Filter collection detection limit

Water (1000 mL) is transferred into a 2-L separatory funnel and extracted with two portions of 50 mL of dichloromethane for 30 min with a mechanical shaker, and the extracts are collected in a 200-mL Erlenmeyer flask. The combined extracts are filtered through anhydrous sodium sulfate into a 300-mL round-bottom flask and evaporated to dryness with a rotary evaporator under vacuum. The residue is dissolved in 1 mL of n-hexane and an aliquot is analyzed by GC/NPD or GC/lTD under the conditions described in Section 2.2.3. Recoveries from water samples fortified with 0.0002 and 0.001 mgL of pendimethalin were in the range 94-110% by GC/NPD and 91-111% by GC/lTD. The detection limit was lower than 0.0001 mgL with both methods. 

Water samples were collected in midsummer, precipitation-free periods, following protocols of Ficklin Mosier (1999). On-site measurements include pH, specific conductance, alkalinity, acidity, dissolved oxygen, turbidity, and water temperature. Samples were collected in 1-litre polypropylene bottles and filtered onsite (0.45 pm) with disposable filters. Subsamples for cation analysis were placed in acid-rinsed polypropylene bottles and acidified with ultra-pure HNO3. Filtered, unacidified sub-samples for anion analysis were refrigerated until analyzed. In 2007, samples for Hg analysis were collected no Hg was detected at detection limits of 0.02 pg/L, precluding collection in 2008. 

With analytical methods such as x-ray fluorescence (XRF), proton-induced x-ray emission (PIXE), and instrumental neutron activation analysis (INAA), many metals can be simultaneously analyzed without destroying the sample matrix. Of these, XRF and PEXE have good sensitivity and are frequently used to analyze nickel in environmental samples containing low levels of nickel such as rain, snow, and air (Hansson et al. 1988 Landsberger et al. 1983 Schroeder et al. 1987 Wiersema et al. 1984). The Texas Air Control Board, which uses XRF in its network of air monitors, reported a mean minimum detectable value of 6 ng nickel/m (Wiersema et al. 1984). A detection limit of 30 ng/L was obtained using PIXE with a nonselective preconcentration step (Hansson et al. 1988). In these techniques, the sample (e.g., air particulates collected on a filter) is irradiated with a source of x-ray photons or protons. The excited atoms emit their own characteristic energy spectrum, which is detected with an x-ray detector and multichannel analyzer. INAA and neutron activation analysis (NAA) with prior nickel separation and concentration have poor sensitivity and are rarely used (Schroeder et al. 1987 Stoeppler 1984). 

The detection limit for measuring a species by filter collection and subsequent analysis is determined either by the uncertainty in the filter blank, that is, the variability in the amount of a species in an unexposed filter, or by the limit of detection imposed by the analytical method. Most frequently, the limit of detection is determined by blank variability, thus care in preparation and handling of filters is important when using such systems for aircraft sampling, particularly if sampling is conducted in remote areas where concentrations are low. In this regard, it is crucial to establish the blank... 

The time-space resolution that may be achieved with filter sampling techniques is dependent on the collection rate, limit of detection, and ambient concentrations. In aircraft applications, filters are typically operated at high flows (100-500 L/min) to maximize the mass accumulation rate. At these flow rates, sampling times on the order of 20 to 30 min are generally sufficient for measurement of substances in the urban troposphere. For sampling in the upper troposphere or in areas remote from pollutant sources, collection times of several hours may be necessary to obtain measurable quantities of material. 

Chemical Composition Aerosol composition measurements have most frequently been made with little or no size resolution, most often by analysis of filter samples of the aggregate aerosol. Sample fractionation into coarse and fine fractions is achieved with a variety of dichotomous samplers. These instruments spread the collected sample over a relatively large area on a filter that can be analyzed directly or after extraction Time resolution is determined by the sample flow rate and the detection limits of the analytical techniques, but sampling times less than 1 h are rarely used even when the analytical techniques would permit them. These longer times are the result of experiment design rather than feasibility. Measurements of the distribution of chemical composition with respect to particle size have, until recently, been limited to particles larger than a few tenths of a micrometer in diameter and relatively low time resolution. One of the primary tools for composition-size distribution measurements is the cascade impactor. 

The analysis of particle bound PAH involves collection of PAH bound to dust particles on 0.8 pm glass fiber or silver membrane filters, desorption of PAH from the particles into a suitable organic solvent, and analysis of the extract by a capillary GC using an FID. Between 500 and 1000 L air at a flow rate of 120 L/h is recommended for sampling, which can give a detection limit of 0.15 to 0.50 pg/m3 for each compound (Riepe and Liphard, 1987). The method suggests the installation of an absorber resin, such as XAD-2 or Tenax, after the... 

Table I lists several XRD analytical methods recently developed in the NIOSH laboratories. For each analyte, the analytical range, detection limit and analytical precision are listed. The method numbers refer to the NIOSH Manual of Analytical Methods (2.). As indicated in the table, there are several NIOSH methods available for free silica analysis. Method No, P CAM 109 incorporates the internal standard approach as developed by Bumsted (3.), The other two methods S-315 and P CAM 259 are based on the substrate standard method. The major difference between the two is the direct sampling on silver membrane filters (S-315). This paper will address the various methods of quantitation, sample collection and procedures for matrix absorption corrections that have been used in this laboratory for the analysis of crystalline particulate contaminants in the workplace.

XRF is the simplest of these methods. It allows bulk analysis of solid or liquid samples with detection limits of approximately 0.1 pg. The method can thus only compete with radiometric methods for the longest lived of radionuclides. It has approximately the same sensitivity for Th as alpha spectrometry but has the advantage that little sample preparation is required and that analysis is rapid and easily automated. XRF would be the method of choice for measurement of airborne thorium collected onto filter papers, for example. 

Atmospheric particulates, collected on Whatman 41 cellulose filters, are decomposed with sulfuric acid and hydrogen peroxide for subsequent determination of antimony and bismuth and with sulfuric acid and nitric acid for tin. Each element is analyzed independently by hydride generation/atomic absorption spectrometry. The optimization of instrumental as well as chemical parameters is described. The precision of the entire procedure is generally better than 10%. The detection limits are 0.25 ng m" for antimony and tin and 0.13 ng m for bismuth if 400 m of air are filtered and a 2 ml aliquot of the initial 50 ml sample solution is analyzed. 

Background Caused by Filters. Since all of the particles were collected on membrane filters it was necessary to determine the blank metal concentrations in the filter. This enabled an estimation of how many particles must be collected in order that the levels of the metals were significantly greater than the blank filter. For this study, both neutron and flame atomic absorption spectrometric analyses were used and the results are shown in Table I. The analyses by neutron activation were made on the filter directly whereas those by atomic absorption spectrometry were obtained by extracting the filter with nitric acid (16M Ultrex). There are apparent differences between the two sizes of membrane filters which are probably related to the fact that these filter sets were obtained at different times. Also, while the metal blanks within a particular batch of filters vary by negligible amoimts, the variations between batches are considerable. These determinations are near the detection limits for both techniques, and therefore there are considerable uncertainties associated with the results. However, these blanks did indicate the minimum level of metals which must be collected if the analyses are to be significant. 

Most AEC methods use direct sample injections vial a sample loop, thus achieving adequate detection limits." " Aqueous sample matrices usually do not require any sample pretreatment other than possibly filtration prior to injection. For matrices such as airborne particles, soil, or filters from air collection devices, organic acids are commonly transferred into an aqueous solution which can then be injected directly into the ion chromatograph. ° ° ... 

In a study of residential exposure to Cr-laden wastes in Hudson County, New Jersey, Lioy etal., 1992) collected 3-day and 4-day composite samples using low-volume air samplers equipped with 37-mm Teflon filters. Total Cr in airborne particles was determined using XRF with a detection limit of 2 pg g (shown in Table 11.3), and extractable Cr (not shown) was determined using nitric acid/sulfiiric acid digestion and ICP-AES. The mean ratio of extractable Cr to total Cr was 0.3. 

The analysis of air particulate matter by TXRF permits the determination of K, Ca, Cr, Mn, Fe, Cu, Zn, Sr and Pb with detection limits of O.i ng/m. Either the particles may be removed from the filter into an acidic solution or the sample carrier may be used as a plate in impactors for air particulate collection (Schmelling et al., 1997a) (the precision varies between 5 and 20%). Markert (1996) reports the determination of 34 elements in NIST-SRM-1633a with very good analytical parameters. 

Detection limits were determined using the 3filter material (Table 2). Detection limits are below 1 ng cm for all elements except Fe. Detection limits estimated for a 24-h collection using a high volume sampler (62.5 m h ) range from 0.001 ng m for Cd to 1 ng m for Fe. Detection limits are largely dependent on the sampler used with important specifications including sampled volume and filter size. The method presented here provides an improvement of at least 10-fold over the method described by Tanaka et al. (1998). 

Coneentrations of seleeted elements in 8 PMio filters collected in Boston are presented in Table 4. All elements had eoneentrations above detection limits... 

---