Identification of Compounds in South African Stream Samples Using Ion Composition Elucidation

Andrew H. Grange,¹ Papo M. Thomas,² Mathebula Solomon,² and G. Wayne Sovocool ¹

Environmental Sciences Division, NERL,
U.S. EPA, P.O. Box 93478,
Las Vegas, NV 89193-3478

Dept. of Water Affairs and Forestry,
Institute for Water Quality,
Pretoria, South Africa

This poster presented at the 51st annual meeting of the American Society for Mass Spectrometry dealt with identifying compounds that are not on a target list.&nbsp Demonstrated was the power of ICE used with a double focusing mass spectrometer to help identify compounds for which multiple mass spectral library matches were found, to deconvolute mass spectra based on ion compositions, to identify compounds providing few mass spectral peaks with signals greater than the chemical noise, and to identify compounds not found in mass spectral libraries.&nbsp Over 100 compounds were identified or tentatively identified in stream sample extracts from South African streams.

INTRODUCTION

Many target compounds have been identified and quantified in surface waters from industrialized countries.&nbsp A major study by the USGS targeted 95 compounds in streams throughout the US.¹&nbsp Analytical methods for target compounds usually employ clean-up procedures to remove potential mass interferences and utilize selected ion recording (SIR) to provide low detection limits.&nbsp Such an approach, however, could overlook non-target compounds that might be present and that could pose risks to ecosystems or to humans.&nbsp In an ideal world, all compounds present would be identified, quantified, and evaluated for toxicity.

Ion Composition Elucidation (ICE)

The US EPA's Environmental Chemistry Branch is identifying as many compounds as possible in combined acid and base/neutral extracts (1 mL total) of six 4-L stream samples collected near Johannesburg, South Africa, using Ion Composition Elucidation (ICE), a high resolution mass spectrometric technique developed in-house for a Finnigan MAT 900S double focusing mass spectrometer.&nbsp This Selected Ion Recording (SIR) based technique measures the exact masses of an ion and its +1 and +2 isotopic mass peak profiles that arise from heavier isotopes such as ¹³C, ²H, ¹⁵N, ¹⁷O, ¹⁸O, ³³S, and ³⁴S.&nbsp The abundances of the +1 and +2 profiles relative to the monoisotopic ion's profile are also measured for compounds by the technique illustrated in Figure 1.²

Note: to better view details of Figures 1 through 5, click on the image. A larger version of the figure will open, and may require scrolling to view in its entirety. To view the image in a separate window, click on the image. When finished, click the Back button in your browser.

Figure 1.&nbsp An output of partial mass peak profiles plotted from areas under chromatographic peaks observed in ion traces for each m/z ratio corresponding to a dot.&nbsp Each of the m/z 304, 305 (+1 in mass), and 306 (+2 in mass) partial profiles was plotted from seven ion traces, and the lock and calibration mass profiles were each plotted using five ion traces.&nbsp Three exact masses and two relative abundances were calculated from these profiles.&nbsp The corresponding theoretical values for partial profiles from C₁₂H₂₁N₂O₃PS⁺, the molecular ion of diazinon, are:&nbsp 304.10105, 305.10396, 306.09944, 15.18%, and 6.11%.

Automated comparison of measured and calculated values of these three exact masses and two relative abundances generally provides a unique composition for the apparent molecular ion.³&nbsp Determining the exact masses of fragment ions then provides their compositions.&nbsp This methodology realizes the full potential of high resolving power for separating components by exact mass on the time scale of a chromatographic peak.&nbsp Mass spectral interpretation based on ion compositions often provides tentative identifications when there are:

	multiple library matches,
	convoluted mass spectra,
	poor quality mass spectra, or
	no library matches.


Multiple Library Matches

Two background subtracted mass spectra are shown in the top part of Figure 2.&nbsp M/z 133, 104, and 78 were entered in the NIST library program and 10 plausible matches with three different molecular ion compositions were found.&nbsp Thus far, ICE was used to determine the molecular ion composition of the first analyte, C₈H₇NO, which limited its possible identity to six compounds (assuming the analyte mass spectrum was in the NIST library).&nbsp The two isocyanatobenzenes are unstable in water, but could form from decomposition of precursor compounds in the injection port.&nbspThe azido compound is probably
	Figure 2 (This is only a portion of the figure. To view entire figure, click the image above)
unstable or explosive.&nbsp These three compounds are not likely to persist in a stream.&nbsp For the remaining 4 isomers, the NIST library provided 11 synonyms for 1H-indol-4-ol, while the others had three or less.&nbsp This compound was purchased and provided the same retention time as the analyte.&nbsp The second analyte is most likely 5-methylbenzotriazole, since the eight library synonyms included two trade names.&nbsp This compound has been quantified by the USGS in US waters.

Convoluted Mass Spectra
When hundreds of compounds are present in an extract, coelution of multiple components is common.&nbsp The background-subtracted (middle) mass spectrum in Figure 3 did not correspond to a single compound and no NIST library matches resembled the mass spectrum.&nbsp A loss of 4 from the m/z 139 ion to produce the m/z 135 ion was unlikely, but both ions could have been fragment ions of an unobserved molecular ion.&nbsp ICE was used to determine the compositions of the m/z 135 and 139 ions.&nbsp Each had an integer number of rings and double bonds suggesting each was a molecular ion and that the mass spectrum resulted from at least two compounds.&nbsp Entering C₇H₅NS and C₇H₉NO₂ into the NIST library provided five good matches and one good match with multiple ions in the convoluted mass spectrum, respectively.&nbsp Benzothiazole had three commercial names, while the other four compounds had none.&nbsp Benzothiazole was purchased and it provided the same retention time as the analyte.&nbsp The compound providing the single library match for the m/z 139 ion is not readily available.&nbsp Determination of the molecular ion's composition provides support for its tentative identification in addition to the library match.&nbsp The library matches are the top and bottom mass spectra in Figure 3.&nbsp Most of the ions in the convoluted mass spectrum are accounted for, but m/z 121 was produced from a third unidentified compound.
		Figure 3

Poor-Quality Mass Spectra
The negative mass defects for ions in the poor-quality, background-subtracted mass spectrum displayed in Figure 4 and the presence of m/z 249, 251, and 253 ions suggested an ultra-trace level of a halogenated compound was contributing to the mass spectrum.&nbsp ICE determined the m/z 249 ion's composition was C₆H₁₂Cl₂PO₄⁺.&nbsp M/z 249, 251, and 223 were entered into the NIST library program and two compounds were found that contained a P atom and at least 4 O atoms.&nbsp One provided ions at m/z 279 and 281, which were absent from the analyte mass spectrum.&nbsp The remaining compound was a flame retardant, tris(2-chloroethyl)phosphate (C₆H₁₂Cl₃PO₄) for which the [M-Cl]⁺ isotopic cluster yielded the highest-mass ions observed.&nbsp It was purchased and found to have the same retention time as the analyte.
		Figure 4

No Library Mass Spectra
The background subtracted mass spectrum in Figure 5 was not found in the NIST library.&nbsp The molecular ion was determined to be C₈H₇NO₂S₂⁺ using ICE.&nbsp Ordinarily, the compositions of the fragment ions would also be determined and candidate molecules consistent with all ions would be hypothesized.&nbsp However, in this fortuitous example, a chemical company's catalog contained only one compound with this composition.&nbsp In addition, the fragment ions observed were plausible based on the compound's structure. It was purchased and its retention time was the same as that of the analyte.
		Figure 5

Compounds Identified and Tentatively Identified

The analytes listed in Table 1 and their standards had the same retention times.&nbsp The analyte mass spectra when compared to those of the standards ranged from nearly identical to containing only a few ions in common.&nbsp The quality of the background-subtracted mass spectra and library matches were visually judged to be poor, fair, good, or excellent as indicated by one to four stars.&nbsp Often for poor quality mass spectra, the composition of the molecular ion or other high-mass ion (key ion in the table) was determined and entered into the NIST library before candidate compounds became apparent.&nbsp These ion compositions provided support for many tentative identifications before the standards were purchased.&nbsp For low-level analytes and for instances where significant mass interferences were present, the ion compositions provided important additional confirmation of each compound's identity.&nbsp Additional compounds will be added and blanks in the table will be filled as this work progresses.

Table 1 - For further information contact grange.andrew@epa.gov

In Table 2 are listed analytes for which library matches were found and for which the composition of the molecular ion or a high-mass ion (key ion) was consistent with the library match.&nbsp For analytes with standards available for purchase, most of these tentative identifications will be proven correct.&nbsp Where standards are not available, the ion composition strengthens the tentative identification. Compounds of anthropogenic origin could be traced upstream to their sources to confirm their identities based on their industrial, commercial, or household usage.&nbsp Some analytes could be byproducts or degradation products.

Table 2 - For further information contact grange.andrew@epa.gov

At least 40 additional analytes not listed in either table provided plausible library matches and
will be investigated further.

Importance of Software Versatility

The versatility of the Finnigan ICL^TM and ICIS^TM software made ICE studies of the hundreds of analytes in these complex extracts possible.&nbsp Using resolving powers of 3000 and 10,000, up to 32 ions were studied for each injection; each ion was from a different analyte.&nbsp Template files in Lotus^TM 9.1 based on the center masses of ions, start and end times for each SIR group, and the mass resolution, all entered by the user, provided analyte specific ASCII files.&nbsp These files were run as procedures by the data system to prepare SIR menus and to tune the instrument (ICL), and after data acquisition, to display ion chromatograms, to integrate peak areas, and to write an ASCII report file of m/z ratios and areas (ICIS).&nbsp This ability to execute ICL and ICIS procedures provided to the data system as ASCII files of instructions is essential for applying ICE to such complex extracts.&nbsp Grange and Sovocool suggest that all research-grade instruments should provide software with this degree of versatility to enable operators to explore and automate new ways of using the instruments.&nbsp ICE is a product of such exploration.

Conclusion

Streams in South Africa contain many of the same compounds found in streams in North America and Europe.
Ion Composition Elucidation (ICE) is a valuable tool for identifying non-target compounds in complex environmental samples.

References
1	"Pharmaceuticals, Hormones, and Other Organic Wastewater Contaminants in U.S. Streams, 1999-2000: A National Reconnaissance" DW Kolpin, ET Furlong, MT Meyer, EM Thurman, SD Zaugg, LB Barber, and HT Buxton Environ. Sci. Technol., 2002, 36, 1202-1211.
2	"Determination of Elemental Compositions from Mass Peak Profiles of the Molecular (M), M+1 and M+2 Ions" Grange AH, Donnelly JR, Sovocool GW, Brumley WC Anal. Chem., 1996, 68, 553-560.
3	"A Mass Peak Profile Generation Model to Facilitate Generation of Elemental Compositions of Ions Based on Exact Masses and Isotopic Abundances" Grange, A.H., Brumley, W.C., J. Amer. Soc. Mass Spectrom, 1997, 8, 170-182.
	Final drafts of references 2, 3, and other detailed information about ICE and its applications is available at: http://www.epa.gov/nerlesd1/chemistry/ice/default.htm

Analytical Environmental Chemistry
ICE Home Page

Environmental Sciences | Office of Research & Development
National Exposure Research Laboratory
Author: Andrew Grange
Email: grange.andrew@epa.gov

Identification of Compounds in South African Stream Samples Using Ion Composition Elucidation

Local Navigation