6.1. BIOLOGICAL SAMPLES

The data on analytical methods for detecting 1,3-DNB and 1,3,5-TNB and their metabolites in biological media are very limited. The few methods that have been used are discussed in the following section and summarized in Table 6-1.

1,3-DNB and its metabolites have been determined in the blood and urine of rodents fed the compound (Bailey et al. 1988; McEuen and Miller 1991; Nystrom and Rickert 1987). The methods to detect 1,3-DNB include high-resolution gas chromatography (HRGC) with electron capture detection (ECD), high-performance liquid chromatography (HPLC) with radioactivity detection (RAD) (for radiolabeled compounds) or ultraviolet (UV) detection or liquid scintillation counting (LSC) (for radiolabeled compounds), gas chromatography (GC) with mass spectrometry (MS), and spectrophotometry. It should be noted that the HPLC/RAD method is not suitable for the determination of 1,3-DNB and its metabolites in humans since it requires exposure to radiolabeled compounds. A reported method for quantitating 1,3-DNB and its metabolites in blood and urine by HRGC/ECD has a limit of detection in the low ppb range, and both recovery (≈110%) and precision (±3% coefficient of variation [CV]) of the method were excellent (Bailey et al. 1988). The reported methods based on HPLC separation and detection/quantification of radioactivity (McEuen and Miller 1991; Nystrom and Rickert 1987) are not suitable for monitoring human exposure because they depend on exposure to a radiolabeled dose of 1,3-DNB. GC/MS is a sensitive and highly selective method of detecting 1,3-DNB in blood and urine but has only been used for qualitative confirmation of the compound (McEuen and Miller 1991). It is possible that a modification of the method could be used for quantification of 1,3-DNB and its metabolites in these matrices. A spectrophotometric method for determining nitrobenzene in urine has been developed (Dangwal and Jethani 1980). This method could also be used to determine 1,3-DNB and 1,3,5-TNB in urine because it is somewhat selective for many nitro and amino benzene-based compounds. However, it is not as useful as HRGC/ECD because it is not as selective, and the sensitivity is several orders of magnitude less (ppm). An enzyme-linked immunosorbent assay (ELISA) has been proposed for the determination of 1,3-DNB in biological samples (Miller et al. 1991). The method showed good specificity and comparable recovery of 1,3-DNB from blood samples when compared with HPLC/UV and HPLC/radiochemical detection. No information was located that specifically discussed the detection of 1,3,5-TNB and metabolites in blood and urine by any method.

TABLE 6-1

Analytical Methods for Determining 1,3-DNB and 1,3,5-TNB in Biological Samples.

Methods have been developed for the detection of both 1,3-DNB and 1,3,5-TNB in extracts from hand-swab samples. The methods employ HRGC/ECD, HRGC/thermal energy analyzer (TEA), and HPLC/electrochemical detection (EC). Data are inadequate for a comparison of the sensitivity and reliability of these methods. Both HPLC and HRGC are effective in separating the analyte from other nitro compounds and contaminants (Douse 1985; Lloyd 1983). Detection by TEA (Douse 1985) and EC using a pendant drop mercury electrode (PDME) (Lloyd 1983) is more selective than ECD (Douse 1985). All three detectors were sensitive to ppb levels of analyte. The PDME has a unique advantage in that a new mercury drop can be formed between samples, and therefore, this detector is not subject to degeneration from contamination build-up. This makes the PDME highly reproducible (precision of 1.8% CV).

6.2. ENVIRONMENTAL SAMPLES

A large variety of methods has been described for the detection of 1,3-DNB and 1,3,5-TNB in environmental samples. These include GC or HRGC (combined with ECD, TEA, MS, nitrogen-phosphorus detection [NPD], or flame ionization detection [FID]), and HPLC (combined with UV and/or photoconductivity [PC] detection). Other methods that do not need chromatographic separation before quantitation, including MS, cyclic and differential pulse voltammetry, spectrophotometry, and assays based on chemical oxygen demand (COD) and total organic carbon (TOC), have also been used or tested. Table 6-2 contains a summary of several representative methods for determining 1,3-DNB and 1,3,5-TNB in various environmental media. Methods in which nitrobenzene was the analyte have been included when the methods apply to 1,3-DNB and 1,3,5-TNB as well. For several of the methods included in Table 6-2, nitrobenzene is the analyte being investigated, although the method should also be useful for analysis of 1,3-DNB and/or 1,3,5-TNB (when nitrobenzene data were used, this is indicated in the Sample column).

The few methods located for analysis of nitro compounds in air have not been well characterized for 1,3-DNB or 1,3,5-TNB. Only one method considered the analysis of 1,3-DNB specifically. The remainder were for analysis of nitrobenzene but could also be applied to the analysis of 1,3-DNB or 1,3,5-TNB. Most air samples are preconcentrated by collection on a solid sorbent prior to measurement; however, grab samples or liquid impingers have been used in some methods for the collection of 1,3-DNB. HRGC and GC with FID have been used to measure 1,3-DNB and nitrobenzene in air samples (Andersson et al. 1983; Cooper et al. 1986; Kebbekus and Bozzelli 1982). Under the experimental conditions used, reliability was adequate, with accuracy ranging from 52 to 84% and precisions ranging from 6 to 30% CV (Andersson et al. 1983; Kebbekus and Bozzelli 1982). The lower recoveries and precisions were obtained with nitrobenzene (Kebbekus and Bozzelli 1982) and could be substantially different for 1,3-DNB and 1,3,5-TNB. However, tests with 1,3-DNB and 1,3,5-TNB would have to be conducted to determine reliability parameters for these compounds. Sensitivity for HRGC/FID using nitrobenzene as the analyte was in the low ppt (Kebbekus and Bozzelli 1982). A comparison of HRGC with either NPD or FID showed that NPD was far more selective for nitro compounds than FID (Cooper et al. 1986). The detection limit for HRGC/NPD under the conditions used was in the. low ppb, and the authors could not quantify nitrobenzene using HRGCYFID. A spectrophotometric method has been developed for detection of nitro and amino benzene-based compounds in air (Dangwal 1981). Since the method cannot differentiate between the various nitro- and amino-benzene compounds, it is substantially less selective than other available methods. The sample preparation is more complex than with other tested methods and involves extraction with carbon tetrachloride, a potentially hazardous chemical. In addition, the sensitivity is several orders of magnitude less (ppm) than the’sensitivity of the HRGC and GC methods.

Both GC (high and low resolution) and HPLC may be used to separate nitrobenzene compounds in water. The most common detector for HPLC analysis is UV. For the GC methods, several detectors have been used, including NPD, ECD, FID, TEA, and MS. The sensitivity of the methods varies from sub to a mid-ppb range depending on the method, contamination level of the sample, efficiency of the extraction procedure, selectivity of the method, specific analyte (nitrobenzene, 1,3-DNB, or 1,3,5-TNB), and other method variables. The sensitivity of the GC-based methods is in the low-ppb range, with the limited data suggesting that ECD and FID may be slightly more sensitive than the other tested detectors (Austem et al. 1975; Richard and Junk 1986). However, TEA (Feltes et al. 1990), nitrogen-phosphorus (Army 1986), and MS (EPA 1986a; Feltes et al. 1990; Stemmler and Hites 1987) detectors are considered much more selective. GC/MS using a low-resolution instrument in electron ionization mode is the method recommended by EPA (EPA 1986a) because of its selectivity and sensitivity. However, other studies have shown that negative chemical ionization (NCI) and electron capture NCI are more sensitive and reliable than the EPA method (Feltes et al. 1990; Stemmler and Hites 1987). HPLC provides an alternative to GC-based methods (Army 1983, 1989; Feltes and Levsen 1989). The HPLC methods available are sensitive (detection limits ranging from low ppt to low ppb), relatively selective, reproducible, and reliable, in that they maintain the integrity of samples (sample decomposition may occur in heated zones of GC injectors). HPLC is also simple and rapid, requiring little sample preparation. A modification of the HPLC/UV method couples two UV detectors and a PC detector (Army 1989). This arrangement of detectors can improve the selectivity of HPLC substantially. An MS technique that allows direct injection of a water sample has also been tested (Yinon and Laschever 1982). The detection limit was only in the low-ppm range, but its selectivity makes it a good method for screening samples for further analysis. Assays based on COD and TOC (Roth and Murphy 1978) are well-established standard methods for determining organic pollution in water, but they are not selective for nitrobenzene compounds.

TABLE 6-2

Analytical Methods for Determining 1,3-DNB and 1,3,5-TNB in Environmental Samples.

HPLC/UV was the only method located for measuring nitrobenzene, 1,3-DNB, and 1,3,5-TNB in soil (Army 1985a; Bauer et al. 1990; Jenkins and Grant 1987; Jenkins et al. 1989). This method, developed by the Army, has been extensively tested and has been proven to be selective and reliable, giving high recoveries and good precision for complex samples. Sensitivity in the low-ppm range has been reported. A similar method, with less rigid sample clean-up, had recoveries for nitrobenzene that varied widely (Grob and Cao 1990). This shows the importance of sample extraction and clean-up with regard to results when the matrix is complex.

Cyclic voltammetry and differential pulse voltammetry have been used to analyze wine, beer, and cider for nitrobenzene (Lorenzo et al. 1988). While no detection limits were reported, amounts as low as 0.5 mg/L were easily detected and precision was excellent (±5% CV). An advantage of this method is that the analyte can be measured by direct insertion of the electrode in the solution. The method should also apply to the detection of 1,3-DNB and 1,3,5-TNB in solutions because it is based on polarographic determination of the nitro group. However, it is not as selective as HPLC- and GC-based methods. MS/MS has been investigated as a screening method for explosives (McLuckey et al. 1985), but no data on the sensitivity and reliability of this method were available. A supercritical fluid capillary chromatographic method with FID detection has been proposed for the determination of a broad range of compounds (including nitroaromatics) in solid wastes (Pospisil et al. 1991). The method was used to chromatograph over 270 compounds on a single column within 1 hour.

6.3. ADEQUACY OF THE DATABASE

Section 104(i)(5) of CERCLA, as amended, directs the Administrator of ATSDR (in consultation with the Administrator of EPA and agencies and programs of the Public Health Service) to assess whether adequate information on the health effects of 1,3-DNB and 1,3,5-TNB is available. Where adequate information is not available, ATSDR, in conjunction with NTP, is required to assure the initiation of a program of research designed to determine the health effects (and techniques for developing methods to determine such health effects) of 1,3-DNB and 1,3,5-TNB.

The following categories of possible data needs have been identified by a joint team of scientists from ATSDR, NTP, and EPA. They are defined as substance-specific informational needs that, if met, would reduce or eliminate the uncertainties of human health assessment. This definition should not be interpreted to mean that all data needs discussed in this section must be filled. In the future, the identified data needs will be evaluated and prioritized, and a substance-specific research agenda will be proposed.