PII: S0003-4878(00)00083-1

Ann. occup. Hyg., Vol. 45, No. 6, pp. 429435, 2001 2001 British Occupational Hygiene Society Published by Elsevier Science Ltd. All rights reserved Printed in Great Britain. 0003-4878/01/$20.00

A Comparison of International Silica (a-Quartz) Calibration Standards by Fourier Transform Infrared Spectrophotometry
DAVE K. VERMA* and DON S. SHAW
Occupational and Environmental Health Laboratory, McMaster University, 1200 Main Street West, Hamilton, Ontario L8N 3Z5, Canada
Downloaded from http://annhyg.oxfordjournals.org/ by guest on April 17, 2014

Seven international silica (a-quartz) standards were examined for relative purity to the US National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 1878 Respirable a-quartz by Fourier TransformInfrared Spectrophotometry (FTIR). The standards examined have been used in North America, the UK, Australia and Germany. The 189 samples analyzed included NIST-SRM 1878, Min-U-Sil 5, Ottawa Silica Sand, Sikron F600, A9950 (AUST 1), DQ12-Robock, DQ12-Bergbau. Size distributions of the standards were determined by Coulter Counter to be broadly similar with equivalent spherical volume median diameter ranging between 1.2 and 3 mm. The results showed the standards to differ by as much as 30% in relative purity. Consequently, an internationally agreed upon calibration standard is urgently needed. Min-U-Sil 5 based NIST-SRM 1878 or Sikron F-600 are the two most likely candidates. Any agreed standard must have a well characterized size distribution and closely match the respirable dust criteria. It should also be studied by both infrared spectrophotometry and X-ray diffraction techniques. 2001 British Occupational Hygiene Society. Published by Elsevier Science Ltd. All rights reserved Keywords: respirable silica; silica analysis; calibration standard; Fourier transforminfrared spectrophotometry; particle size

INTRODUCTION

Occupational exposure to crystalline silica (-quartz) occurs in mining, manufacturing and construction industries. It is well known that prolonged inhalation of dust containing -quartz can cause silicosis a brotic lung disease (Anon, 1975). The International Agency for Research on Cancer (IARC) has recently revised the classication of silica to Group 1 Human Carcinogen (IARC, 1997). The American Conference of Governmental Industrial Hygienists (ACGIH) has also classied silica as a category A2 carcinogen, i.e. a suspected human carcinogen (Anon, 1999). It has been reported that the variable health effects, in terms of brogenesis and carcinogenesis, are related to the inherent characteristics of crystalline silica determined by the origin of the sample and

Received 20 March 2000; in nal form 12 September 2000. *Author to whom correspondence should be addressed. Tel.: +1-905-525-9140; fax: +1-905-528-8860; e-mail: vermadk@mcmaster.ca

external factors, i.e. association with or contamination by substances other than silica which might activate or blunt its brogenicity and carcinogenicity (Fubini, 1998; Donaldson and Borm, 1998). It has been earlier suggested by Altree-Williams (1982) that variable crystallinity may be related to brogenicity and he hypothesized that the Crystallinity Index (CI) based on an X-ray powder diffraction measurement may be related to the specic toxicity and brogenesis of quartz. The epidemiological evidence of silica, silicosis and lung cancer have also been recently reviewed (Soutar et al., 2000; Hessel et al., 2000; Finkelstein, 2000). Finkelstein concluded that the risk of silicosis following a lifetime exposure at the current OSHA (USA) standard of 0.1 mg/m3 is likely to be at least 5-10% and the lung cancer risk is likely to be increased by 30% or more. On the other hand, Hessel et al. (2000) dispute the validity of IARCs designation of silica as a Group 1 carcinogen, stating that data on humans demonstrates a lack of association between lung cancer and exposure to crystalline silica.

429

430

D. K. Verma and D. S. Shaw

A doseresponse study of silica and silicosis among Ontario hardrock miners was conducted (Verma et al., 1989) using the potassium bromide disc method of silica analysis using Infrared Spectrophotometry (IR) as described by Dodgson and Whittaker (1973). The method, with minor modications, was also issued as NIOSH Method 7602. As part of the epidemiological study an investigation was undertaken to compare the measurement of -quartz by Xray diffraction (XRD) and IR (Verma et al., 1992). In that study, it was noted that there may be signicant differences in the quantied results obtained by using various -quartz calibration standards in these analytical techniques. At the time of the study, in the late eighties, there was no universally accepted international -quartz standard for use with these analytical methods. Various methods of silica analysis have been critically reviewed. An important early review discussed the issues related to the commonly used method (Anderson, 1975). Since then, another major review updated this work and identied the present and future needs (Madsen et al., 1995). The issues and controversy in the measurement of crystalline silica was a topic of an international conference, summarized by Miles (Miles, 1999). For risk estimates, metaanalysis of doseresponse studies are often conducted where dose measured in various studies are combined. The silica dose measured in different studies could be different for several reasons, one being the purity of calibration standards. As noted earlier, no internationally accepted consensus for an -quartz standard exists. If there were such a standard, intermethod variability due to different standards could be eliminated. A few studies, all of which were conducted in the early 1980s, compared the then available -quartz standards by XRD and or IR (Dewell and Ambidge, 1980; Altree-Williams et al., 1981; Biggins, 1982). These studies recommended the need for further comparative studies and advocated the development of a consensus international standard. In Canada, Ottawa Silica Sand (a product of Ottawa, Illinois, USA) was used as the -quartz standard in the Province of Ontario, whereas in the rest of Canada and across North America, generally Min-U-Sil 5 was used (Verma et al., 1992). This study was undertaken to compare seven international -quartz standards by FTIR for their relative purity. The purity of quartz calibration standard is related to crystallinity, particle size, presence and absence of an amorphous layer and other impurities. Ideally, an absolute pure standard will be 100% crystalline with no amorphous layer and impurities and of the appropriate particle size. However, no such material exists. An -quartz calibration standard may not be 100% crystalline silica (-quartz) nor free from amorphous layers and impurities, but its analyti-

cal response can be assumed to be from a 100% pure -quartz. For this investigation NIST-SRM 1878, formerly known as NBS-SRM 1878, was assumed to be 100% pure. The analytical purity of all other -quartz standards have been measured compared to NISTSRM 1878 and referred to as relative purity. The specic objectives were: 1. To determine the size distribution of seven international -quartz standards and; 2. To compare the relative purity of the standards by using an FTIR.

MATERIALS AND METHODS

Agencies and researchers involved in silica analysis in the USA, Canada, the UK, Australia, Germany and Scandinavian countries were contacted and asked for details of the standards they use, and if possible, to supply samples of the standards. Our own laboratory has performed silica analysis since 1978 and had a supply of a number of standards used in North America. We were able to obtain seven standards used in various countries. These -quartz standards and their sources are listed in Table 1. Although some of these standards may have come from the same source (i.e. Sikron F-600 and A9950), they were treated as separate standards for this investigation. For the purpose of a more complete characterization of all seven standards, particle sizing was performed. Particle sizing was carried out by using a Coulter Counter Model TAII, (Coulter Corp., Miami, Florida) equipped with a 50-m diameter aperture, permitting particle sizing in the range of 120 m. A Coulter Counter Channelyzer, Model C1000 (Coulter Corp., Miami, Florida) was also used in conjunction with the Model TAII. These instruments were calibrated using L2 (2 m) and L5 (5 m) latex monospheres (Coulter Corp., Miami, Florida). All standards were analyzed using a Nicolet Model 510P FTIR, (Nicolet Instruments Corporation, Madison, Wisconsin, USA), using NIOSH Method 7602 (Anon, 1994). NIST-SRM 1878 was used as the silica calibration standard for the instrument, consequently all the bulk standards were analyzed with respect to the NIST standard. The bounds of the instruments calibration curve ranged from 5 to 500 g. No sample preparation of the standards was required prior to silica analysis and particle sizing except for Ottawa Silica Sand. Ottawa Silica Sand, as received, was a coarse river sand-like material. The coarse Ottawa Silica Sand was mixed and ground in a Spex Mixer Mill (Spex Industry Inc., Metuchen, NJ, USA) then sieved to a less than 10 m diameter precision sieve (ATM Corporation of Milwaukee, Wisconsin, USA). The less than 10 m fraction of sand was used for both particle sizing and silica analysis. Nominal weights selected for each quantied silica

Downloaded from http://annhyg.oxfordjournals.org/ by guest on April 17, 2014

International silica calibration standards

431

Table 1. List of international -quartz standards

Name
Min-U-Sil 5 NIST-SRM 1878 Ottawa Silica Sand

Origin
USA USA USA

Where Used
USA, Canada USA, Canada and other AIHA accredited laboratories Ontario, Canada UK and Germany Australia Germany Germany

Source
Pittsburgh Sand and Glass Company, USA and NIOSH in USA National Institute of Science and Technology, MD, USA Elliott Lake Laboratory, Energy Mines and Resources Canada (Geoffery Knight) Health and Safety Executive, UK (Dr P. A. Elwood) Stephen Altree-William, The Australian National University, and Bernard Jordan, Deakin University, Australia British Cast Iron Research Association, UK (Dr P. D. Biggins) British Cast Iron Research Association, UK (Dr P. D. Biggins)
Downloaded from http://annhyg.oxfordjournals.org/ by guest on April 17, 2014

Sikron F-600 (also labelled Germany by HSE as A9950) A9950 (AUST) No 1 (subset Germany of UK A9950 derived from Sikron F-600) DQ12-Robock Germany DQ12-Bergbau Germany

standard were 20, 50, 100, 150, 200, 250, 300, 350 and 400 g. Each standard was mixed and ground with 250 mg of potassium bromide (KBr) in a Vibromill (Beckman Coulter Inc., Fullerton, California), and pressed in an evacuable 13-mm diameter die (Beckman Coulter Inc., Fullerton, California) at 8 t pressure to form a pellet. Three individual quantitative FTIR analyses were performed by rotating each pellet in 120 increments through 360. The mean value of the three rotations was used for each data point. Primary quantication of the characteristic crystalline -quartz doublet was based upon the peak height absorbance at 798 cm1, with the less sensitive secondary singlet located at 696 cm1 used as a conrmatory tool. Three pellets were prepared from each standard and analyzed at the appropriate nominal weights. The order in which the series of nominal weights were analyzed was randomized such that the series of 20-g nominal weight standards did not necessarily precede the 50-g nominal weight series. A total of 189 samples were analyzed.
RESULTS

The results of size distribution by Coulter Counter of the standards are given in Table 2 and shown graphically in Fig. 1. The volume median diameters of all seven standards were found to be between 1.2 and 3 m. The analytical results of all seven calibration standards obtained by FTIR, shown in Fig. 2, are plotted on the Y axis with NIST-SRM 1878 as the Gold Standard on the X-axis. The equation of the line for each standard, based on 27 pairs of observations are also shown in Fig. 2.
DISCUSSION

standards are therefore important. It has been shown that as particle size increases XRD peak height also increases but IR peak height decreases (Bhaskar et al., 1994). This effect was also reported by Addison (1991) with coal mine dust samples. It should be noted that mineralogical qualities other than crystallinity and size distribution also have an effect. Amorphous silica, probably as a disordered shell around a crystalline core of comminuted quartz, has the effect of reducing the absorbance coefcient while the normal effect of particle size reduction is to increase it (Duyckaerts, 1959). The ideal particlesize distribution of the calibration standard should match the sample being analyzed as closely as possible. In the case of the respirable dust sample, which consists predominantly of ne particles less than 10m aerodynamic diameter with a median aerodynamic diameter cut point of 3.54 m, the calibration standards should also be composed of ne particles. The particle-size distributions of the seven standards were found to be approximately log-normally distributed with median equivalent spherical volume diameter ranging between 1.2 and 3 m, as shown in Table 2 and Fig. 1. The volume median diameter and mass median diameter of the log-normally distributed particle size distribution can be considered approximately the same (Reist, 1984). The volume median diameter given by the Coulter Counter can also be converted to aerodynamic mass median diameter. Ogden et al. (1983) have shown that for coal particles, equivalent spherical volume diameter obtained by the Coulter Counter is equivalent to aerodynamic diameter. By knowing the density and the shape factor, the equivalent spherical volume diameter can be converted to aerodynamic diameter as follows: da =

Analytical methods for silica such as IR and XRD are particle-size dependent. The particle-size distribution of samples to be analyzed and the calibration

s K
0

s d2 s

432

D. K. Verma and D. S. Shaw

Table 2. Cumulative particle size distribution of silica standards, analyzed by Coulter Counter (particle-size as equivalent spherical volume diameter)
Normalized cumulative % total volume
Particle diametera
0.6 0.8 1.0 1.3 1.6 2.0 2.5 3.2 4.0 5.0 6.4 8.0 10.1 12.7 16.0 20.2
a

Ottawa Silica Min-U-Sil 5 Sand

1 5 10 17 25 35 45 56 70 82 90 92 96 98 99 100 0 6 16 32 49 68 83 92 96 98 99 99 99 99 100 100

NIST SRM Sikron F-600 1878

0 7 17 33 51 70 82 88 91 92 93 93 93 98 99 100 1 4 9 15 23 34 47 62 74 83 88 90 92 93 99 100

A9950

DQ12Bergbau
0 7 17 31 45 60 72 82 89 95 97 97 98 98 99 100

DQ12Robock
1 16 38 58 66 70 73 75 78 79 83 84 89 92 99 100

0 5 12 23 37 51 65 76 84 90 93 95 96 97 99 100

Downloaded from http://annhyg.oxfordjournals.org/ by guest on April 17, 2014

Equivalent spherical volume diameter.

Fig. 1. Cumulative particle size distribution (as equivalent spherical volume diameter) by Coulter Counter.

where: da ds K s s0 is aerodynamic diameter, is equivalent spherical volume diameter, is shape factor, is density of the particle, and is 1 g/cm3 (unit density sphere for aerodynamic particle).

dards of 1.6 m would convert to aerodynamic median diameter of 2.18 m as: da =

2.6 (1.6)2 = 2.18 1 1.4

By substituting a K value of 1.4 and density of 2.6 mg/cm3 for silica from the literature (Mercer, 1973) in the equation above, the median equivalent spherical volume diameter of the NIST-SRM 1878 stan-

Likewise Sikron F-600s median diameter of 2.7 m would convert to an aerodynamic median diameter of 3.68 m. These standards provide a good match for the respirable dust curves (50% acceptance at 3.5 m old ACGIH curve, 50% at 5 m BMRC curve and 50% at 4 m currently recommended by

International silica calibration standards

433

Downloaded from http://annhyg.oxfordjournals.org/ by guest on April 17, 2014

Fig. 2. Relationships between NIST-SRM 1878 and other standards as analyzed by FTIR: (a) Min-U-Sil 5 (b) Ottawa Silica Sand (c) Sikron F-600 (d) A9950 (AUST 1) (e) DQ12-Robock (f) DQ12-Bergbau.

the European and international Committees as well as the ACGIH). In fact, Sikron F-600 would provide a closer match to the newly agreed European Community and ACGIHs respirable dust criteria of 50% at 4 m. It should be noted that the size distribution of NIST-SRM 1878 by the Coulter Counter gave an equivalent spherical volume median diameter of 1.6 m, which could also be considered as the mass median diameter for the log-normally distributed aerosol. Size distribution information provided with the NIST-SRM 1878 standard, as determined by a sedimentation method, was found to be between 0.33 and 5 m with a mass mean spherical diameter of 1.62 m (Anon, 1983). The Coulter Counter result of 1.6 m was thus remarkably similar to that obtained by the Sedimentation method. The volume median diameter of other standards given in Table 2 could thus be compared to each other with assurance. Biggins 1982 examined the particle size distribution of eight -quartz standards namely; X-7488, Bahai, Fyle, Min-U-Sil 5, Min-U-Sil 10, Min-U-Sil 15, DQ 12, DQ 120. The mass median diameter of

these standards ranged from 1.85 to 5.15 m. However, many of the standards examined by Biggins are no longer in use or not available. NIST-SRM 1878 is nearly equivalent to Min-U-Sil 5, as shown in Fig. 2(a), which is not surprising since NBS-SRM 1878 was likely made from the latter. The ground Ottawa Silica Sand standard shows a marked deviation in purity, relative to NIST-SRM 1878, of approximately 30% [Fig. 2(b)]. Sikron F-600 had a difference of approximately 20% [Fig. 2(c)] relative to NIST-SRM 1878. A9950 (Aust 1) was also nearly equivalent to Sikron F-600 [Fig. 2(d)], probably due to the fact that A9950, used in UK, was likely derived from Sikron; this was the standard A9950 (Aust 1) used in Australia. DQ12-Robock is equivalent to NIST-SRM 1878 [Fig. 2(e)] while DQ12-Bergbau [Fig. 2(f)] was nearly equivalent to the NIST standard. Addison (1991) investigated the accuracy and precision of coal mine dust sample analysis within the European Community with the aim of harmonization of standard reference materials (i.e. -quartz calibration standard) and analytical methods. Thirty dif-

434

D. K. Verma and D. S. Shaw Peter Eller of USA. We thank Lorraine Shaw and Tracey Tuttle for their help in preparing this manuscript.

ferent samples from three UK coal mines, with widely different mineral compositions and particle size ranges, were analyzed by XRD and IR in ve European laboratories, one French, two German, one Belgian and one from the UK. Three laboratories employed DQ12 as the calibration standard, one used French IR87 and another used A9950. The study concluded that particle size and mineralogical interferences can contribute to differences of up to 30% in quartz content, despite efforts to control for both these factors. Since the early studies of Dewell and Ambidge (1980), Altree-Williams et al. (1981) and Biggins (1982), to the best of our knowledge, only one study has been published comparing Sikron F-600 to NISTSRM 1878 (Jeyaratnam and Nagar, 1993). In that study, Jeyratnam and Nagar reported that Sikron F600, which was assumed to be 100% -quartz in comparison to NIST-SRM 1878, should be taken as containing 92.5% -quartz for the bulk sample and 95.5% -quartz for on-lter analysis. Our results indicate the Sikron F-600 to be about 80% of NIST-SRM 1878, when analyzed by FTIR. This difference could be because of the use of the an alternate technique (i.e. IR as opposed to XRD) by Jeyaratnam and Nagar (1993). It was demonstrated by Dewell and Ambidge (1980) and Biggins (1982) that the crystallinity of -quartz standards could in fact be different and vary signicantly in their XRD and IR responses, as subsequently observed by Verma et al. (1992). It is therefore important that calibration standards should be evaluated for their relative purity by both IR and XRD methods. High inter-laboratory variability exists in the US Prociency Analytical Testing (PAT) program for silica, where as many as 12 different -quartz calibration standards have been used by the participating laboratories. The same silica exposure sample given to two different laboratories could result in silica concentrations which may differ by a factor of 2 or more (Eller et al., 1999a,b). A similar high inter-laboratory variability has been observed in the UKs Workplace Analysis Scheme for Prociency (WASP) program for silica. A well characterized and commonly accepted silica calibration standard would be of signicant help in reducing some of the variablility. The results of our study indicate that the two possible candidates are Min-U-Sil 5 based NIST-SRM 1878 or Sikron F-600. The size distribution of Sikron F-600 matches the respirable dust criteria more closely than NIST-SRM 1878. However, it remains to be seen whether Sikron F-600 can be obtained in sufcient quantity to satisfy this need.
AcknowledgementsWe are grateful to our colleagues in North America, Australia, the UK and Europe who generously provided us with the information and samples. In particular, we thank Dr P. D. Biggins and Dr P. Elwood of UK, Mr Stephen Altree-Williams and Mr Bernard Jorden of Australia, and Dr

REFERENCES Anon. Criteria for recommended standard: occupational exposure to silica. NIOSH (HEW) publication #75-120. Cincinnati (OH): NIOSH, 1975. Anon. Certicate of analysis. Standard reference material 1878: respirable alpha quartz (quantitative X-ray powder diffraction standard). Washington (DC): NIST (previously known as National Bureau of Standards), 1983. Anon. Silica crystalline by IR method #7602, In: Eller PM, editor. NIOSH manual of analytical methods, DHHS (NIOSH) publication #94-113. 4th edn. Cincinnati (OH): NIOSH, 1994. Anon. Threshold limit values for chemical substances and physical agents and biological exposure indices. Cincinnati (OH): ACGIH, 1999. Addison J. Improvements of analysis of mineral component of coal mine dusts. Institute of Occupational Medicine report. TM/91/10, Edinburgh, 1991. Altree-Williams S. Quartz: correlation between crystallinity index and brogenicity ([letter to Editor]). Ann Occup Hyg 1982;25:4558. Altree-Williams S, Byrnes JG, Norman MB. Reference material for the quantitation of -quartz in respirable dust by X-ray diffraction. Ann Occup Hyg 1981;24:34756. Anderson PL. Free silica analysis of environmental samples a critical review. Am Ind Hyg Assoc J 1975;36:76778. Bhaskar R, Li J, Xu L. A comparative study of particle size dependency of IR and XRD methods for quartz analysis. Am Ind Hyg Assoc J 1994;55:6059. Biggins PDE. Standard quartz samples compared by X-ray powder diffractometry and infrared analysis (BCIRA Report #1472). BCIRA J 1982;30:23553. Dewell P, Ambidge PF. choice of standard quartz for X-ray and infrared analysis of respirable dust samples for quartz content (BCIRA Report #1399). BCIRA J 1980;28:41117. Dodgson J, Whittaker W. The determination of quartz in respirable dust samples by infrared spectrophotometry I. The potassium bromide disc method. Ann Occup Hyg 1973;16:37387. Donaldson K, Borm PJA. The quartz hazard: a variable entity. Ann Occup Hyg 1998;42:28794. Duyckaerts G. The infrared analysis of solid substances a review. Analyst 1959;84:20114. Eller PM, Feng HA, Song RS, Key-Schwartz RJ, Esche CA, Groff JH. Prociency analytical testing (PAT) silica variability, 19901998. Am Ind Hyg Assoc J 1999;60:5339. Eller PM, Key-Schwartz RJ, Song R, Edward SL, Schlecht PC. Silica method modication for improved inter-laboratory precision. The Synergist 1999;10(11):234. Finkelstein MM. Silica, silicosis and lung cancer: a risk assessment. Am J Ind Med 2000;38:818. Fubini B. Surface chemistry and quartz hazard. Ann Occup Hyg 1998;42:52130. Hessel PA, Gamble JF, Gee JB, Gibbs G, Green FH, Morgan WK, Mossman BT. Silica, silicosis and lung cancer: a response to a recent working group report. J Occup Envir Med 2000;42(7):70420. IARC. Silica, some silicates, coal dust and para aramide brils [68]. IARC monograph on the evaluation of the carcinogenic risk of chemicals to humans, Lyon (France), IARC, 1997. Jeyaratnam M, Nagar N. Comparison of -quartz standard Sikron F-600 (HSE A9950) with the NIST respirable quartz standard SRM 1878 for bulk and on-lter analysis. Ann Occup Hyg 1993;37:16779. Madsen FA, Rose MC, Cee R. Review of quartz analytical methodologies: present and future needs. Appl Occup Envir Hyg 1995;10:9911002.

Downloaded from http://annhyg.oxfordjournals.org/ by guest on April 17, 2014

International silica calibration standards Mercer TT. Aerosol technology in hazard evaluation. New York: Academic Press, 1973. Miles WJ. Issues and controversy: the measurement of crystalline silica; review papers on analytical methods. Am Ind Hyg Assoc J 1999;60:396402. Ogden TL, Barker D, Clayton MP. Flow-dependence of the Casella respirable-dust cyclone. Ann Occup Hyg 1983;27:26171. Reist PC. Introduction to aerosol sciences. New York: McMillan Publishing Company, 1984.

435

Soutar CA, Robertson A, Miller BG, Searl A, Bignon J. Epidemiological evidence on the carcinogenicity of silica: factors in scientic judgement. Ann Occup Hyg 2000;44:314. Verma DK, Sebestyen A, Julian JA, Muir DCF, Schmidt H, Bernholz CD, Shannon HS. Silica exposure and silicosis among ontario hardrock miners: II. Exposure estimates. Am J Ind Med 1989;16:1328. Verma DK, Sebestyen A, Julian JA, Muir DCF. A comparison of two methods of sampling and analysis of free crystalline silica -quartz. Appl Occup Envir Hyg J 1992;7:75863.

Downloaded from http://annhyg.oxfordjournals.org/ by guest on April 17, 2014