Silica

 Exposure Measure Amount Compare to Limit Protection

The following questions were received over time from industrial hygiene professionals in laboratories and the field. These include persons employed in foreign, federal, state and local governments; private laboratories; manufacturers; industry; construction; and insurance. If you have a question that does not appear in the list below, try using your browser's "Find" feature (usually a control F command) to locate key words on this page.

Evaluating Exposure Levels:

How does the OSHA Permissible Exposure Limit (PEL) of 10/(%Quartz + 2) compare to the American Conference of Government Industrial Hygienists' (ACGIH) old Threshold Limit Value (TLV) of 0.1 mg/m3?

Most OSHA PELs are set values for a single air contaminant such as cadmium or a related family of contaminants such as the polyaromatic hydrocarbons (PAHs). The PEL for respirable dust containing quartz differs considerably in that it is a function that varies between a value of 0.1 mg/m3(when the material is pure quartz) up to a value approaching the Particulates Not Otherwise Regulated (PNOR) OSHA PEL of 5 mg/m3. The PEL does not apply below 1% quartz, so the highest it can get is 3.3%. (When the concentration drops below 1%, the PNOR PEL of 5 mg/m3 applies.) It can be shown mathematically that when this PEL function for respirable dust containing quartz is divided into the TWA exposure, the resulting standardized concentration (or exposure severity) is the sum of the standardized concentrations for the separate quartz and PNOR exposures. This derivation requires only simple algebra, but is available in the literature [Frank J. Hearl, "Mixture Formula Justified," Letters to the Editor, AIHA Journal57, June 1996, p 575 and also in Frank J. Hearl, "Guidelines and Limits for Occupational Exposure to Crystalline Silica," in Silica and Silica-Induced Lung Diseases, V. Castranova, V. Vallyathan, and W. E. Wallace Eds., CRC Press Inc. pp.15-22.] [Note that the derivations count the quartz twice because it is not subtracted from the dust exposure in determining the PNOR exposure. If the quartz were subtracted, the quartz standard of 0.098 mg/m3 would result, which is only very slightly more restrictive than the 0.1 mg/m3 proposed in the (now vacated) Final Rule.] Through the same approach, it is easy to derive the PEL for the mixture of respirable dust containing quartz and cristobalite [PEL = 10/(%Quartz + 2(%Cristobalite) + 2)]. The use of these PELs achieves the same result as using the mixture calculation specified in 29 CFR 1910.1000 for exposures to substances having an additive effect on the body or target organ system.

How do I perform a crystalline silica mixture calculation when the sample contains both quartz and cristobalite?

The Silica eTool provides a discussion and example of the calculations. An Advisor Genius calculator is also available which will perform the calculations.

I am in the construction industry and can't locate any lab that analyzes impinger samples collected for quartz in terms of millions of particles per cubic foot (mppcf) using microscopy. How should I sample and analyze for quartz?

That methodology is obsolete. The impinger samples need to be counted within 24 hours. Errors associated with the analyses are unknown and the method does not provide information about the particulate composition. Separate samples must be taken to determine the % crystalline silica in the samples. Therefore,  the microscopy method is impractical. Instead, resample using the cyclone and collect on tared low-ash PVC filters. Analyze by XRD, IR, or by a Chemical Method as appropriate for your sample matrix and apply the equivalent OSHA PEL for respirable dust containing quartz, 10/(%Quartz + 2) in terms of mg/m3. For details, see Appendix F of Memorandum of OSHA Secretary Joseph A. Dear, May 2, 1996, Special Emphasis Program (SEP) for Silicosis.

How do I interpret a company's crystalline silica sampling data when it doesn't list the percentage quartz and all I have are the TWA exposure mg/m3 in terms of quartz and the corresponding TWA exposure mg/m3 for the mixture of respirable dust containing quartz? For example, say I'm told that the quartz exposure was 0.066 mg/m3 and the mixture exposure was 3.0 mg/m3.

There are two mathematically equivalent ways you can determine the standardized exposure of the mixture:
1. From the response to FAQ #1 above, one could obtain the sum of the standardized exposure to quartz at an exposure limit of 0.100 mg/m3 and the standardized exposure to the mixture at a PEL of 5.0 mg/m3. In the example exposures given above, the summed standardized exposure is 0.066/0.100 + 3.0/5.0 = 0.66 + 0.60 = 1.26. This method may be how the company tracks exposures. The practical advantage is that the component standardized exposures are determined separately, which may help in evaluating the engineering actions needed to bring the exposure to the mixture under control. The company may be using the 0.050 mg/m3 respirable quartz REL or TLV as a more restrictive exposure limit; it is still important that it consider the additive effect of the exposures to both quartz and PNOR.

2. For compliance purposes calculate the percentage quartz by taking 100 times the ratio of the TWA quartz exposure divided by the TWA mixture exposure. (The equal volumes of air in the concentration units cancel out leaving a weight ratio.) In the example exposures given above, the percentage quartz is 100 (0.066/3.0) = 2.2%. From this the PEL is 10/(2+2.2) = 2.38 and the standardized exposure would be 3.0/2.38 = 1.26, which is the same result as in (a) above.
Note: Calculating the standardized exposure both ways can provide a check on your math, but results from methods (a) and (b) may differ slightly due to round off of intermediate results.

How do I use the percentage quartz results from the analysis of the bulk samples to determine whether the OSHA PEL for respirable dust containing quartz has been exceeded on the air samples?

Please Don't. Analytical results on the quartz content of the air samples are necessary to evaluate whether the OSHA PEL is exceeded. Results for bulk materials do not necessarily represent the composition of the respirable dust suspended in the air at the time of sampling. Bulk materials are analyzed primarily to establish the potential for exposure or to assist the laboratory when interferences are encountered or may be present in the air samples. See Memorandum of OSHA Secretary Joseph A. Dear, May 2, 1996, Special Emphasis Program (SEP) for Silicosis, for recommendations on collection of bulk samples.

We need OSHA's sampling and analysis error (SAE) for the analysis of crystalline silica in order to apply it to the results we get from our private contract lab. Where can we obtain OSHA's SAE?

This request indicates some confusion with regard to industrial hygiene analyses and the meaning of SAE. Generally speaking, neither NIOSH nor OSHA mandate either specific analytical procedures or analytical equipment. Good science should prevail, and when permitted, other methods may be used. Each IH laboratory should determine its own SAE values based on the analysis of quality control samples, spiked samples or other tests of recovery. The private contract lab should be contacted for its SAE estimates for the time period of the analyses.

The SAE is not a universal constant but is a time-varying function of instrument and analyst performance for particular analytes. The OSHA PEL for dust containing quartz is given by PEL = 10 / (%Quartz + 2) , where the %Quartz in the respirable dust is calculated from the analytically determined mass of quartz in the gravimetrically determined mass of the dust. The standardized concentration is a measure of exposure severity to which the SAE is applied to confidently determine whether an overexposure has actually occurred. The standardized concentration is: Standardized Concentration = Exposure / PEL, where the Exposure term refers to the gravimetrically measured time-weighted average concentration (mg/m3) of respirable dust containing quartz. So the application of the SAE for respirable dust containing quartz involves two analyses: Gravimetric and quartz analyses.

The OSHA SAE for quartz analysis is provided to the OSHA compliance officer in the same report as the analytical results. Assuming a 5% error in the flow rate, the SAE for quartz analysis is typically around 20% for 95% confidence.

So strictly-speaking, the SAE for such analyses contains weighted components for the errors associated with the flow rate, gravimetric, and quartz portions of the analyses. The function for the overall SAE (at 95% confidence) that can be derived for the combined errors of the two analyses varies between about 10% for respirable dust with little quartz to 20% for respirable dust that is nearly pure quartz. Using the SAE for quartz given with the analytical report, therefore, provides a better than 95% confidence level. The PEL calculation for respirable dust containing both quartz and cristobalite is given by:

PEL = 10 / [2 + %Quartz + 2(%Cristobalite)],

so a similar situation arises.

Regarding Air Sampling

How does a PVC filter with a 5-micron pore size capture submicron size dust?

The low-ash 5-micron PVC filter is the preferred sampling medium for respirable dust containing crystalline silica and is used in the American Industrial Hygiene Association (AIHA) Performance Analytical Testing (PAT) Program. The 5-micron pore size filters reduce problems associated with sample loading and back pressure. This condition is important to maintaining a constant sampling rate in dusty work environments.

There are several reasons generally attributed as to why these filters work well in this application.

The pore size in this type of membrane filter is just one description of the filter medium. The 5-micron size for this filter is not the physical dimension of holes at the surface of the filter; it is based on the pressure drop across the filter. The electron microscope shows the surfaces to be a maze of PVC strands presenting passageways much smaller than 5 microns. The "smooth" surface appears as a spongy mat on which impaction can occur. Pore paths are not linear so particles do impact surfaces.

 This is an image of the smooth shiny side of clean 5-micron filter (Omega P503700). This is the rough, dull side of the same clean filter. Click the image to enlarge. Click the image to enlarge.

Such filters do not behave merely like sieves and have good collection efficiencies for air suspended dusts well below the pore size. The Omega Specialty Instruments 5-micron PVC membrane filter P503700 is used in the MSA 800019 cassette in the OSHA automated weighing program. In capture efficiency tests, the P503700, for example, retained 99.9% of respirable coal dust. Particles are captured primarily by impaction and diffusion (Morris Katz, Ed., Methods of Air Sampling and Analysis, 2nd Ed., American Public Health Association, Washington, DC 1977, p 199 and University of Minnesota Short Course on Aerosol and Particle Measurement, Aug. 22-24, 1994). Dust particles tend to agglomerate (clump or stick together) due to Van der Waals, polar and ionic surface features, and hydrogen-bonding forces. [Crystalline silica surfaces exposed to the air rapidly become covered with silanol (Si-OH) groups that can ion exchange and hydrogen bond.] Through agglomeration, smaller particles join together and with larger particles to be more readily captured on the filter than isolated particles. Filter loading also increases filter efficiency; as the filter begins to pick up dust, the effective pore size decreases. Dusts are rarely spherical. For example, quartz dust produced by abrasive and grinding processes generally appears under the microscope like shards of broken glass. So in some orientations they often present a larger cross section for capture than their equivalent spherical diameter may indicate. Most aerosols in the workplace carry a static electrical charge. When they contact the filter they induce an opposite charge to form at the surface enhancing the surface attraction. Also, uncharged particles can develop charge; when different materials (the aerosol and the membrane) come in contact they become oppositely charged and attracted to one another.

[Note: Currently available PVC filters of the smaller pore size (0.5 micron) do not dissolve properly in tetrahydrofuran (THF) and therefore cannot be analyzed by OSHA method ID-142.]

Can I sample crystalline silica using low-ash 0.4-micron pore size AA (MCEF) filters instead of low-ash PVC filters?

Generally speaking, no.

The weights of MCEF filters are affected greatly by slight variations in humidity and are not suitable for the gravimetric portion of the analysis of respirable dusts containing crystalline silica.

Additionally, MCEF filters do not dissolve properly in THF and therefore the crystalline silica collected on MCEF cannot be readily analyzed by OSHA method ID-142.

Nylon has a problem with developing a surface static charge during use. I've been told that I can use a metal cyclone operated at a specific flow rate to give the desired 50% cutpoint at 3.5 microns. Why do we still use nylon cyclones rather than the newer metal cyclones or conductive plastic cyclones?

Manufacturers often refer to the 50% cutpoint of 3.5 microns as being the critical distinction, but the other cutpoints specified in 29 CFR 1910.1000 Table Z-3 are equally important. These sampling efficiencies are at various aerodynamic particle sizes that compare well with both the then-current model for respirable dust deposition in the human respiratory tract and with the sampling efficiency of the AEC instrument consisting of a 1-cm Dorr-Oliver cyclone made with a nylon body. (The nylon body is not to be confused with the sampler holder which may have a metal protective tube surrounding the cyclone body.) Currently the only cyclone we are aware of that has the characteristic sampling efficiencies listed in 29 CFR 1910.1000 Table Z-3 is the Dorr-Oliver nylon cyclone. A device that does not develop a static charge would probably give more reproducible sampling and represent an improvement, but until an equivalent conductive device becomes available, we use the Dorr-Oliver nylon cyclone. Please inform the Salt Lake Technical Center if you are aware of another cyclone that matches the characteristics in 29 CFR 1910.1000 Table Z-3.

How do I perform a leak test of the Dorr-Oliver cyclone?

The Cincinnati Technical Center has put together a set of instructions on leak testing procedures. Click here for the document.

I need to sample a sand-blasting operation in a room that has no forced ventilation. The blasting lasts only 20 minutes each day. The employee wears a hood only during the operation. The employee spends 8 hours a day in the same room with other employees performing other activities. How long should I sample?

Sample the whole 8-hour work day.

If respirable silica dust is generated during the blasting, it will remain suspended long afterward. The air is rarely still enough for the dust to completely settle. The activities of the employee and his/her coworkers would also resuspend dust that has settled on their clothes. Dust can also be resuspended when workers pass by or over dusty surfaces, such as the factory floor and when they perform any inappropriate cleanup operations such as dry sweeping or using an ordinary shopvac. While IHs are often tempted to sample only during the dust generation operation, doing so will underestimate the actual exposure – the result of such sampling may easily show up as non-detected whereas the exposure may be very much larger. For example, if the 20 minute sampling gave a result just below the detection limit (say, 10 micrograms of quartz) and the air concentration averaged just half that level throughout the rest of the day, the total quartz exposure would be about 0.150 mg/m3 – well over the TLV. If this dust was a typical % quartz composition of say, 20%, the total weight collected over the day would have been 0.77 mg/m3 giving a severity of exposure of 0.77/[10/(2+20)]=1.7 or 1.7 times the OSHA PEL. The IH should consider such consequences of short-duration sampling for any air contaminant.

The material safety and data sheet (MSDS) for a product consisting of food-grade calcium silicate indicates that (for example) it contains:
19% CaO
67% SiO2
6 to 8% H2O and
less than 0.1% free silica.
How can that be; isn't SiO2 "free" silica? Should I sample for quartz?

No and no.

The description indicates that the SiO2 is not "free silica." "Free silica" includes amorphous silica and the crystalline forms that are not chemically combined with any other elements. That is to say, silica that is not "free" is chemically bound in another compound. In this material, the element silicon is present in the form of silicate compounds and is not in the form of a compound consisting of SiO2 as implied by the analytical data. The confusion here is due to the use of the mineralogist's convention in representing the chemical composition of the product as component oxides. This use is a common and often practical way to report analytical results and, in this case, does not represent a mixture of the compounds CaO, SiO2, and H2O. Users of MSDSs should be aware of this common convention, but the distinction may not be emphasized even in undergraduate chemistry courses. The material was probably analyzed for Ca, Si, and weight loss on heating. When the results of the analysis of the product were reported, the gravimetric factors for the oxides were used with oxygen making up the remainder of the composition. This assumed oxide composition satisfies the chemical combining capacity of Ca and Si for compounds formed under normal conditions. The practical value of using oxides can be demonstrated by looking at the mineralogist's oxide representation of some of the various calcium silicate compounds that might be present in this product:
CaSiO3 or CaO.SiO2
Ca2SiO4 or 2CaO.SiO2
Ca3SiO5 or 3CaO.SiO2
The downside is that the convention does not always address health concerns.

Note: The term, "free silica" is often used loosely to refer to quartz; quartz is a "free silica," but not all "free silica" is quartz.

We are removing some old fire brick, fiber glass (or rock wool) heat insulation from a furnace. Can these materials contain crystalline silica?

The hazards in removal need to be evaluated – if no cutting, breaking, grinding, etc., is involved there may be no hazard.

In these materials devitrification (conversion of the glassy to the crystalline state) may have occurred. Devitrification can occur in fire brick, obsidian, pumice, fused silica, fiber glass, and rock wool held at elevated temperatures producing disordered crystalline silica forms. These materials should be analyzed for the type and amount of crystalline silica present before removal so that proper protection can be provided to the employee. Even without good XRD quantitative confirmation, crystalline silica may be qualitatively identified.

Just what is "free silica?"

The term, "free silica" is often used loosely to refer to quartz; quartz is a "free silica," but not all "free silica" is quartz. Amorphous silica and cristobalite are also examples of "free silica."

Regarding Analytical Procedures

We need OSHA's detection limit DL for the analysis of crystalline silica in order to apply it to the results we get from our private contract lab. Where can we obtain OSHA's DL?

As discussed above re. SAEs, contact your contract lab for the detection limit data they have determined for their analyses.

The qualitative detection limit is usually defined as the amount of analyte equivalent to three times the standard deviation of the instrument response for low level spikes on sampling media using the slope of the calibration curve of low level calibration standards. The low level spikes may be made either at a single low level to obtain the standard deviation, or the standard error of estimate of the regression for several low level spikes bracketing the detection limit may be used instead of the standard deviation at a single level. Ten or more spiked samples are recommended depending on whether spikes are at a single or several levels.

The quantitative detection limit is usually defined as 3.33 times the qualitative detection limit defined above.

Can AIHA Proficiency Analytical Testing® (PAT) data for a laboratory from recent rounds be used to estimate the analytical error portion of the sampling and analysis error (SAE)?

No.

The PAT program is designed to help consumers select laboratories that are proficient. In the PAT program analyses of quartz, the "true" values against which a laboratory's results are compared are based on results from reference laboratories that are a subset of the participating laboratories. Assuming that the PAT samples were made from accurately delivered consensus reference material and that the participants all used the same techniques, instrumentation and methodology, and that the samples are not otherwise flawed so as to introduce bias, the best accuracy that can be achieved by consensus analyses is limited by the standard error of the precision of that analysis [SD/(n)½, where SD is the standard deviation in the results among the n reference labs].

However, the PAT program is not suited to achieve the best accuracy: In the case of crystalline silica analyses, the analytical equipment and methods vary between labs. X-ray diffraction, infrared spectroscopy (IR), and colorimetric (chemical) analyses are all in use. The reference material used in these analyses may also differ in homogeneity and quality from the National Institute of Standards and Technology (NIST, formerly the NBS) standard reference material SRM 1878. NIST SRM 1878 is strongly recommended as a consensus standard for calibrations and tests of analytical accuracy.

The current method of PAT quartz sample generation is by aerosol generation using "5 micron" Min-U-Sil 5 without cyclones. In addition to any errors in the generation process, this "total dust" approach introduces a sampling error that may not duplicate the sampling error associated with the use of a cyclone.

In the PAT program, these generation and sampling errors are recognized as significant and are evaluated in statistical tests conducted on sub-batches and batches of PAT samples by the contract laboratory that prepares them. Part of each batch is sacrificed for characterization. The criterion of acceptability at each concentration level (sub batch) of PAT quartz samples permits the characterization consisting of 7 samples to have a maximum CV of 30% at each concentration level of quartz. The corresponding 95% confidence range would be about 42%. Additionally, a maximum average CV of 20% (for 28 samples) each set of 4 concentration levels (batch) is permitted. These criteria are currently under review by the AIHA and NIOSH. The results obtained by participants in the PAT program therefore include both the analytical error the participating laboratories introduce and an unknown but potentially large amount of error introduced in the generation and sampling of the aerosol. These latter errors may vary batch to batch.

Therefore, more appropriate data for monitoring analytical recovery and variability in the analysis of respirable quartz might be obtained by labs using SRM 1878 (or comparable reference material) to prepare blind QCs for their in-house quality control program. Quality control samples are typically prepared from aliquots of well-dispersed suspensions of the SRM in a solvent or are prepared gravimetrically. NIST prepares its SRM 2679a Quartz on Filter Media by delivery of a water suspension of SRM 1878 using a thickening agent to ensure homogenous suspensions. The true value is determined spectroscopically with a demonstrated precision of 4 to 11% depending on spike level. The deposit is overlain with a layer of clay. The filter medium is well suited for some IR methods, but unfortunately does not dissolve in the solvent used for XRD analyses.

Gravimetric QCs may give more accurate delivery and may be prepared by weighing the SRM or equivalent material onto tared filters or by delivering SRM to a filter as an aerosol (as in the case of the direct-on-filter methods). (Particle size effects can have opposite effects on recoveries in XRD and IR. The effect of particle size on the colorimetric method also strongly depends on analyst technique.)

How can we resolve possible misidentification of opal C or other materials that mimic cristobalite in the analysis of cristobalite?

One resolution of this problem requires a combined approach using a modified Talvitie chemical treatment (see description and reference in OSHA ID-142 method). In order to optimize the conditions for removal of the interferent, recovery tests using cristobalite-spiked bulk material is needed to ensure that cristobalite is not being removed along with the opal-C.

This procedure is tedious and is only performed when absolutely necessary. At the OSHA SLTC, the efficient use of resources is brought about by the procedures for analyzing compliance air samples for cristobalite given in method ID-142. The type of industrial operation determines whether cristobalite is analyzed. Air samples suspected of containing cristobalite are first screened on the most intense X-ray diffraction peak (JCPDS 11-695 an undistorted cristobalite). Based on the screening, if the exposure is above the detection limit and possibly over a significant fraction of the OSHA permissible exposure limit (PEL), an attempt is made to confirm on the next 2 most intense peaks. A consensus Standard Reference Material (SRM) for respirable quartz analyses is currently manufactured by National Institute of Standards and Technology (NIST, formerly the NBS). The best confirmation is achieved both quantitatively and qualitatively. We calibrate the instrument to provide quantitative results on all four quartz diffraction peaks; typically three are used. The silver membrane used to support the dust is used as an external 2-theta calibration; the peaks should agree within 0.05 degrees 2-theta of the location of the NIST Standard Reference Material. If the presence of cristobalite is confirmed, a wide angle scan is performed on the bulk material(s) that may be supplied with the air samples to look for potential interfering phases. We discuss any unusual situations that arise with the compliance officer who performed the inspection.

The process I need to sample involves silica (quartz, fused silica, diatomaceous earth, etc.) that reaches a temperature of only 1100 degrees C – too low for the formation of cristobalite. Why should I analyze for cristobalite?

Contrary to the usual diagram for the conversion of silica to tridymite (the diagram shows cristobalite requiring higher temperatures), silica in the temperature regime of 867 and 1470 degrees C will most often convert to a disordered cristobalite phase favored by free-energy considerations rather than tridymite. The extent of disordering depends in part on the presence of trace contaminants and on the thermal history of the material (sequence and time of heating at the various temperatures encountered in the process.

XRD is not as sensitive to submicron crystalline silica. Is this an important consideration?

No. For at least two reasons:

The smallest quartz particles are generally less significant in relation to the total weight of quartz in a sample. The larger particles in an aerosol generally represent most of the mass of the aerosol because the mass varies as the cube of the diameter. It takes 125,000 particles, 0.1 micron in size, to equal the mass of a single particle 5 microns in size.

Due to the presence of an amorphous silica layer on crystalline silica particles, a smaller proportion of the mass of small particles is crystalline silica. The amorphous layer is estimated to be approximately, 0.03 micron thick (N.J. Elton, P.D. Salt, and J.M. Adams, "The Accurate Determination of Quartz in Kaolins by X-ray Powder Diffraction," Issues and Controversy: The Measurements of Crystalline Silica, International Symposium, August 20-21, 1992, Cambridge, MA. CMA Crystalline Silica Panel.) A 0.03-micron layer on a spherical 1-micron diameter quartz particle represents 17% of the volume of the particle leaving about 83% crystalline. The relative volume of the amorphous layer increases rapidly with decreasing diameter; a 0.5-micron diameter quartz particle would be about 68% crystalline, and a 0.1-micron diameter "quartz" particle would be about 6.4% crystalline. These are high estimates because dusts are rarely spherical. As a result, the amount of crystalline silica decreases rapidly as particles become smaller.

 Silica