Jump to main content.


Ion Composition Elucidation (ICE)  

Mass Measurements by an Accurate and Sensitive Selected Ion Recording Technique

Andrew H. Grange1, Joseph R. Donnelly1, William C. Brumley2, Stephen Billets2 and G. Wayne Sovocool2

1Lockheed Environmental Systems & Technologies Co., 980 Kelly Johnson Drive, Las Vegas, Nevada 89119.
2U.S. Environmental Protection Agency, National Exposure Research Laboratory,
Environmental Systems Monitoring Laboratory, P.O. Box 93478, Las Vegas, Nevada 89193.

ABSTRACT

Trace-level components of mixtures were successfully identified or confirmed by mass spectrometric accurate mass measurements, made at high resolution with selected ion recording, using GC and LC sample introduction.  Measurements were made at 20,000 or 10,000 resolution, respectively.  Under these conditions, accurate mass determinations from magnet or accelerating potential-electrostatic sector scans would not be feasible.  Chlorinated diphenyl sulfides, aralkylamine lubricant additives, and a sulfonated azo dye were analyzed with mass measurement errors of less than 5 ppm.

INTRODUCTION

Increasing the sensitivity and the resolving power of mass spectrometers has been an important goal for approximately 80 years.  The early purpose for high resolution was to improve the accuracy and precision of atomic mass measurements.(1-3)  Studies of organic compounds followed, largely by the petroleum industry.(4)  High mass resolution is now used routinely to separate analyte ions from interferences having similar masses and to determine accurate masses of the ions.

The accurate masses of molecular and fragment ions are valuable data for identifying unknowns.  For each accurate mass, a limited set of possible elemental compositions can be calculated whose masses fall within the accuracy limits of the experimental mass measurement.

Traditionally, accurate masses have been determined by using peak matching(1) in the accelerating potential-electrostatic sector (KVE) scanning mode, typically within an accuracy range of 1-5 ppm.(3)  The KVE scanning mode employs a fixed magnet current and varies accelerating potential and electrostatic sector voltages to scan a narrow mass range that includes two calibrant masses.  At 10,000-20,000 resolution, many high-resolution mass spectrometer (HRMS) data systems require long scan cycle times, for example, 2.7-5.4 s with the VG 11-250 system for scans 14 Da wide when perfluorokerosene (PFK) is the calibrant.  However, the requirements of a relatively large amount of sample and long scan time are not problems for single analytes introduced by the probe.  Accurate mass measurements also can be determined in magnet scanning mode, with some sacrifice in accuracy and sensitivity.  This technique affords accuracies in the range of 5 ppm at 5,000 resolution for a complex environmental fly ash extract with GC sample introduction.(5)

Accurate mass measurements by magnet or KVE scan are often impractical for samples introduced by chromatographic techniques because the amount of sample is small, the residence time in the HRMS source is short, the data consume much system memory, and the long reset time between scan ranges prevents measurement of components with retention times differing by less than 1 min during a single analysis.  Only two or three measurements can be taken across a typical capillary GC peak, for example, and resolutions above 10,000 generally cannot be achieved because of sampling time and sensitivity limitations.  Consequently, gas chromatographic peaks cannot be time-profiled as an aid to determining which ions in a mass spectrum arose from a single component, and accurate masses cannot be determined for low-abundance ions or trace components because sensitivity is inadequate.

The U.S. Environmental Protection Agency (EPA) Environmental Monitoring Systems Laboratory (EMSL) in Las Vegas, NV, conducts research in the analytical separation and identification of unknown compounds in hazardous wastes and other environmental samples.  Although analysis by GC/MS has proven to be convenient and to provide important results, it is limited in that unknowns can be difficult to identify by their mass spectra alone.  Therefore, it is important to acquire on-line, accurate mass measurements for trace-level GC peaks.  This research describes the sensitivity enhancement and reduced measurement time realized by applying selected ion recording (SIR) to accurate mass measurements.  This report also describes the use of peak profiling for quality assurance of GC/MS data.

Environmental samples frequently contain trace-level components that cannot be identified with confidence by computerized, or manual mass spectral interpretations.  Often, mass spectral interferences remain after background subtraction, or reference spectra for these trace level components are not available in commercial data bases.  For unknowns, accurate mass measurements greatly limit the number of possible elemental compositions.  Of these compositions, only a fraction will be structurally rational.  High precision and accuracy in such mass measurements are crucial to reduce the number of possible elemental compositions for an unknown to a workable number, making identification much more practical.  Accurate mass measurement by GC/HRMS-KVE scanning generally entails compromises in mass resolution (less than 10,000) and sensitivity that are impractical for trace analysis.  As a result, low-level components may be undetectable.  Components in environmental samples with the same retention time having accurate masses separated by only a few ppm provide the appearance of a single component with an accurate mass that is the weighted average of the two.  The KVE technique is also unable to provide chromatographic peak profiling (intensity of a given accurate mass over time) to ensure the masses studied belong to the component of interest.

For the SIR mode, as in the KVE mode, the magnet current is fixed, while the accelerating potential and electrostatic sector voltage are varied.  In contrast to the KVE mode, however, they are varied in a stepwise fashion to monitor discrete m/z ratios for the dwell times specified.  The resulting improvement in the signal-to-noise ratio increases sensitivity and provides greater resolving power for accurate mass measurements.

The SIR accurate mass technique is applicable to other mass spectrometric techniques including liquid secondary ionization mass spectrometry (LSIMS) , where higher mass resolution can better discriminate against the chemical noise arising from use of a liquid matrix.  This liquid sample introduction technique is utilized to analyze nonvolatile, polar, and thermally labile compounds.(6)

We report here four examples where compounds present at trace levels in mixtures were successfully identified or confirmed by SIR accurate mass measurements made at high resolution.  For these examples, accurate mass determinations from magnet or KVE scans would not have been feasible.  In the first three examples, specific components of environmental extracts were studied by GC/HRMS.  In the fourth example, a sulfonated azo dye was analyzed by LC/HRMS.

EXPERIMENTAL

Samples.  The process waste still-bottom sample was obtained from Kenneth Wang of the U.S. EPA National Enforcement Investigations Center, Denver, CO.  The environmental extracts, prepared according to a standard method,7 were obtained through the EPA Contract Laboratory Program (CLP).  Aralkylamine lubricant additives, structurally similar to those in the environmental extracts, were obtained from a commercial supplier to serve as reference standards.  The Acid Orange 8 dye sample was purchased from Aldrich Chemical Co. and was used without purification (ca. 65% dye content).

Reagents.  Glycerol (Mallinckrodt spectrophotometric grade), ammonium acetate (Mallinckrodt AR grade), hexane, methanol, and methylene chloride (Baxter's GC2 grade) were used for LC/MS experiments.  Organic solvents and purified water (Barnstead nanopure grade) were filtered through 0.2-mm pore diameter filter paper to remove particulates.

Apparatus and Materials.  A 30-m x 0.25-mm i.d. Restek RTx-5 GC column was interfaced to the ion source for GC/HRMS analyses.  Nanoscale LC (n-LC) columns were prepared from 75-mm i.d. fused silica capillary tubing (Polymicro Technologies, Phoenix, AZ) and C18 bonded silica beads (Shandon ODS-Hypersil, Shandon Southern Products, Cheshire, U.K.).  Silica particles (10-mm diameter; YMC-GEL, Yamamura Chemical Laboratories, Kyoto, Japan) were used to prepare the frit at the end of the column.  Column preparation using a magnetically stirred reservoir has been described.(8)   An Isco SFC-500 microflow syringe pump was used to prepare columns and to drive the matrix through a 150-mm i.d. fused silica capillary (Polymicro Technologies) installed in the dynamic FAB probe which contained the n-LC column.  A Spectra-Physics SP8700XR LC gradient elution pump provided the mobile phase to the n-LC column through a 1:99 (column:waste) splitter.  The matrix flow was 1.0 mL/min.  Mobile phase flows of 1-10 mm/min were used.

The GC or n-LC column was interfaced to a VG 70-250SE high-resolution mass spectrometer equipped with the VG 11-250 data system and a Gateway 486-66 computer as auxiliary data system.  Both a Hewlett-Packard 5890 GC and a direct liquid inlet were available for sample introduction.

Procedure.  A GC/HRMS Analysis.  Approximately 20 mg of solid sample was dissolved in 500 mL of methylene chloride for 1-mL on-column injections into the GC/MS without further purification.  The CLP environmental extracts were prepared according to the EPA Statement of Work.7  The GC temperature program was as follows:  50 oC for 3 min, 12 oC/min to 230 oC, 25 oC/min to 310 oC, final hold of 10 min.  Positive ion 70-eV electron ionization mass spectra were acquired in selected ion recording mode with perfluorokerosene (PFK) calibrant, an accelerating potential of 8 kV, and an ion source temperature of 250 oC, and the GC transfer line was maintained at 280 oC.

B.  LC/MS (n-LC/LSIMS) Analysis.  An accelerating potential of -8 kV was used with a mass resolution of 10,000 and a photomultiplier voltage of 500 V for n-LC/LSIMS analyses.  A VG dynamic FAB probe with a hemispherical tip was used as the n-LC/LSIMS interface with a VG Cs ion gun providing the primary ion beam (beam energy, 22 kV; emission current, 0.8 mA).

C.  Mass Peak Profiles from SIR data.  A mass peak profile is displayed from SIR data by monitoring m/z ratios across the mass profile at discrete intervals and then plotting the areas as a function of mass.  As illustrated in Figure 1, the area under the chromatographic peak in each ion chromatogram is integrated for each m/z ratio in the SIR descriptor (SIR mass) to provide a point on the mass peak profile.

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

Figure 1.  (a) Mass peak profile for an alkylated diphenylamine standard with a molecular weight of 393.3396, plotted from the areas under the ion chromatograms for m/z ratios across the profile, using a mass increment of 5 ppm.  A mass resolution of 32,500 was used for analysis of the authentic standard. (b)-(f) Ion chromatograms for five SIR masses used to plot the mass peak profile.

Use of chromatographic peak areas rather than maxima enhances the S/N ratio of the profile through signal averaging.  A reference ion must also be monitored to make corrections for small calibration errors.

The number of points monitored across each mass peak profile is a function of the mass resolution and of the mass increment, which is determined by the mass range observed and the number of m/z ratios permitted by the data system.  The resolution and range are tailored to the goal of each analysis.  When the identity of an ion is to be confirmed, its theoretical mass is used as the center mass, and 5-ppm mass increments are used with a mass resolution of 20,000.  When an accurate mass is desired for a mass peak observed in a full scan mass spectrum, an accurate mass survey using a resolution of 3,000 and a mass increment of 100 ppm is made around that mass.  The large mass range ensures all ions responsible for the nominal mass are located, and the resolution provides some separation of multiple ions.  Often, an ion due to the calibrant is observed when a PFK lock mass is used.  Only three points are observed across each profile, and a coarse estimate of the accurate mass is obtained.  An estimated mass is then used as the center mass in a SIR descriptor to determine the accurate mass of a target ion.

For an accurate mass determination, the SIR mass increment is chosen to provide 10 points across the profile.  Seventeen points are monitored to ensure the entire profile is observed, permitting the mass resolution to be estimated from the profile width at 5% of the profile's maximum height.  Five points are monitored across the top portion of the lock mass ion to provide calibration for each determination.  The mass increment for each ion is calculated as either (center mass) / (resolution x 10) or (center mass) /200,000, whichever is larger.  The lower limit, corresponding to 5 ppm, is the minimum increment for the digital-analog converter of the data system.

The VG 11-250 data system cannot plot the mass peak profiles from the chromatographic peak areas.  Therefore, the data were exported to a personal computer using Reflection 4, and commercial spreadsheet and word processing programs were used to plot and print the ion profiles in different formats.(9-11)  Procedures in LOTUS 1-2-3 were written to prepare the SIR descriptors after input of the center mass and resolution by the user.  Multiple SIR descriptors to study a series of chromatographic peaks can be prepared for each injection.  The transition time between SIR descriptors is only 5 s, rather than 1 min as required for KVE scans.  Hence, accurate masses can be determined for compounds with different masses that have retention times that differ by less than 1 min.  The LOTUS spreadsheet automatically calculates the SIR masses and prepares an ASCII file, which is executed as a command file by Reflection 4 to type the SIR masses into the SIR descriptor(s).  Preparation time for each experiment is minimal and typing errors are avoided.  To provide the partial mass peak profile for the lock mass, a chromatographic peak was simulated by operating the valve between the mass spectrometer's source and analyzer sections as a shutter.  One such simulated chromatographic peak is shown in Figure 3c.  The VG data system integrates the area between the two times when the shutter was closed.

RESULTS AND DISCUSSION

SIR Accurate Mass Measurement Technique.  Magnet scan, KVE scan, and SIR accurate mass measurements were made in triplicate at 10,000 resolution on 2,3,3',4-tetrachlorobiphenyl introduced by GC.  The detection limits (at 2.5:1 signal-to-noise ratio) and accuracies of these techniques could then be compared under similar conditions.  The first method, magnet scanning, gave a detection limit of 2 pg and a deviation from the theoretical mass of 31 ppm.  The KVE scan had a detection limit of 1 pg and a deviation of 0.2 ppm.  The SIR measurement had a detection limit of 6 fg and a deviation of 0 ppm.  A 10-fold reduction in ion beam intensity occurred when the resolution was increased from 10,000 to 20,000.  Thus, at 20,000 resolution, the SIR technique could provide approximately a factor of 17 better detection limit than the KVE scan technique at 10,000 resolution.

Like the KVE scan, SIR measurements produced mass peak profiles.  In contrast to the KVE scan, the SIR technique also allowed plotting of individual accurate mass chromatograms against time.  In the SIR mode, a cycle time of 0.8 s was selected, allowing 10 or more measurements across chromatographic peaks of at least 8 s duration.  The system memory requirements for the acquired data were at least a factor of 10 less than those for scan data.  The greater sensitivity of the SIR mode permitted use of higher mass resolution while monitoring ions of low abundance from trace components in complex mixtures.

The high resolution and sensitivity of the SIR technique, combined with the ability to plot mass and chromatographic peak profiles, provided valuable quality assurance for accurate mass measurements.  These features gave evidence that the correct mass was measured and that the mass belonged to a single component.  Separation from species present in the sample having nearly identical retention times and from contaminants such as column bleed and calibrant present in the mass spectrometer source were demonstrated.

Mirex and a Diazene.  An example is shown in Figures 2 and 3 where the high resolution and sensitivity of the SIR technique, combined with mass chromatographic peak data, allowed differentiation of species with similar retention times and accurate masses in an environmental sample from a waste site.

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

Figure 2.  Left-hand traces: mass peak profiles for selected ions, all having a nominal mass of 336 Da, as acquired and separated from coincident ions having similar accurate masses. Mass resolutions of (a) 3000, (b) 4500, (c) 10000, (d) 20000 with 10-ppm SIR mass steps, and (e) 20000 with 5-ppm SIR mass steps were used to select and acquire accurate masses of ions with increasing accuracy and selectivity as resolution was increased and SIR mass step increments were reduced in size. Right-hand traces: reference mass profiles.

.

Figure 3 - For further information contact grange.andrew@epa.gov

Figure 3.  (a-c) Ion chromatograms for the peak and shoulder maxima in Figure 2a, normalized to the largest signal.  Magnification for the second trace is 10x, and 100x for the third trace.  (d-e) Ion chromatograms for the peak maxima in Figure 2d, normalized to the larger signal.  In this example, 20,000 resolution provided mass separation of two species that were not well resolved at 10,000 resolution.

Initially, a retention time window that included time before and after the chromatographic peak of interest was chosen to determine if interferences were present.  At 3,000 resolution (Figure 2a), a large profile with two shoulders was seen; at least three species having nominal mass 336 Da were present near the retention time of interest.  The smaller shoulder that maximized at SIR mass 335.9631 Da was due to C9F12+ (theoretical mass, 335.9808 Da) from the calibrant, PFK.  The mass increment was 33.5 mDa, and the difference between the experimental and theoretical masses was 17.7 mDa, about one-half step.

The largest mass peak profile (m/z 335.7952) and larger shoulder (m/z 335.8959) were resolved into two discrete profiles in Figure 2b using 4,500 resolution.  The most abundant ion corresponding to SIR mass 335.8064 Da was the C1035Cl337Cl3+ ion of mirex (theoretical mass, 335.8043 Da).  The mass increment was 12.0 mDa, and the SIR mass at the profile's apex was 2.1 mDa larger than the theoretical mass.

The smaller profile in Figure 2b was partially resolved into two profiles in Figure 2c at 10,000 resolution.  At 20,000 resolution (with 3.4-mDa or 10-ppm SIR mass steps; Figure 2d), the two profiles were completely resolved.  The accurate mass measurement with the greatest accuracy and precision, shown in Figure 2e, was acquired at 20,000 resolution with a SIR mass step of 1.7 mDa (5 ppm).  The central mass in the SIR descriptor, 335.9205 Da, was selected as the expected accurate mass based on the best library match to the full scan mass spectrum.  This mass was confirmed in three of three accurate mass determinations, verifying the elemental composition to be that of bis(3,4-dichlorophenyl)-diazeneoxide, C12H6ON235Cl337Cl.

Experimental accurate masses in Figure 2 were influenced by SIR descriptor step size and by resolution.  The SIR mass corresponding to the apex of each profile is usually the one closest in value to the mass of the ion.  In Figure 2d, m/z 335.9190 is closer in value to 335.9205 Da than to the next larger SIR mass, 335.9223 Da.  The theoretical mass, 335.9205 Da, was the center mass in Figure 2e and, as expected, provided the maximum chromatographic peak area.  If greater accuracy is required when the theoretical mass is not included in the SIR descriptor, a peak shape equation or a weighted average of the most intense points on the profile could be employed to calculate the maximum (the experimental accurate mass value).

The complexity of the sample was also shown by the mass chromatograms (Figure 3).  The three top mass chromatograms show the three components of nominal mass 336 Da from Figure 2a that chromatograph well (Figure 3a), that chromatograph poorly (Figure 3b), and that are continuously present (Figure 2c; in this case, a PFK component).  Figure 2d shows the poorly chromatographed component at m/z 335.8955 that was mass resolved (at 20,000) from the desired analyte (Figure 2e) at m/z 335.9205.

Chlorinated Diphenyl Sulfides.  The second application of SIR accurate mass measurements acquired using GC/HRMS was the confirmation of tetra- through nonachlorodiphenyl sulfides as minor components in a process waste sample.  This class of compounds has been reported in pulp mill effluents and waste incinerator stack gases.(12)  The low-resolution mass spectra were consistent with the proposed structures.  The isotopic distributions for the molecular ions best matched those for ions containing 4-9 chlorine atoms, and the molecular ions had the expected negative mass defects.  Further, no [M-33]+ ions indicative of a Ph-SH structure were observed.

The possibility that the compounds were diphenylphosphines was investigated.  The mass difference between PH and S is only 0.0095 Da; hence, mass resolutions of 34,000-52,000 are required to completely separate the mass peak profiles of these two compound classes.  However, the customary 10% valley between mass peak profiles of equal height was not required to distinguish between two compounds, with only one expected to be present.  Since errors of plus or minus 1 mass increment are equally probable (and errors of plus or minus 2 or more increments are not observed under normal operating conditions), the average of several determinations serves to estimate the accurate mass to within 2.5 ppm (0.0012 Da for nonachlorodiphenyl sulfide) when a mass increment of 5 ppm is used.

Accurate masses were determined for the molecular ions of these trace-level chlorinated compounds.  Table 1 lists the experimental results.

Table 1.  Chlorinated Diphenyl Sulfides Found in Process Waste Sample

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

Using a mass increment of 5 ppm, the correct accurate mass for a polychlorodiphenyl sulfide [M]+ ion was obtained for 19 of 23 determinations.  Two of the determinations were high and two were low by 1 mass increment.

Aralkylamines.  Another example of elemental composition assignment by SIR accurate mass measurement using GC/HRMS was the identification of aralkylamine lubricant additives in a soil sample extract from a waste site where petroleum products had been recycled.  Review of the CLP GC/MS data package revealed the presence of a series of unidentified compounds having odd-mass molecular ions.  The methylene chloride extract, prepared by the contract laboratory using a standard environmental analytical method,7 was reanalyzed by GC/HRMS.  Accurate mass measurements of molecular and fragment ions using the SIR technique yielded elemental compositions consistent with interpretations based on masses and fragmentation patterns observed at low resolution.  Mass spectral features and accurate masses of components in the CLP extract matched those of available commercial additive standards.  Structures of commercial high-pressure and antioxidant lubricant additives were consistent with the data obtained in this study for these compounds (see Table 2).

Table 2.  Lubricant Additives Found in Environmental Sample Extract

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

a Other isomers are possible. Ph, phenyl; Ar, p-[PhC(CH3)2]C6H4-; Ar', p-tert-octyl-C6H4-; Ar", p-phenyl-C6H4-.
b Accurate mass and composition of fragment ion.
c Average used as correct mass.
d -6.7 ppm, -1.7 ppm, +3.3 ppm, +8.3 ppm.
e -7.0 ppm, -2.0 ppm, +3.0 ppm, +8.0 ppm.

For the standards, the correct accurate mass was obtained in 20 of 22 determinations using 5-ppm increments and 30,000 mass resolution (the practical resolution limit when pure standards are injected).

The arithmetically possible elemental compositions were determined for an ion with a mass of 405.2453 Da between atom limits of 10-33 C, 0-68 H, 0-6 O, 0-6 N, 0-4 S, 0-4 P, and 0-4 Si.  For error limits of 2.5, 5.0, and 10 ppm, 13, 31, and 63 possible compositions were found, respectively.  Hence, only about one-fifth as many compositions are possible when the mass error limit is reduced from 10 ppm to 2.5 ppm.  The 10-ppm error limit has been referenced for KVE scans and magnet scanning GC/MS determinations.(13)  The 2.5-ppm limit is attainable with the SIR technique.  The desirability of error limits tighter than 10 ppm for some situations has been discussed with reference to the recent recommendation for error limits of 5 ppm.(14)  For unknowns in environmental extracts, identification can be difficult because a number of exact mass compositions are possible within the accuracy of the accurate mass measurement; reducing the possible compositions is therefore important.  Complementary techniques such as GC/IR may not be available or sufficiently sensitive to detect the component of interest.

If a SIR mass is very close to the theoretical mass of an unknown, that SIR mass will usually be the ordinate at the apex of the experimental mass peak profile, or the point of maximum relative area in Figure 1. If the theoretical mass of the unknown is between two SIR masses, the experimentally determined apex will usually correspond to one or the other in the individual determinations, more often to the SIR mass closer to the theoretical mass of the unknown. The apex mass obtained most frequently should not be in error by more than 2.5 ppm when a mass increment of 5 ppm is used. The weighted average of the two SIR masses and the frequencies associated with their appearance as apex will provide the best estimate of the accurate mass for the unknown. To obtain a good average, 3-5 determinations were usually made.

For example, SIR masses 287.1675, 393.3378, and 405.2453 Da in Table 2 were very close to the theoretical masses and were the ordinate at the apex in all measurements. However, experimental accurate masses m/z 423.2953 and 505.4646 were weighted averages because in these cases two adjacent SIR masses (423.2946 and 423.2967 Da; 505.4636 and 505.4661 Da) were found to be the experimental maxima in some of the determinations. These results indicated that the theoretical masses for the two unknowns were between the SIR masses and would be best approximated by weighted averages of the SIR masses times their relative frequencies [e.g., 0.2(505.4636 Da x 3 + 505.4661 Da x 2) = 505.4646 Da]. The theoretical masses for the compositions assigned to the unknowns were very close to the weighted averages; these assigned compositions also matched the full scan mass spectral features of the compounds. In all cases, the experimentally determined accurate masses were within 2.5 ppm of the theoretical values.

Acid Orange 8.  The SIR accurate mass measurement technique was found to be suitable for another sample introduction and ionization mode. A commercial dyestuff, Acid Orange 8, with approximately 65% dye content was analyzed. This sulfonated azo dye was introduced by n-LC and analyzed by negative ion LSIMS.15  Accurate masses were determined using m/z 367 from the glycerol matrix as the lock mass.  The low abundance of the glycerol ions used for calibration limited the mass resolution to 10,000.  The correct mass, 341.0596 Da, was observed 14 times out of 27; 1 mass increment low (341.0562 Da) was seen six times, and 1 increment high (341.0630 Da) was observed seven times.

While mass calibration requires two reference masses as for other high-resolution techniques, the SIR method is unique in that it requires only one reference (lock) mass during data acquisition after the voltage steps are calibrated.  This advantage can be significant with LSIMS, where there is a high background and a weak calibrant signal from the commonly used glycerol matrix.  The lower mass calibration ion can be monitored with a low flow of matrix, but the higher mass caibration ion can be observed only with increased matrix flow.  After calibration, the matrix flow is reduced, and the flow of effluent from the LC column is increased to maximize sensitivity for analytes.

CONCLUSIONS

In these studies, the SIR accurate mass measurement technique provided very accurate and precise mass measurements for trace-level components of complex samples.  The technique is sufficiently general and rugged for use under El or LSIMS ionization with GC or LC sample introduction.  By this technique, accurate masses of bis(3,4-dichlorophenyl)diazene oxide, chlorinated diphenyl sulfides, aralkylamines, and a sulfonated azo dye were measured.  The technique typically provided the correct accurate mass the majority of the time and did not vary from that value by more than 1 mass increment (generally + or - 5 ppm), as set by the SIR descriptors.  The mass resolution, scan speed, sensitivity, precision, and accuracy afforded by this technique for accurate mass measurement were superior to those of the conventional KVE scanning method.  Chromatographic and mass peak profiles were obtained at 20,000 mass resolution.  This procedure provided good sensitivity with isolation of the desired component from analytical interferences.  The SIR technique was able to identify unknowns and confirm structure assignments in complex environmental samples for trace components in the presence of interferences, components that could not be identified using KVE scans, GC/IR, or NMR due to the complex nature of the samples and very low concentrations of the components being studied.

ACKNOWLEDGMENT.  The U.S. Environmental Protectjon Agency (EPA), through its Office of Research and Development (ORD), partially funded and collaborated in the research described here.  This work has been subjected to the Agency's peer review and has been approved as an EPA publication.  Mention of trade names or commercial products does not constitute endorsement or recommendation for use.

 REFERENCES

  1. Nier. A O. J. Am. Soc. Mass Spectrom. 1991, 2, 447-452.

  2. Svec, H. J. "Mass Spectrometry-Ways and Means-A Historical Prospectus" Retrospective Lectures, 32nd ASMS Conference on Mass Spectrometry and Allied Topics; San Antonio, TX, 1984: pp 47-63.

  3. Beynon, J. H. Biomed. Mass Spectrom. 1981, 8, 380-383.

  4. Meyerson, S. Org. Mass Spectrom. 1986, 21, 197-208.

  5. Sovocool. G. W.; Mitchum. R K; Tondeur. Y.; Munslow, W. D.; Vonnahme, T. L; Donnelly, J. R. Biomed. Environ. Mass Spectrom. 1988, 15, 669-676.

  6. Moseley, M. A.; Deterding, L. J.; deWit, J. S. M.;Tomer, K B.; Kennedy, R.T.; Bragg, N.; Jorgenson. J. W. Anal. Chem. 1989, 61, 1577-1584.

  7. U.S. Environmental Protection Agency, Statement of Work for Low/Medium Level Organics, Invitation for Bid Contract No. OLMO1.8, March 1991.

  8. Moseley, M.; Deterding, L.; Tomer, K; Jorgensen. J. Anal. Chem. 1991, 63, 1467-1473.

  9. Grange. A. H.; Brumley, W. C. Rapid Commun. Mass Spectrom. 1992, 6, 68-70.

  10. Grange, A. H.; Brumley , W. C. "Automated Writing of Selected Ion Recording (SIR) Descriptors and Plotting of Mass Peak Profiles from SIR Data" Proceedings of the 40th ASMS Conference on Mass Spectrometry and Allied Topics; Washington, DC. 1992; pp 1143-1144.

  11. Grange, A. H. "Interfacing Spreadsheet and Word Processing Software to the VG 250 Data System" Proceedings of the 40th ASMS Conference on Mass Spectrometry and Allied Topics; Washington. DC, 1992; pp 1145-1146.

  12. Sinkkonen. S.: Kolehmainen. E.; Laihia, K; Koistinen. J.; Rantio. T. Environ. Sci. Technol. 1993, 27, 1319-1326.

  13. Hunt, D. F.; Sethi. S. K. In High Performance Mass Spectrometry:  Chemical Applications; Gross. M. L., Ed.; ACS Symposium Series 70; American Chemical Society: Washington DC. 1978; p 175.

  14. Gross. M. L. J. Am. Soc. Mass Spectrom. 1994, 5, 57.

  15. Brumley. W. C.; Brownrigg. C. M.; Grange. A H. J. Chromatogr., 1994. 680, 635-642.


 

Analytical Environmental Chemistry
ICE Home Page

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


Local Navigation


Jump to main content.