Environ. Sci. Technol. 1990,24,1832-1836
Anion-Exchange Chromatography Particle Beam Mass Spectrometry for the Characterization of Aromatic Sulfonic Acids as the Major Organic Pollutants in Leachates from Stringfellow, California I n Suk Kim, Fassil I . Sasinos, Robert D. Stephens, and Mark A. Brown"
Hazardous Materials Laboratory, California Department of Health Services, Berkeley, California 94704 ~~
Anion-exchange liquid chromatography with detection via both particle beam mass spectrometry (electron impact and negative chemical ionization) and UV absorption spectrophotometry is used to characterize target and nontarget compounds in lyophilizates from three Stringfellow hazardous waste leachates-upstream, downstream, and a charcoal-treated mixture. Lyophilization retains essentially all the organic carbon, which is 513, 46.8, and 453 ppm, respectively. Charcoal treatment removes priority pollutants but not most the total organic carbon (TOC),which is primarily highly polar, nonextractable, and nonpurgeable material. Besides 4-chlorobenzenesulfonic acid (from 53 to 69% of the TOC), seven additional major aromatic sulfonic acids are tentatively identified. All are sulfonated and chlorinated aromatic byproducts probably from DDT manufacture. Relative ratios of the aromatic sulfonic acids are not significantly altered by the charcoal treatment. The potential impact upon human health of these highly polar halogenated organic compounds is unknown.
Introduction A general problem in characterizing organic pollutants in aqueous leachates from hazardous waste sites and groundwater monitoring wells is that most of these pollutants are too polar, nonvolatile, or thermally labile to be analyzed via conventional gas chromatography based methods. Analyses of groundwater monitoring well Samples from 16 hazardous waste sites show that a total pyrolysis method invariably found substantially higher total organic halocarbon levels relative to a gas chromatography-mass spectrometry (GC-MS) based analysis, a discrepancy attributed to a significant nonvolatile component of the total organic halocarbon content at these sites ( I ) . Similarly, liquid chromatography (LC) with thermospray mass spectrometry detection has been used to confirm that approximately half the unidentified total organic halocarbon in leachates from BKK, a hazardous waste site located in southern California, is the nonconventional pollutant 4-chlorobenzenesulfonic acid (PCBSA). Only 4% of the total organic halocarbons had been previously identified by standard GC methods (2). Thermospray liquid chromatography-mass spectrometry (LC-MS) has seen limited application for analyses of difficult environmental analytes, including organic acids and esters (carboxylic and sulfate esters) via ion-exclusion, reverse-phase, and anion-exchange chromatography in both positive and negative ionization modes (3-5). Recently, particle beam LC-MS has been shown to be effective both as a qualitative and as a quantitative method for a wide range of chemical classes. These include ionic organic compounds such as paraquat hydrochloride, chlorophenoxy acid herbicides and daminozide, and extremely lipophilic compounds such as fenbutatin oxide (with a molecular weight of >1000) and endosulfan (refs 6-9; for a general review, see ref 6). Organic components in aqueous leachate Samples from the Casmalia and Stringfellow, CA, hazardous waste sites, and in drinking water from Santa Clara, CA, 1832
Environ. Sci. Technol., Vol. 24, No. 12, 1990
have been successfully resolved via anion-exchange liquid chromatography particle beam mass spectrometry (6, 7). These samples had been historically difficult or impossible to analyze by conventional analytical methods. The Stringfellow US. EPA Superfund site in California poses specific analytical problems common to many waste sites that may be addressed best via LC-MS methods. Most organic compounds contained in aqueous leachates from this site are not characterized by GC-MS-based methods. Analysis of Stringfellow bedrock groundwater shows that less than 1% of the total dissolved organic materials are identifiable via purge and trap analysis and are compounds such as acetone, trichloroethylene, etc., whose physical properties are ideally suited for GC-MS separation and confirmation (10). Most of the organic materials contained in these leachate samples are highly polar, nonpurgeable, and nonextractable compounds that have not been previously characterized. The major waste stream originating from Stringfellow sampled at upstream and downstream locations is shown to have 45 and 4070, respectively, of the total organic carbon as PCBSA (measured by ion chromatography and UV detection). This compound, a waste product from the manufacture of DDT, was known to be present because of a history of disposal of "sulfuric a c i d waste (11). Conventional reversed-phase chromatography fails to resolve or give any retention using any combination of elution solvents of the organic materials in Stringfellow leachates (6). PCBSA and its two isomers 2- and 3-chlorobenzenesulfonic acids have been detected by anion-exchange chromatography particle beam mass spectrometry (7). The utilization of inductively coupled plasma mass spectrometry as a detector with anion-exchange chromatography of Stringfellow leachates shows that all of the major organic components contain both chlorine and sulfur and are consistent with being other chlorinated aromatic sulfonic acids (7). The present study employs anion-exchange chromatography, electron impact (EI) and negative chemical ionization (NCI) particle beam mass spectra, and UV absorption spectra to characterize the major organic components of the total organic carbon contained in Stringfellow hazardous waste leachates.
Materials and Methods Liquid Chromatography Particle Beam Mass Spectrometry. Instrumentation consists of a HewlettPackard 5988A mass spectrometer equipped with a Hewlett-Packard particle beam LC interface and 1090 HPLC (Hewlett-Packard Co., Palo Alto, CA). Ionization modes are electron impact and negative chemical ionization with isobutane as a reagent gas. Mass spectrometry conditions are essentially the same as specified by the manufacturer: source temperature 250 "C, electron energy 70 (EI) or 200 eV (NCI), scan time 2 s, source pressure approximately 2 X (EI) or 1 X Torr (isobutane, NCI), helium pressure for the particle beam interface 50 psi. LC methods are initially developed on a Hewlett-Packard 1050 LC equipped with a 1040 diode array detector and 79994A
0013-936X/90/0924-1832$02.50/0
0 1990 American Chemical Society
Table 1. Gradient Solvent System for Anion-Exchange Chromatography of Lyophiliiatrr of Stringfellow Aqueous Leachates
Table I1 Concentrations (ppm) of Total Organic Carbon (TOC) of 4-Chlorobenzenesulfonic Acid (PCBSA) in Upstream, Downstream, and Charcoal-Treated" Stringfellow Aqueous Leachates
rnlvpnt %
time, min
water
acetonitrile
0
83 52 5
2 33 80 Rn
35 60 7n
F,
ammonium acetate (0.1 M)
sample
TOC ppm + SDb (% of upstream)
15 15 15
upstream downstream charcoal treated
513 22.7 (100) 46.8 7.6 (9.1) 453 3.0 (88.3)
15
Chem Station for data acquisition. The diode array detector is used to produce UV spectra of individual peaks of materials resolved via anion-exchange chromatography. Anion-exchange chromatography columns are made hy SGE (Ringwood, Australia) (Model 250GL-SAX, 25 cm X 2 mm). Elution is with ammonium acetate buffer and acetonitrile gradients (Table I). Flow rates are 250 wL/m. Tentative identifications are based upon a combination of electron impact (fragmentation) and negative chemical ionization (molecular weight) data, UV absorption spectra (presence or absence of chromophores), and available information about materials that have been disposed of at the Stringfellow site. Qualitative elemental analysis via ICP-MS of each peak from previous work (7) indicating the presence of sulfur and chlorine was also considered. Confirmation of the structure of PCBSA is possible by comparison and spiking with an authentic sample of this compound. Other authentic standards are not available. Aqueous Sample Preparation. Sensitivity limitations indicate that the direct detection of components in aqueous samples via particle beam LC-MS is usually not possible without prior extraction and concentration. Limits of detection for PCBSA are approximately 250 ng in both electron impact and negative chemical ionization-full scan. Lyophilization (freeze drying) is used for recovery and concentration, although the volatile fraction is sacrificed. Thus, an aqueous sample (5-2000 mL) is freezedried (Freezemohile 12 SL, VirTis Co., Gardiner, NY) over 1-72 h, the residue is extracted with methanol (2-200 mL), the inorganic salts are precipitated by addition of equal amounts of acetone, and finally the filtered soluble phase is evaporated under reduced pressure. This precipitation step may he repeated for samples containing very high levels of inorganic salts. The final residue is redissolved in methanol (0.25-20 mL) for injection. Total Organic Carbon Analysis. TOC is determined via a Dohrman DC 180 total organic carbon analyzer (Rosemount Analytical Division, Santa Clara, CA). It is measured initially for the whole aqueous leachate sample and then for a lyophilized sample reconstituted to its original volume in distilled water. Results and Discussion Total Organic Carbon a n d PCBSA Concentration Measured in Stringfellow Aqueous Leachates. Total organic carbon (TOC) and PCBSA concentrations of the upstream, downstream, and charcoal-treated mixture leachates samples are shown in Table 11. The lyophilization process of these aqueous leachates does not result in a significant loss of TOC, suggesting that these analytical methods are being applied to the major portion of the organic pollutants present in these samples. Since the proportion of the upstream and downstream leachates that is mixed for charcoal treatment is unknown, the amount of TOC removed by the treatment cannot be precisely determined. However, the treated leachate has 88.3% of the TOC compared to the upstream leachate, suggesting
PCBSA ppm SDb (% of TOC)
* *
334 + 17.2 (69) 27.6 f 2.6 (60) 241 f 7.3 (53.2)
'Charcoal treatment is used at the Stringfellow site for removal of priority pollutants from the leachate stream. bSD, standard deviation based upon three samples.
-
: 80 E
Downstream I
I
1
60
a
"
40
e
20 e
.
O
1
2
3
4
5
6
7
8
9 1 0
PEAK NUMBER
Figure 1. Percent peak area of peaks 1-10 in me rbwnstream versus charcoaUreat4 mixture relative to the upstream Sample of Sbingfelbw aqueous leachate (= 100%)(anionexchange dnomalograDhv whh UV absorption detection at 230 nm).
UV - 265 nm
J U L P J " LJ 0 IO 20 30 40 50 min Figure 2. UV absorption anion-exchange chromatography of String fellow leachate samples with tentatively assigned structures of wm pounds in peaks 3-10.
that charcoal treatment does not remove the majority of the aromatic chlorinated sulfonic acids present. Figure 1 shows the relative percent peak area of the 10 resolved peaks and indicates that the charcoal treatment selectively removes the late-eluting peaks in anion-exchange chromatography. Anion-Exchange Chromatography w i t h UV Absorption Spectrophotometry Detection. The anionexchange chromatograms of the upstream Stringfellow lyophilizate with UV absorption spectrophotometry (230 and 265 nm) with tentatively assigned structures of individual eight peaks are shown in Figure 2. There are at least 14 different major peaks present in this chromatogram. UV spectra of peaks 2-14, showing two distinct X Environ. Sci. Technoi., Vol. 24, No. 12, 1990 1833
Table 111. UA Absorption of Model Benzenoic Compounds with a-Saturated and Unsaturated Substituents;,,,A and t h e Largest Extinction Coefficient in t h e 244-268-nm Region (12) model compd a-saturated toluene ethylbenzene 2-chlorotoluene a-unsaturated chlorostyrene benzophenone 4-chlorobenzophenone
a-substituent
log e (Am=)
H CH3 CH3
2.35 (268) 2.30 (260) 2.46 (265)
=C =O =O
4.04 (244) 4.27 (252) 4.26 (259)
maxima in the regions of 210-230 and 255-270 nm are consistent with the presence of aromatic benzenoid chromaphores (see Table I11 for comparison with model compounds in the longer wavelength region) (12). Relatively stronger long-wavelength absorption a t the 265-nm detection window for peaks 7-9 (Figure 2, bottom trace) suggests that they have a-unsaturated substitution. For example, model aromatic compounds without a unsaturation have nearly 2 orders of magnitude smaller extinction coefficients in the region of 244-268 nm (Table 111). Anion-Exchange Chromatography with Particle Beam Mass Spectrometry Detection. Figure 3 shows both electron impact and negative chemical ionization particle beam LC-MS corresponding to the UV detection chromatograms seen in Figure 3. Chromatography con-
0
10
20
30
40
50 min
Figure 3. Particle beam mass spectrometry anion-exchange chromatography with electron impact (top trace) and negative chemical (bottom trace) ionization of Stringfellow leachates, with mass spectra of peaks 9 and 10.
ditions are identical with those used with UV absorption detection, although particle beam detection was carried
Table IV. Anion-Exchange Particle Beam Mass Spectra of Leachates from Stringfellow Groundwater with Electron Impact and Negative Chemical Ionization Mass Spectrometry peak no. 1
2 3
major ions and their source, m / z ( % base ion, fragment) electron impact ionization negative chemical ionization 627 - 631 (19, M-), 592 - 601 (40, M - CI), 522-530 (100, M - 3C1), 508-516 (82), 497-505 (221, 485-490 (42, M - 4C1) not observed 191-194 (5, M-, M - H, 1 Cl), 157 (30, M - Cl), 156 (100, M - HCl)
4 5 6 7
191-192 ( < l ,M-, M - H, 1 Cl), 157 (30, M - Cl), 156 (100, M - HC1) 329-334 (11,M - 1, M-, 2 C1); 294-296 (100, M - HCl) 191-192 (
