PAPER

www.rsc.org/jaas | Journal of Analytical Atomic Spectrometry

Reversed phase ion-pairing HPLC-ICP-MS for analysis of organophosphorus chemical warfare agent degradation productsw
Douglas D. Richardson, Baki B. M. Sadi and Joseph A. Caruso*
Received 16th March 2005, Accepted 6th January 2006 First published as an Advance Article on the web 23rd January 2006 DOI: 10.1039/b503857j The separation and analysis of three organophosphorus chemical warfare degradation products is described. Ethyl methylphosphonic acid (EMPA, the major hydrolysis product of VX), isopropyl methylphosphonic acid (IMPA, the major hydrolysis product of Sarin (GB)), and methylphosphonic acid (MPA, the nal hydrolysis product of both) were the analytes and were separated by reversed phase ion-pairing high-performance liquid chromatography (RP-IP-HPLC) with the use of myristyl trimethylammonium bromide as ion-pairing reagent and an ammonium acetateacetic acid buer system (pH 4.85). An Agilent 7500ce inductively coupled plasma mass spectrometer (ICP-MS) equipped with collision/reaction cell technology was coupled to the chromatographic system for detection of 31P and 47PO1. Historically, ICP-MS detection of phosphorus has been limited due to its high rst ionization potential (10.5 eV) and the presence of severe nitrogen polyatomic interferences (such as 14N16O1H1 and 15N16O1) overlapping its only isotope at m/z = 31. Implementation of an octopole reaction cell with helium as the cell gas allowed for removal of the nitrogen polyatomic interferences and reduction of background signal. Detection limits for EMPA, IMPA, and MPA were found to be 263, 183 and 139 pg mL1, respectively, with separation in less than 15 min. The developed method was successfully applied to the analysis of spiked environmental water and soil samples.

Introduction
Recent increases in terrorist activity and the threat of chemical weapon attacks have lead to the demand for a rapid and reliable method for the analysis of chemical warfare agents (CWA) and their degradation products. As a result of the Chemical Weapons Conventions (CWC), which banned the production, acquisition, retention, and direct or indirect transfer of chemical weapons, destruction of all chemical weapons held in reserve was mandated.1,2 These chemicals, which include nerve and vesicant agents, pose a deadly threat not only to the human population but also to vital aqueous and agricultural resources (Table 1.).1,36 Based on these facts the development of a sensitive and selective analytical technique for the analysis of CWA and their degradation products is of high importance to ensure homeland security. Phosphorus containing nerve agents, along with their degradation products, present diculties for ultra-trace analysis due to their high polarity, low volatility, and lack of a good chromophore. Direct analysis of CWA degradation products provides an indirect technique for CWA detection. Previous studies have successfully utilized methods such as gas chromatography-mass spectrometry (GC-MS), ion mobility-mass spectrometry (IMMS), and liquid chromatography-mass spectrometry (LC-MS) for the analysis of organophosphorus containing degradation products with detection limits in the
Department of Chemistry, University of Cincinnati, McMicken College of Arts and Sciences, Cincinnati, Ohio 45221-0172, USA E-mail: joseph.caruso@uc.edu; Web: http//www.uc.edu/plasmachem/ w Presented at the 2005 Winter Conference on Plasma Spectrochemistry, January 30February 3, 2005, Budapest, Hungary.

ng mL1 range.4,5,7 However, considering the lethal doses reported in Table 1, lower detection limits in the pg mL1 range are desirable for such nerve agents and their degradation products. To achieve this lower level detection requires a more selective analytical detection technique such as inductively coupled plasma mass spectrometry (ICP-MS). Elemental speciation analysis by ICP-MS allows for high sensitivity, low level detection, and elemental selectivity, making it the instrument of choice for ultra-trace elemental speciation studies.814 Phosphorus (m/z = 31) analysis by ICP-MS, until recently, was limited due to its high rst ionization potential (10.5 eV) and polyatomic interferences, including 14N16O1H1 and 15N16O1 (m/z = 31).15 Sector-eld MS detection with ICP sources does provide a potential resolution enhancement but at the expense of losing part of the beam. For elements with high ionization potentials the throughput is clearly diminished. Recent developments in collision/reaction cell technology16,17 have allowed for the analysis of elements prone to isobaric and polyatomic interferences through removal by collisional dissociation (collision energy c bond energy), chemical reaction, and/or energy discrimination.9 Recently Sadi et al.9 applied ICP-MS detection to phosphorus containing herbicides through the use of state of the art collision/reaction cell technology. Separation of two phosphorus based general purpose herbicides, as well as aminomethylphosphonic acid (AMPA) by reversed phase ionpairing high-performance liquid chromatography (RP-IPHPLC)18 achieved detection limits of less than 32 pg mL1.9 In this study reversed phase ion-pairing chromatography was coupled with ICP-MS detection for the analysis of three organophosphorus degradation products of Sarin (GB) and
This journal is
 c

396 | J. Anal. At. Spectrom., 2006, 21, 396403

The Royal Society of Chemistry 2006

Table 1 Chemical warfare agents and degradation products Chemical warfare agent Agent liquid LD50/mg kg1a 0.14 Chemical warfare degradation products Degradation product pKa 2.16 Degradation product oral human LDLO/mg kg1

24

2.24

143428b

See above

pKa1 = 2.41 pKa2 = 7.54

Vapor form LD50 values range from B0.092 mg-min m3 (Agent MSDS).

Cerilliant MSDS.

VX. A detailed description of the ion-pairing chromatography employed as well as the selection of optimum ion-pairing reagent is provided. Helium collision/reaction cell optimization experiments for the removal of polyatomic interferences through collisional processes and by application of an appropriate energy barrier are also described. Analytical gures of merit for each species studied, ethyl methylphosphonic acid (EMPA), isopropyl methylphosphonic acid (IMPA), and methylphosphonic acid (MPA), are presented. Finally, the developed technique was applied to spiked samples of Little Miami (Ohio) River water, tap water, and soil samples.

Experimental
Reagents The three chemical warfare degradation products (ethyl methylphosphonic acid (EMPA), isopropyl methylphosphonic acid (IMPA), and methylphosphonic acid (MPA)) used were obtained from Cerilliant (Austin, TX) as 1 mg mL1 certied reference materials (CRMs). CRMs are used as standard analytical solutions for analysis of Schedule 1, 2, or 3 toxic chemicals, their precursors, and/or degradation products as mandated by the CWC for verication.1,5 Stock solutions of 10 mg mL1 for each degradation product were prepared through dilution in HPLC buer. Further dilutions of these stock solutions in HPLC buer, as well as preparation of standard mixtures over the range 20400 ng mL1, were performed as needed. Instrument tuning was accomplished through the use of a 30 ng mL1 adenosine 5 0 -triphosphate (Sigma, St. Louis, MO), corresponding to a phosphorus concentration of 5 ng mL1.
This journal is
 c

The reagents utilized throughout this experiment were of analytical grade prepared fresh daily through dilution of stock standards with DDI water. All water was prepared by passing through a NanoPure (18 MO) treatment system (Barnstead, Boston, MA). A 50 mmol L1 ammonium acetate (Fisher Scientic, Fairlawn, NJ) solution with 5 mmol L1 ion pairing agent and 2% methanol (TEDIA, Faireld, OH) at pH 4.85 was used as the chromatographic buer. Ion pairing agents used in these experiments consisted of: tetrabutylammonium hydroxide (Aldrich, Milwaukee, WI), myristyl trimethylammonium bromide (Aldrich, Milwaukee, WI), and hexadecyl trimethylammonium bromide (Sigma, St. Louis, MO). The buer was prepared fresh from stock solution before starting the experiments. Adjustment of the pH was accomplished through addition of glacial acetic acid (Fisher Scientic, Fairlawn, NJ). Helium gas (Matheson Gas Products, Parisppany, NJ), with a purity of 99.999%, was used as the collision/reaction gas throughout all experiments. Gas ow rate was optimized during instrument tuning prior to each experiment and controlled by the mass ow controller provided with the instrument (Fig. 1.) Environmental water samples used in this experiment were collected in polypropylene bottles from laboratory tap water and the Little Miami River in Cincinnati, OH. Both water samples were ltered through 0.20 mm Nalgene nyloncellulose acetate (CA) syringe lters (Nalge Nune International Corporation, Rochester, NY) to remove unwanted particulate matter from entering the system. Environmental soil samples were collected from top soil outside of the laboratory and potting soil kindly provided by the greenhouse at the University of Cincinnati. Preparation of the soil samples consisted
J. Anal. At. Spectrom., 2006, 21, 396403 | 397

The Royal Society of Chemistry 2006

Table 2 HPLC-ICP-MS instrumental parameters ICP-MS parameters Forward power Plasma gas ow rate Auxiliary gas ow rate Carrier gas ow rate Nebulizer Spray chamber Sampling depth Sampling and skimmer cones Dwell time Isotopes monitored (m/z) Octopole reaction system HPLC parameters Instrument Flow rate Injection volume Buer 1500 W (with shielded torch) 15.6 L min1 1.0 L min1 1.20 L min1 Glass expansion Microconcentric E2 1C (Scott double channel) 6 mm Nickel 0.1 s 31 P and 47PO1 He (Flow optimized prior to experiment) Agilent 1100 HPLC 0.5 mL min1 100 mL 50 mM ammonium acetate; 2% Methanol 5 mM Myristyl trimethylammonium bromide 4.85 Alltima C8 (3.2 150 mm) 5 mm

pH Column Fig. 1 Separation comparison of three ion-pairing reagents: tetrabutylammonium hydroxide (TBAH), hexadecyltrimethylammonium bromide (HAB), myristyltrimethylammonium bromide (TTAB).

A detailed description of ICP-MS operating conditions is provided in Table 2. Electrospray mass spectrometry (ESI-MS). The ESI-MS spectra were acquired in negative ion mode using a Finnigan LTQ (Thermo Electron Corporation, San Jose, CA) linear ion trap mass spectrometer equipped with an Ion Max source and Xcalibur (version 1.4 SR1) data software. Parameters were optimized by insertion of an EMPA (10 mg mL1) standard solution at 3 mL min1 using a syringe pump. The source parameters were as follows: sheath gas ow, 35 units; auxiliary and sweep gas ow, 0 units; spray voltage, 4.00 kV; and capillary temperature, 300 1C. Full scan MS and MS/MS spectra were acquired by insertion of the sample solutions at 3 mL min1. The collision energy for MS/MS experiments was 41%. Lyophilized peak fractions were dissolved in DDI water (14 mg mL1) and analyzed following the procedure described above.

of placing 1.0 g of solid material in 5.0 mL of DDI water and stirring for 15 min. The resulting solution was ltered through 0.20 mm Nalgene nyloncellulose acetate syringe lters (Nalge Nune International Corporation, Rochester, NY). All environmental samples were processed as blanks and 100 ng mL1 spiked mixtures (prior to ltration) of ethyl methylphosphonic acid, isopropyl methylphosphonic acid, and methylphosphonic acid. Instrumentation HPLC conditions. An Agilent 1100 (Agilent Technologies, Palo Alto, California) high-performance liquid chromatograph (HPLC) equipped with a binary pump, autosampler, vacuum degasser, thermostated column compartment, and diode array detector was used for the separation of the three chemical warfare degradation products. A C8 column (Alltima C8, 100 A, 3.2 150 mm, 5 mm, Alltech Associates Inc, Deereld, IL) with a guard column (Alltima C8, 7.5 3.0 mm, 5 mm, Alltech Associates Inc, Deereld, IL) was used for all separation experiments. A detailed description of the HPLC separation conditions is provided in Table 2. Inductively coupled plasma mass spectrometer (ICP-MS). An Agilent 7500ce (Agilent Technologies, Tokyo, Japan) ICP-MS equipped with shielded torch and collision/reaction cell technology was used for the element specic detection of 31P and 47 PO1 throughout this experiment. The collision/ reaction cell consisted of an octopole ion guide operated in rf only mode which also served for the removal of polyatomic interferences. Electronic coupling of the ICP-MS with the HPLC was accomplished through the use of a remote cable which allowed for simultaneous starting prior to each chromatographic run.
398 | J. Anal. At. Spectrom., 2006, 21, 396403

Results and discussion

Owing to the charged nature of the compounds of interest, ion-pairing chromatography was investigated as the chromatographic separation technique. The acid dissociation constants for the chemical warfare degradation products are: methylphosphonic acid (MPA) pKa1 2.41, pKa2 7.54; ethyl methylphosphonic acid pKa1 2.16, and isopropyl methylphosphonic acid pKa1 2.24 (Table 1). Based upon the acid dissociation constants a buer system (acetic acidammonium acetate; pKa 4.8) at pH 4.85 was used in the separation experiments. It was believed that the hydrophobicity and dierence in eective charges of the dierent species would allow for separation by the proposed chromatography. Tetrabutylammonium hydroxide (TBAH) was rst investigated as the ion-pairing reagent for the separation of the three species of interest. Baseline resolution was not achieved between the ion-pairs formed with this compound.
This journal is
 c

The Royal Society of Chemistry 2006

Fig. 2 Separation of 100 ng mL1 mixture of MPA, EMPA and IMPA.

Hexadecyltrimethylammonium bromide (HAB) was then investigated as an ion-pairing reagent due to its longer alkyl chain: however, baseline resolution was also not achieved. Myristyl trimethylammonium bromide (TTAB), consisting of an intermediate length alkyl chain, was then investigated as an ion-pairing reagent. A comparison of the separation achieved with each ion-pairing reagent is provided in Fig. 2. Myristyl trimethylammonium bromide along with an ammonium acetateacetic acid buer (pH 4.85) and 2% methanol for the mobile phase allowed separation of methylphosphonic acid, ethyl methylphosphonic acid, and isopropyl methylphosphonic acid with the selected column in less than fteen minutes (Fig. 3). ICP-MS detection Element specic detection by ICP-MS is a popular analytical technique based on the high sensitivity and selectivity oered by this instrument. In this experiment instrument selectivity was vital because of the need for element specic detection of phosphorus (m/z = 31) and the complex nature of the environmental matrices analyzed. Recently, phosphorus analysis by ICP-MS has grown in popularity due to the ability to remove nitrogen-based polyatomic interferences to 31P and the ability to ionize phosphorus suciently in spite of its high rst ionization potential. Other researchers depended upon the formation of PO1 (m/z = 47),19,20 or the use of high-resolution mass spectrometers to dierentiate between the polyatomic interferences and the phosphorus signal at m/z = 31.12,2123 Monitoring of PO1 in these experiments was performed to ensure no loss of 31P signal by oxide formation. This work consisted of the use of a helium collision cell for the removal of 14N16O1H1 and 15N16O1 interferences through a collision/energy discrimination process. Any fragmentation of the polyatomic interferences would need to overcome the nitrogenoxygen bond energy by using helium.9,21 After overcoming the polyatomic interferences with collisional dissociation, selective ion transmission through adjustment of the pole bias plays a vital role in analyte response. Helium was chosen as the collision gas for all experiments due to its light/nonThis journal is
 c

Fig. 3 Separation of blank and 100 ng mL1 spiked Little Miami River (Ohio) water.

reactive nature to allow for reduction of the background signal at m/z 31. Optimization of the helium gas ow rate was accomplished through the use of a mass ow control valve and constant introduction of 30 ng mL1 adenosine 5 0 -triphosphate (corresponding to 5 ng mL1 phosphorus) in buer. Phosphorus response versus helium ow rate was plotted and the ow rate corresponding to optimal signal with minimal background (buer signal m/z = 31) was selected (Table 2.). The gas ow used ranged from 3.54.0 mL min1 helium for all experiments based upon the optimization results. Analytical gures of merit Calibration curves were prepared through the use of standard mixtures ranging from 20 to 400 ng mL1. All regression coecients (r2) values were acceptable with the lowest value being 0.993. Detection limits (3s) based on three times the standard deviation of seven replicates of the blank peak areas (IUPAC) for the analysis of MPA, EMPA, and IMPA were found to be 139, 263, and 183 pg mL1, respectively. The detection limits for these three species are an improvement of at least one order of magnitude compared with those reported
J. Anal. At. Spectrom., 2006, 21, 396403 | 399

The Royal Society of Chemistry 2006

Table 3 Chemical warfare degradation product detection limits Chemical warfare degradation Analytical products method Ion mobility mass spectrometrya Detection limits/ ng mL1 56017005

LC-ESI-TOFb Electrophoresis microchip with contactless conductivity detectorc

8010003 488624

RP-IP-HPLC-ICP-MSd

0.1390.263e

Based on concentration producing a signal three times that of the noise. b Estimated in SIM mode at concentrations down to 50 ng mL1 for signal-to-noise ratio of 3 : 1. c Estimated from signal-tonoise characteristics (S/N = 3) of the response for 150 ng mL1 mixture. d Based on IUPAC. e This work.

in other analytical techniques (Table 3), although these detection limits were calculated based upon a concentration that would give a signal three times that of the noise. The precision for repeated injections of a 20 ng mL1 standard mixture was less than 1% for retention times and less than 6% for peak areas. Column recovery was calculated to evaluate the eciency of the sample preparation and separation techniques. These values ranged from 6986%. The analytical gures of merit are summarized in Table 4. Complex samples Investigation of complex sample matrix eects on the method led to the collection of Little Miami (Ohio) River water, tap water, top soil, and potting soil samples. Samples were treated with the sample preparation procedure described in the Experimental section. Figs. 4 and 5 show blank and spiked Little Miami River and laboratory tap water samples. An unknown peak around 600 seconds is observed in both samples and overlaps with the EMPA spike, resulting in an increased peak area. Sadi et al. reported unknown organophosphorus comTable 4 Analytical gures of merit based on 20 ng mL1 mixture Chemical warfare Detection Column degradation limit/pg recovery product mL1 MPA EMPA IMPA 139 263 183 86.2 69.2 73.0 RSD (%) peak area 2.75 5.39 5.96 RSD (%) retention time 0.38 0.55 0.65

Fig. 4 Separation of blank and 100 ng mL1 spiked laboratory tap water.

pounds in raw river water samples: however they were not identied.9 Figs. 6 and 7 show both blank and spiked top soil and potting soil samples. The blank chromatograms do not show the presence of any unknown peaks; however, the trend of the spiked top soil sample shows an increase in the EMPA peak compared with the standard sample chromatogram (Fig. 3.). Characterization eorts Initial characterization eorts of the unknown peak overlapping EMPA in the standard separation consisted of electrospray ionization mass spectrometry (ESI-MS) and time elapsed hydrolysis experiments. The interfering peak from blank Little Miami River water injections (retention time 600 s) was collected multiple times (>15), ltered through a strong cation exchange column to remove the ion pairing agent, and lyophilized. Mass spectrometric analysis of standard EMPA showed a strong molecular ion peak at m/z 123 (data not shown). Analysis of the concentrated river water fractions did not show the presence of EMPA (no m/z 123 peak). Based upon the mass spectrometric data EMPA was determined not to be the unknown species present in water and soil samples.
This journal is
 c

400 | J. Anal. At. Spectrom., 2006, 21, 396403

The Royal Society of Chemistry 2006

Fig. 5 Separation of blank and 100 ng mL1 spiked top soil.

Fig. 6 Separation of blank and 100 ng mL1 spiked potting soil.

Time elapsed experiments consisted of spiking DDI water with 10 mg mL1 EMPA at ambient temperature in a clear sample vial, resulting in hydrolysis to MPA. The sample was then analyzed by the developed method 2.5 hours after spiking (Fig. 7A). A large peak for EMPA was observed at 660 s with a minor peak for MPA seen at 450 s. The resulting sample was spiked with 10 mg mL1 MPA and an increase in peak area was observed at 450 s (Fig. 7B). The shift in retention times for the water spiked samples is attributed to dierences in their ionic strength and pH. Data collected from the time elapsed experiment conrms previous predictions that the unknown and interfering species in river water and soil samples was not EMPA due to the lack of hydrolysis to MPA in these complex samples. Investigation of other phosphorus containing species commonly found in environmental sample matrices led to the analysis of monobasic phosphate (pKa1 2.12, pKa2 7.2, pKa3 12.3) as the interfering species. Ammonium phosphate (SCP Science, Quebec Canada), potassium phosphate (Fisher Scientic, Fairlawn, NJ), and sodium phosphate (Fisher Scientic, Fairlawn, NJ) were analyzed by the developed separation method (pH 4.85; Fig. 8). All three phosphate salts showed retention times of 600 s overlapping with the EMPA standard. It is interesting to note that the net charges of EMPA and
This journal is
 c

monobasic phosphate are nearly identical, based upon their pKa1 values of 2.16 and 2.12, respectively. Identical net charges of these species at the buer pH of 4.85 can account for their overlapping migration through the ion-pairing separation.

Conclusion
In this work, the coupling of ion-pairing reversed phase HPLC with ICP-MS equipped with a collision/reaction cell allowed for trace analysis of three organophosphorus chemical warfare degradation products: MPA, EMPA, and IMPA. Ion-pairing chromatography oered the best separation based on interactions of the analyte between the stationary and mobile phases as well as slight charge dierences between the species of interest. This method provides a highly sensitive and selective technique with baseline separation of the three species within 15 min and detection limits of less than 263 pg mL1. Application of the developed method to environmental water and soil samples demonstrated the RP-IP-HPLC-ICP-MS technique to have high potential for complex sample speciation analysis. Improvements in the chromatographic resolution through investigation of alternative liquid and gas
J. Anal. At. Spectrom., 2006, 21, 396403 | 401

The Royal Society of Chemistry 2006

Fig. 7 (A) Time elapsed analysis of EMPA (10 mg mL1) in DDI water after 2.5 hours. (B) 10 mg mL1 MPA spike.

chromatographic separation techniques coupled with atomic mass spectrometric detection are currently under way.

Acknowledgements
The authors would like to thank Bryan M. Gamble of the FDA Forensic Chemistry Center (Cincinnati, OH) for ESIMS support. We also acknowledge Agilent Technologies for continued support of our work as well as NIEHS Grant ES04908 (supplement 3P42-E5004908-15S4) for partial funding of this research. Daisy-Malloy Hamburg and Kevin M. Kubachka are thanked for designing the graphical abstract illustration.

Fig. 8 Separation of 1.85 mg mL1 ammonium phosphate, 5 mg mL1 potassium phosphate, and 5 mg mL1 sodium phosphate.

References
1 United States Bureau of Arms Control and Disarmament Agency, Chemical Weapons Convention (CWC), Washington DC, April 29, 1997. 2 J. F. Weimaster, W. T. Beaudry, P. C. Bossle, M. W. Ellzy, L. G. Janes, D. W. Johnson, J. M. Lochner, S. G. Pleva and J. H. Reeder et al., J. Chem. Technol. Biotechnol., 1995, 64, 115128. 3 Q. Liu, X. Hu and J. Xie, Anal. Chim. Acta, 2004, 512, 93101. 4 J. R. Smith and M. L. Shih, J. Appl. Toxicol., 2001, 21, S27S34. 5 W. E. Steiner, B. H. Clowers, L. M. Matz, W. F. Siems and H. H. Hill Jr, Anal. Chem., 2002, 74, 43434352. 6 R. M. Wester, H. Tanojo, H. I. Maibach and R. C. Wester, Toxicol. Appl. Pharmicol., 2000, 168, 149152.

7 R. M. Black and R. W. Read, J. Chromatogr. A, 1998, 794, 233244. 8 J. A. Caruso, B. Klaue, B. Michalke and D. M. Rocke, Ecotoxicol. Environ. Saf., 2003, 56, 3244. 9 B. B. M. Sadi, A. P. Vonderheide and J. A. Caruso, J. Chromatogr. A, 2004, in the press. 10 J. A. Caruso, K. L. Sutton and K. L. Ackley in Elemental Speciation: New Approaches for Trace Element Analysis, eds. J. A. Caruso, K. L. Sutton and K. L. Ackley, 2002. 11 A. Montaser, Inductively Coupled Plasma Mass Spectrometry, Wiley-VCH, New York, 1998. 12 J. S. Becker, S. F. Boulyga, C. Pickhardt, J. Becker, S. Buddrus and M. Przybylski, Anal. Bioanal. Chem., 2003, 375, 561566. 13 S. Wilber and E. McCurdy, Agilent Application Note, 2001, pp. 59884286EN. 14 P. Leonhard, R. Pepelnik, A. Prange, N. Yamada and T. Yamada, J. Anal. At. Spectrom., 2002, 17, 189196.

402 | J. Anal. At. Spectrom., 2006, 21, 396403

This journal is

 c

The Royal Society of Chemistry 2006

15 C. D. Stalikas and C. N. Konidari, J. Anal. At. Spectrom., 2001, 907, 119. 16 S. D. Tanner, V. I. Baranov and D. R. Bandura, Spectrochim. Acta, Part B, 2002, 57, 13611452. 17 D. R. Bandura, V. I. Baranov and S. D. Tanner, Anal. Bioanal. Chem., 2001, 370, 454470. 18 P. C. Bossle, D. J. Reutter and E. W. Sarver, J. Chromatogr., 1987, 407, 399404. 19 A. H. Kudzin, D. K. Gralak, J. Drabowicz and J. Luczak, J. Chromatogr. A, 2002, 947, 129141.

20 Z. H. Kudzin, D. K. Gralak, G. Andrijewski, J. Drabowicz and J. Luczak, J. Chromatogr. A, 2003, 998, 183199. 21 C. Sietho, I. Feldmann, N. Jakubowski and M. Linscheid, J. Mass Spectrom., 1999, 34, 421426. 22 M. Wind, I. Feldmann, N. Jakubowski and W. D. Lehmann, Electrophoresis, 2003, 24, 12761280. 23 S. Kozono, S. Takahashi and H. Haraguchi, Anal. Bioanal. Chem., 2002, 372, 542548. 24 J. Wang, M. Pumera, G. E. Collins and A. Mulchandani, Anal. Chem., 2002, 74, 61216125.

This journal is

 c

The Royal Society of Chemistry 2006

J. Anal. At. Spectrom., 2006, 21, 396403 | 403