MS

Chemosphere 46 (2002) 1451–1455 www.elsevier.com/locate/chemosphere

Optimization of the ionization conditions for the trace analysis of PCDD/PCDF with ion trap MS/MS Yukio Kemmochi *, Kaori Tsutsumi, Ken-ichi Futami Environmental R&D Department, EBARA Corporation, 4-2-1 Honfujisawa, Fujisawa 251-8502, Japan

Abstract Commercial ion trap mass spectrometer provides easy-to-operate MS/MS analysis for the determination of PCDD/ PCDF. The limit of quantification is appropriately low (0.2 pg for 23478-P5CDF) because all the stages are performed in the trap and sample losses associated with the ion transportation are minimized. However, if excessive ions are injected into the trap, its electrical fields are distorted and an overall reduction in performance arises. Ionization condition is an important parameter as it affects the amount of the total ions produced. If the amount of interfering compounds are negligible, such as standard solution or cleaned-up sample, lower ionization condition (e.g. electron energy: 30 eV, emission current (EC): 150 lA) is preferable. On the contrary, in case excessive interfering ions are coexisting with PCDD/PCDF, such as crude extract or semi-cleaned-up sample, the ionization condition should be high (e.g. electron energy: 90 eV, EC: 350 lA) for the reproductive quantification. Ó 2002 Elsevier Science Ltd. All rights reserved. Keywords: Ion trap; Space charge; Tandem mass spectrometry; Analysis; Polychlorinated dibenzo-p-dioxins and dibenzofurans; Electron energy; Emission current; Ionization condition; Reproducibility

1. Introduction Conventional PCDD/PCDF analysis with ‘‘single mode-of-the-operation’’ mass spectrometer requires labour intensive multi-step clean-up procedures. Tandem mass spectrometry (MS/MS) technique is highly selective due to its characteristic PCDD/PCDF fragment ions produced by the secondary ionization (Karasek and Clement, 1988), so that the interference caused by the sample matrices can be minimized. Along with the gas chromatography, ion trap MS/MS presents highly reliable information which helps identification of the analytes (Plomly et al., 1995; Leonards et al., 1996). Consequently, the MS/MS makes it possible to determine PCDD/PCDF in interfering com-

*

Corresponding author. Fax: +81-466-82-2859. E-mail address: [email protected] (Y. Kemmochi).

pounds and the clean-up procedures are minimized eventually. Theoretically, the instrument can detect all the ions formed in the ion source chamber. The sensitivity is supposed to be increased as the amount of the ions increase. However, if excessive interfering ions are injected into the trap along with PCDD/PCDF, space charge effects (ion–ion coulombic interactions) take place and the accuracy of the analytes mass assignments decreases. As a result, the space charge leads to a significant decline in quality of the analysis. The aim of this study is to perform the reproducible PCDD/PCDF quantitative analysis with ion trap MS/MS. PCDD/PCDF in different matrices; standard solution, wastewater extract (negligible interfering compounds) and soil extract (excessive interfering compounds) were quantified repeatedly and their reproducibility were evaluated. The results showed that the ionization condition was an important parameter for the reproducible PCDD/PCDF analysis with ion trap mass spectrometer.

0045-6535/02/$ - see front matter Ó 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 5 - 6 5 3 5 ( 0 1 ) 0 0 2 6 5 - X

1452

Y. Kemmochi et al. / Chemosphere 46 (2002) 1451–1455

2. Experimental 2.1. Sample preparation Both soil and wastewater extracts were evaporated to approximately 1 ml, then transferred onto a sulfuric acid coated diatomaceous earth column. PCDD/PCDF were eluted with 30 ml of dichloromethane/n-hexane (1:2). The eluate was evaporated to approximately 0.5 ml, then transferred onto a silica-gel column. The analytes were eluted with 12 ml of toluene and transferred to suitable glass tubes and concentrated to 0.2 ml under vacuum. Concurrently, these extracts have been quantified by the conventional PCDD/PCDF analytical method with double-focus mass analyzer (Takasuga et al., 1992) for the experimental references. EPA-1623 CS1 13 C-labeled PCDD/PCDF mixture (Wellington Laboratories, Ontario, Canada) was used as a standard solution. A wastewater extract was analyzed by the conventional method prior to this study. Then, dioxin standards were added to the wastewater extract in order to prepare ‘‘Synthetic’’ sample. The isomers and the concentrations of the standards in the final volume were as follows: 13 Clabeled 12378-P5CDD (5 ng/ml), 13 C-labeled 2378T4CDF (5 ng/ml), 12378-P5CDF (5 ng/ml), 13 C-labeled 12378-P5CDF (5 ng/ml), 23478-P5CDF (0.2 ng/ml), 13 Clabeled 23478-P5CDF (5 ng/ml) and 13 C-labeled 234678H6CDF (5 ng/ml). 2.2. Ionization condition MS/MS analysis was performed on a Polaris (ThermoQuest, Austin, TX, US) ion trap mass spectrometer. The MS/MS condition has already been published (Kemmochi and Arikawa, 1999). PCDD/PCDF were ionized in the external ionization chamber with the

thermal electron emitted from the filament. The voltage/ current of the filament and the temperature of the chamber are the parameters for the ionization. The electron energy (EI) was set from 30 to 100 eV. The emission currents (EC) were 150, 250 and 350 lA. The temperatures of the ion source were 200 and 250 °C. The optimized condition may vary depending on the isomers. In this study, the ionization has been optimized for the following 2378 isomers; P5CDD, T4CDF, P5CDF and H6CDF, since they shared the major part in the TEQ contribution (Oser et al., 1998).

3. Results and discussion 3.1. Sensitivity The ion trap MS/MS was appropriately sensitive as a commercial quadrupole mass spectrometer so that 0.2 pg of 23478-P5CDF in the synthetic wastewater extract could be detected with symmetrical peak shape (Fig. 1). The signal-to-noise ratio was 8 and was acceptable for the quantification. The ratio of m/z; 274.9:275.9:276.9 was 1:0:1 and fitted to the theoretical ratio (0.8:0.1:1). Theoretically, as the electron energy decreases, the response factor of the Mþ ion increases in consequence of the reduction of the fragmentation. EC affects the number of the electrons emitted from the filament. The higher the EC is, the more the electrons are emitted, and then, the more the PCDD/PCDF are ionized. Consequently, higher EC provides higher response factor. As shown in Fig. 2, the theoretical curves were obtained when the EC was under 250 lA. However, while the EC was augmented to 350 lA, and concurrently the electron energy was reduced under 40 eV, the response factor dropped significantly. Theoretically, 350 lA/30 eV was

Fig. 1. 0.2 pg of 23478-P5CDF.

Y. Kemmochi et al. / Chemosphere 46 (2002) 1451–1455

Fig. 2. Ionization condition and response factor.

supposed to provide the highest response factor. But in practice, the result was in inverse relation. This phenomenon can be explained by the space charge effects. At 350 lA/30 eV, excessive ions were produced in the ion volume and injected into the trap. Eventually, the space charge took place and inaccurate mass assignment led lower response factor. 3.2. Reproducibility As far as the amount of the interfering ions in the trap was negligible, the space charge effect did not affect

1453

the reproducibility. The Fig. 3 shows the changing of the chromatograms of the cleaned-up wastewater extract. The Fig. 3B was the chromatogram of 12378-P5CDF at 250 lA/30 eV. As mentioned above, the space charge occurred when the ionization conditions were raised up to 350 lA/30 eV (Fig. 3C) and the response factor fallen sharply (10,079 > 1404). It indicates that 350 lA/30 eV was over-ranged. On the other hand, no peak distortions were observed and the peak ratio of the native and the 13 C-labeled 23478-P5CDF remained the same (0:11 > 0:11). No matter what the ionization conditions were, the relative standard deviation of the native/13 Clabeled peak ratio were almost the same (Table 1). In other words, the response factor was over-ranged, yet stable. Thus the reproducibility was fairly good enough throughout the analyses. The deviation has been calculated followed by the equation below, where X was the quantified value of 2378-isomers (P5CDD, T4CDF, P5CDF and H6CDF). P deviation ¼ 1!N



X  Xmedian Xmedian N 1

2 ð1Þ

On the contrary, in case the excessive interfering ions were coexisting with the target analytes ions in the trap, the space charge led poor reproducibility (Fig. 4 and Table 2). The Fig. 4 shows the changing of the

Fig. 3. Ionization condition and peak shape (cleaned-up wastewater extract).

1454

Y. Kemmochi et al. / Chemosphere 46 (2002) 1451–1455

Table 1 Relative standard deviation of the quantification (wastewater extract) Emission current (lA)

Electron energy (eV) 30

40

50

60

70

80

90

100

150 250 350

0.5% 0.8% 4.4%

1.1% 0.5% 1.1%

0.4% 0.7% 0.3%

0.6% 1.1% 0.7%

1.6% 0.5% 0.5%

0.7% 1.4% 2.0%

0.4% 0.2% 0.3%

3.6% 0.4% 0.6%

Fig. 4. Ionization condition and peak shape (semi-cleaned-up soil extract).

Table 2 Relative standard deviation of the quantification (soil extract) Emission current (lA)

Electron energy (eV) 30

40

50

60

70

80

90

150 250 350

1.0% 104% 511%

1.0% 325% 7233%

1.7% 0.6% 374%

1.3% 4.0% 1.4%

5% 1.0% 11%

1.7% 3.3% 6.6%

1.5% 0.5% 2.3%

chromatograms of the semi-cleaned-up soil extract. As the EC decreased from 350 to 150 lA (Fig. 4, C  A), the sum of Mþ ions decreased since the number of the emitted electrons decreased. The reduction of the interfering ions made the peak distortion restored and the reproducibility improved. In the meantime, as the electron energy increased from 70 to 90 eV (Fig. 4, D  F), the sum of Mþ ions decreased by the fragmentation.

Again, the reduction of the interfering ions made the peak distortion restored and the reproducibility improved. These results showed that the ionization condition were meaningful parameters for the reproducible PCDD/PCDF quantification with ion trap MS/MS. The excessive ions came from semi-cleaned-up soil extract, coexisting with PCDD/PCDF in the trap, could cause the space charge effect and deteriorate the quan-

Y. Kemmochi et al. / Chemosphere 46 (2002) 1451–1455

1455

Table 3 Quantification of 23478-P5CDF (pg/g) (6.5 pg/g by HRMS) Emission current (lA)

Electron energy (eV) 30

40

50

60

70

80

90

150 250 350

6.7 31 59

6.5 49 206

6.3 6.5 52

6.2 6.9 8.1

9.6 6.7 14

6.3 8.8 11

6.1 6.6 6.6

tification. However, if the ionization conditions were set at 90 eV/350 lA, these excessive ions were suppressed and did not affect the reproducibility. Meanwhile, the quantification results at this ionization condition were the same as the experimental references analyzed by the conventional method with HRMS (Table 3). It suggests that as long as the quantification are reproducible, semicleaned-up samples analyzed by ion trap MS/MS and corresponding cleaned-up samples analyzed by HRMS produce the comparable results.

4. Conclusion Excessive interfering ions coexisting with PCDD/ PCDF in the trap cause the space charge effects and lead the analysis to irreproducible quantification. Ionization conditions are meaningful parameters for the reproducibility. Once the ionization conditions are optimized, ion trap MS/MS performs reproducible PCDD/PCDF quantification. Furthermore, as long as the quantification is reproducible, semi-cleaned-up samples analyzed by ion trap MS/MS and corresponding cleaned-up samples analyzed by HRMS produce the comparable results. It indicates that the conventional labor-intensive PCDD/PCDF clean-up procedure can be simplified by

means of the ion trap MS/MS with optimized ionization conditions.

References Karasek, F.W., Clement, R.E., 1988. Basic gas chromatography-mass spectrometry; principles and techniques. Elsevier, Amsterdam, ISBN: 0-444-42760-0. Kemmochi, Y., Arikawa, A., 1999. Rapid PCDD/PCDF screening method for fly ash with ion trap MS/MS. Organohalogen Compd. 40, 161–164. Leonards, P.E.G., Brinkman, U.A., Cofino, W.P., 1996. The use of gas chromatography with ion-trap MS/MS detection for the determination of planar PCBs in biota and sediment. Chemosphere 32, 2381–2387. Oser, H., Thanner, R., Grotheer, H.-H., 1998. Continuous monitoring of ultratrace products of incomplete combustion during incineration with a novel mobile jet-REMPI device. Chemosphere 37, 2361–2374. Plomly, J.B., Mercer, R.S., March, R.E., 1995. Optimization of the quadrupole ion storage mass spectrometer for the tandem mass spectrometric analysis of dioxins and furans. Organohalogen Compd. 23, 7–12. Takasuga, T., Inoue, T., Ohi, E., 1992. Evaluation of a highresolution gas chromatography/high-resolution mass spectrometry method for the determination of dioxin and related compounds. J. Environ. Chem. 2, 599–613.