorption effects and the linear relationship no longer holds. In practice, for dust samples from typical industrial environments, no significant deviation from linearity is found for dust loadings up to about 2 mg. Occasionally, samples with a high mass absorption coefficient for the radiation used may be encountered, eg an iron oxide matrix with CuKa radiation, and the deviation from linearity may begin at filter loadings as low as 0.5 mg. In such cases, the dust loading on the filter should be reduced by sampling for a shorter time. Methods exist to correct for these matrix absorption effects26, 27 but they make the analysis more complicated and may introduce significant overcorrection.28 80 Empirical observation suggests that the linear relationship between diffracted intensity and mass may not hold when the mass of dust is very low, ie close to the origin. The extrapolated straight line calibration does not usually pass through the origin (Figure 3) leading to uncertainty in the detection limit and measurements close to it. Diffracted intensity Table 1: Example of operating and data collection conditions for X-ray diffractometry Diffractometer operating conditions 2.7kW broad focus copper anode X-ray tube run at 50kV, 45mA 1 divergence and scatter slits (used over whole range to avoid changes during data collection) (An automatic divergence slit is suitable if fitted) Receiving slit: 0.3 mm Diffracted-beam graphite monochromator Sample spinner Automatic sample changer Data collection Qualitative scan 6 to 70 20 continuous scan at 1/min or 0.02/sec Quartz 25.65 to 27.65 20 in 0.05 steps, counting for 10 sec at each step or 0.02 steps, counting for 4 sec at each step Quartz 19.90 to 21.90, 49.10 to 51.10 and 59.00 to 61.00 20 in 0.05 steps, counting for 30 sec at each step or 0.02 steps, counting for 12 sec at each step Cristobalite 21.00 to 23.00, 30.45 to 32.45 and 35.10 to 37.10 20 in 0.05 steps, counting for 30 sec at each step or 0.02 steps, counting for 12 sec at each step Figure 3: Supposed behaviour of the XRD calibration line near the origin System calibration 82 Diffraction data are collected for the calibration standards for quartz and cristobalite in the same way as for sample analysis (see below). Since no interfering phases will be present on the calibration standard filters, a qualitative scan is unnecessary. 8 11 Silica Construction Information Sheet CIS36(rev1) HSE Books 1999 12 Occupational exposure limits: Containing the list of maximum exposure limits and occupational exposure standards for use with the Control of Substances Hazardous to Health Regulations 1999 Environmental Hygiene Guidance Note EH40 HSE Books 2002 Supplied with Occupational exposure limits: Supplement 2003 Environmental Hygiene Guidance Note EH40/2002 HSE Books 2003 ISBN 0 7176 2194 4; and Occupational exposure limits: Supplement 2003 EH40/2002 HSE Books 2003 ISBN 0 7176 2172 3 25 Foster RD and Walker RF Sample re-radiation effects in the quantitative analysis of crystalline silica in foundry samples by infrared spectrophotometry Analyst 1981 106 1240-1242 26 Altree-Williams S, Lee J and Mezin NV Quantitative X-ray diffractometry on respirable dust collected on Nucleopore filters Ann Occup Hyg 1977 20 109-126 27 National Institute for Occupational Safety and Health Silica, crystalline, respirable: method 7500 NIOSH Manual of Analytical Methods 3rd Edition Volume 2 1984 13 Summary criteria for occupational exposure limits Environmental Hygiene Guidance Note EH64 HSE Books 1996-2000 ISBN 0 7176 2073 5 and EH64 Supplement HSE Books 2001 ISBN 0 7176 2070 0 and 2002 HSE Books Supplement ISBN 0 7176 2372 6 14 Respirable crystalline silica Chemical Hazard Alert Notice 35 HSE website: http://www.hse.gov.uk/pubns/CHAN35.htm 15 BS EN 482: 1994 Workplace atmospheres - General requirements for the performance of procedures for the measurement of chemical agents European Standard British Standards Institution 16 General methods for sampling and gravimetric analysis of respirable and inhalable dust MDHS14/3 2000 ISBN 0 7176 1749 1 (available free from HSE website: http://www.hse.gov.uk/pubns/mdhs/index.htm) 17 Monitoring strategies for toxic substances HSG173 HSE Books 1997 ISBN 0 7176 14115 28 Anderson CC Collaborative tests of two methods for determining free silica in airborne dust Final report Contract No 210-79-0059 National Institute for Occupational Safety and Health 1983 29 Jeyaratnam M and West NG The determination of acristobalite in airborne dust by X-ray diffraction - theory and practice Advances in X-ray Analysis 1989 32 585-592 30 Biggins PDE The selection of a suitable cristobalite standard to be used in the analysis of airborne dust for cristobalite Report No X.191 (Revised) British Cast Iron Research Association 1986 31 Perrotta AJ, Grubbs DK, Martin ES, Dando NR, McKinstry HA and Huang C-Y Chemical stabilisation of a-cristobalite J Am Ceram Soc 1989 72 441-447 32 Young J, Rea MS and Briggs G The non-formation of a-cristobalite in devitrified commercial-grade aluminosilicate refractory ceramic fibre Br Ceram Trans J 1989 88 58-62 18 BS EN 1232: 1997 Workplace atmospheres. Pumps for personal sampling of chemical agents. Requirements and test methods British Standards Institution 33 Butler MAand Dyson DJ The quantification of different forms of cristobalite in devitrified aluminosilicate ceramic fibres J Appl Cryst 1997 30 467-475 19 Foster RD and Walker RF Quantitative determination of crystalline silica in respirable-size dust samples by infrared spectrophotometry Analyst 1984 109 117-1127 20 Shinohara Y Direct quantitative analysis of respirable cristobalite on filter by infrared spectrophotometry Industrial Health 1996 34 25-34 21 Elton NJ, Salt PD and Adams JM Determination of quartz in kaolins by X-ray powder diffractometry Anal Chim Acta 1994 286 37-47 22 Gordon RL and Harris GW Effect of particle size on the quantitative determination of quartz by X-ray diffraction Nature 1955 176 1135 23 Tossavainen A Determination of quartz on membrane filters by X-ray diffraction Scand J Work Environ & Health 1979 5 379-383 24 Jeyaratnam M and Nagar N Comparison of a-quartz standard Sikron F600 (HSE A9950) with the NIST respirable a-quartz standard SRM 1878 for bulk and onfilter analysis Ann Occup Hyg 1993 37 379-383 34 Dyson DJ, Butler MA, Hughes RJ, Fisher R and Hicks GW The de-vitrification of alumino-silicate ceramic fibre materials - the kinetics of the formation of different crystalline phases Ann Occ Hyg 1997 41 561-590 35 Nagelschmidt G Inter-laboratory trials on the determination of quartz in dusts of respirable size Analyst 1956 81 210-219 36 Edmonds JW, Henslee WW and Guerra RE Particle size effects in the determination of respirable a-quartz by X-ray diffraction Anal Chem 1977 49 2196-2203 37 Kohyama N A new X-ray diffraction method for the quantitative analysis of free silica in the airborne dust in working environment Industrial Health 1985 23 221-234 38 Bhaskar R, Li J and Xu LA comparative study of particle size dependency of IR and XRD methods for quartz analysis Amer Ind Hyg Assoc J 1994 55 605-609 39 Pickard KJ, Walker RF and West NG A comparison of X-ray diffraction and infra-red spectrophotometric methods for the analysis of a-quartz in airborne dusts Ann Occup Hyg 1985 29 149-167 14