Anal. Chem. 1994,66, 230-235
Determination of Carbofuran by On-Line Immunoaffinity Chromatography with Coupled-Column Liquid Chromatography/Mass Spectrometry Geoffrey S. Rule,? Alexander V. Mordehal, and Jack Henion' Drug Testing and Toxicology, New York State College of Veterinary Medicine, Cornell University, 925 Warren Drive, Ithaca, New York 14850
Development of a new technology for combined on-line sample preparation and analyte confirmation is presented. This technology involves the use of antibodies for trace analyte extraction and enrichment directly from a complex matrix. The antibodies are used in columns designed for use with ordinary high-pressure liquid chromatographic equipment. The immunoaffinity columns are combined with conventional reversed-phase LC columns by use of column-switching techniques and coupled directly to an atmospheric pressure ionization quadrupole ion trap mass spectrometer equipped with a pneumatically assisted electrospray (ion spray) interface. A column of aldehyde-activated silica was used to prepare a column specific to carbofuran. This column demonstrated excellent specificity toward carbofuran and showed no binding of another, unrelated compound, fluometuron. Direct extraction and detection of carbofuran was demonstrated at low levels (40 pg/mL) in spiked water, but the real utility of immunoaffinity chromatography (IAC) is demonstrated by the online extraction and detection of carbofuran from a chemically complex, crude potato extract. Samples extracted using an IAC column indicate that superior purification is obtained with IAC in comparison with samples pumped directly onto a reversed-phase trapping column. A detection limit for carbofuran of approximately 2.5 ng/g of potato was obtained using an atmospheric pressure ionization quadrupole ion trap mass spectrometer. There is continued interest in the development of alternative means of sample preparation for trace analyte analysis, especially for chemically complex samples. The reasons for this include the desire to reduce time, expense, and hazardous wastes. At the same time there is considerable interest in developing methods that can be automated, thereby increasing sample throughput and reducing labor. Recently, there has been considerable development in the area of solid-phase extraction (SPE)I**as well as supercritical fluid extraction (SFE)3for alternative methods in sample extraction. Another method which has been suggested as being potentialiy useful for very selective analyte extraction is immunoaffinity chromatography (IAC).4,5
'
Present address: Analytical Chemistry Labs, Cornell University, NYSAES, Geneva, N Y 14456. ( I ) Horack, J.; Majors, R. E. LC-GC 1993, I / , 74-90. (2) Symposium on Solid-Phase Extraction in Environmental and Clinical Chemistry, San Francisco, CA, 1992; J . Chromatogr. 1993,629 (Symposium Volume). (3) King, J. W.; Hopper, M . L. J . AOAC Inr. 1992, 75, 375-378. (4) Van Emon, J. M.; Lopez-Avila, V. Anal. Chem. 1992, 64, 79-88.
230 Analytical Chemistry, Vol. 66, No. 2,January 15, 1994
In recent work,6 we described a method for on-line immunoaffinity chromatography with coupled-column HPLC for extraction and detection of two basic drugs directly from diluted urine solutions. In that report, a protein G immunoaffinity column was used in conjunction with liquid chromatography and mass spectrometry (LC/MS) for identification. Other workers haveutilized IAC for over twodozen small analyte molecule^.^ Some of these reports describe the direct combination of IAC with LC8 and even GC.9 There are also reports in which IAC is used for low-pressure extraction of analytes prior to various means of detection.I0J The objectives of this study were to evaluate the feasibility of combining these columns, for use in high-pressure IAC systems, on-line with LC/MS for analyte confirmation. This is referred to as IAC coupled-column LC/MS. In addition, we wanted to evaluate the selectivity of antibodies for analyte isolation in comparison with an ordinary reversed-phase column. In this report, covalently bound antibodies were utilized to extract the insecticide carbofuran from an environmental water sample and a food matrix. Two different, commercially available, activated columns were evaluated using polyclonal antibodies. Carbofuran was chosen as a model analyte for this study both because of the availability of commercial antibodies to this compound and because of its potential interest with regard to LC/MS. Carbofuran is a carbamate pesticide used to control insect and nematode pests on a variety of agricultural crops. Owing to their thermal instability, the carbamate pesticides are not readily determined by GC or GC/MS without prior derivatization.I2 Most researchers rely upon LC with UV or fluorescence detection for carbamate p e s t i c i d e ~ . I ~ , ~ ~ This report presents a preliminary evaluation of the feasibility of combining immunoaffinity columns on-line with coupled-column LC/MS. The advantages of such a combination include the capability of providing mass spectral ( 5 ) Katz, S . E.; Brady, M. J.-Assoc. Off. Anal. Chem. 1990, 73, 557-560. (6) Rule, G.; Henion, J. J . Chromatogr. 1992, 582, 103-1 12. (7) Farjam, A. Ph.D. Thesis, Free University, The Netherlands, 1991. (8) Nilsson, B. J . Chromatogr. 1983, 276, 413-417.
(9) Farjam, A.; Vreuls, J . J.; Cuppen, W. J. G. M.; Brinkman, U. A. Th.; de Jong, G . J. Anal. Chem. 1991, 63. 2481-2487. (IO) Gaskell, S . J.; Brownsey, B. G. Clin. Chem. 1983, 29, 677-680. ( I I ) Groopman, J . D.; Zarba, A. In Immunoassays for Trace Chemical Analysis; Vanderlaan, M., Stanker, L. H., Watkins, B. E., Roberts, D. W., Eds.; ACS Symposium Series 451; American Chemical Society: Washington, DC, 1985; pp 207-2 14. (12) Bellar, T. A.; Budde, W. L. Anal. Chem. 1988, 60, 2076-2083. (13) Marvin, C . H.; Brindle, I. D.; Hall, C.D.; Chiba, M. Anal. Chem. 1990, 62, 1495-1498. (14) Goewie, C. E.; Hogendoorn, E. A. J . Chromatogr. 1987, 404, 352-358.
0003-27C0/94/0366-0230$04.50/0
0 1994 American Chemical Society
confirmation of trace analytes directly from complex sample solutions via automated, on-line analysis.
EXPERIMENTAL SECTION Materials. The protein G and aldehyde-activated columns, both 2.1 mm i.d. X 30 mm, with 30-pm silica particles, 5001000-A pore size, were kindly supplied by Chromatochem Inc. (Missoula, MT). The N-hydroxysuccinimide ester(NHS-) activated columns (Affi-Prep lo), 4.6 mm i.d. X 30 mm, 0.5-mL bed volume, 40-60-pm beads of poly(methy1 methacrylate), 1000-A pore size, were purchased from Bio Rad Laboratories (Richmond, CA). Polyclonal antiserum to carbofuran (rabbit) was purchased from Biodesign International (Kennebunkport, ME). Purified rabbit IgG was obtained from Sigma Chemical Co. (St. Louis, MO). U1trafiltration membranes (YM-30 Diaflo ultrafilters), used in a Model 8050 stirred ultrafiltration cell, and Centriprep-30 centrifugation concentrators were purchased from Amicon Division (W. R. Grace and Co., Beverly, MA). HPLC Columns and Mobile Phases. The trapping column used for carbofuran was a Brownlee Labs, Inc. (Santa Clara, CA) RP-18 guard column, 3.2 mm i.d. X 15 mm with 7-pm particles. The analytical column used was either a DuPont Zorbax Rx-C18, from MAC-MOD Analytical (Chadds Ford, PA), 4.6 mm i.d. X 150 mm, with 5-pm particles or a Keystone Scientific (Bellefonte, PA) Hypersil ODS, 2.1 mm i.d. X 250 mm, with-5 pm particles. A mobile phase consisting of 35% or 40% ACN in H2O was used for the two columns, respectively. In both cases 5 mM NHdAc was added to the mobile phase and the pH was adjusted to 4.0 with concentrated formic acid. Eluant flow on the 4.6-mm-i.d. column was maintained at 1.OmL/min while that on the 2-mm-i.d. column was 0.4 mL/min. Phosphate-buffered saline (PBS) consisted of 0.01 M sodium phosphate buffer containing 0.15 M NaCl (pH 7.4). The solution used for desorption of analytes from the aldehydeactivated column was either 2% acetic acid (pH 2.6) or 0.2% formic acid (pH 2.6) in water. Instrumentation. The column switching and HPLC system6 consists of a Dionex (Sunnyvale, CA) gradient pump, one Model 5 10 Waters pump, a Waters automated switchingvalve (ASV), (Division of Millipore Corp. Milford, MA), the IAC column, a trapping column, and an analytical column. Additional hardware includes a Rheodyne Model 7 125 injector (Cotati. CA) with a 100-pL Tefzel loop (Upchurch Scientific Inc., Oak Harbor, WA) and a Kratos Analytical Spectroflow 783 variable-wavelengthUV detector (Ramsey, NJ) connected to a Hewlett-Packard 3390A recording integrator (Palo Alto, CA). A Waters Model 680 gradient program controller was used to control the ASV and solvent flow through the analytical column. Mass Spectrometry. The differentially pumped quadrupole ion trap mass spectrometer, designed and constructed in this laboratory, utilizes a pneumatically assisted electrospray (ion spray) interface and atmospheric pressure ionization (API) source.15 The instrument is controlled by the data station and electronics of a Varian (Palo Alto, CA) Saturn I1 ion trap. Briefly, ions are sampled through two orifices (1 30 and (15) Mordehai, A. V.;Hcnion, J . D.Rapid Commun. Muss Spectrom. 1993, 7, 205-209.
600 pm) separating atmospheric pressure from an intermediate-pressure region, and the latter region from a lower pressure region housing the ion trap. A 60-V potential drop, maintained across the two orificies, effects collision-induced dissociation (CID) of analyte molecules in the intermediatepressure region. The ion trap analyzer is maintained at 5 X Torr with helium gas. The electron multiplier is maintained at a pressure of 4 X 10-6 Torr by differential pumping provided by one of two turbomolecular pumps. The first low-pressure region is pumped by a high-speed (1060 L/min) rotary pump. The ion spray LC/MS interface is operated at 3-4 kV, and nitrogen (60 psi) is used as the coaxial nebulization gas. An accumulation time of 250 ms is used over a mass range of 100-300 amu. All experiments on the mass spectrometer were performed using the 2.1-mm4.d. analytical column. IgC Purification. Antibodies were purified on a protein G column in a manner similar to that described previously,6 except that larger amounts of antibody were used. Injections of 50 pL of neat antiserum were made onto the protein G column, and desorption with 2% acetic acid was made after every two injections. The desorbed IgG was collected in a beaker, surrounded by an ice bath, and immediately neutralized by dropwise addition of 2 M sodium hydroxide solution. The IgG solution was desalted and concentrated using an ultrafiltration membrane, with a molecular weight cutoff of 30 000. The initial IgG solution of 75 mL was reduced to 10 mL before transfer to an ultrafiltration tube used with a highspeed centrifuge. These tubes allowed concentration of the solution to 1.5 mL or less. At the same time, this concentration process allowed complete exchange of the Qriginal buffer solution with the recommended buffer for thecoupling reaction (either 0.1 M citrate or 0.1 M 4-morpholinepropanesulfonic acid). Reparation of Immunoaffity Columns. Covalent coupling of purified IgG to the aldehyde-activated column was carried out in 0.1 M citrate buffer (pH 5.5) in the presence of 20 mg/mL NaCNBH3. A 1-mL tuberculin syringe, equipped with plastic Luer hub adapters, was used to wash the column with 10 volumes of citrate buffer. The purified IgG solution was then applied in the same fashion, but with a second syringe attached to the outlet end of the column. This allowed collection of the flow-through protein solution for subsequent reapplication to the front of the column. After this step was repeated three times, the binding reaction was allowed to proceed for 1 h. The column was then flushed with 10 mL of PBS. The column was then connected to an LC pump, and PBS pumped through the column until a stable baseline was obtained at 280 nm. The extent of coupling was monitored during the reaction by injection of small aliquots of the coupling solution onto the protein G column. The IgG was desorbed as described previously,6 and the column effluent was monitored at 280 nm. The N-hydroxysuccinimide coupling reaction was carried out as described in the product literature. To deactivate remaining N H S coupling groups, a 1 M solution of ethanolamine (pH 8) was injected onto the column in 10 100-pL aliquots and pumped through at 0.1 mL/min. The amount of IgG that coupled to the columns was determined from small aliquots of IgG solution that were Analyrcal Chemistry, Vol. 66, No. 2, January 15, 1994
231
Table 1. Sequence of Timed Events for Potato Extract Analysis
time (min) 0.00
10.00 15.00
15.50
20.50
step begin pumping potato extract (or water sample), diluted in PBS to 20 mL, through IAC column at 2 mLimin. Automated switching valve (ASV) in starting position. Column effluent goes directly to waste wash IAC column with PBS for 5 min at 3 mLimin switching valve turned to alternate position. Flow is directed from IAC column through trapping column begin 5-min desorption with pH 2.6 formic acid at 3 mL/min switching valve turned to starting position, to back-flush trapping column on-line to LCiMS return IAC column mobile phase to PBS
kept both before and after the coupling reaction. These solutions were injected onto the protein G column and quantified by use of the peak areas obtained upon desorption with 2% acetic acid. These peak areas were then compared with those obtained from injections of a sheep IgG standard solution. Coupled-Column IAC. The system used for the analysis of samples spiked with carbofuran is essentially the same as described previously.6 Samples are pumped through the IAC column for the desired amount of time before switching to a PBS washing step (Table 1). During both of these steps the IAC column effluent is directed to waste. This is followed by actuating the automated switching value, so that the IAC column effluent is directed through the trapping column. After a short (30 s) reconditioning period, the desorption solution is pumped through the IAC column. The desorbed analyte is trapped on the trapping column and subsequently "backflushed", with the organic mobile phase used to effect separation on the analytical column. This is brought about by returning the ASV to its original position. This sequence of events is completely automated by pneumatic actuation of the ASV, which was controlled by the gradient controller, and by programming the gradient pump. All analyses utilized a desorption step of 5 min with 0.2% formic acid (pH 2.6) at 3 mL/min. Capacity Determination of Anti-Carbofuran IAC Column. A solution of PBS containing carbofuran at a level of 10 ng/ mL was pumped through the IAC column at 3 mL/min with the column connected directly to the detector (220 nm). The baseline was monitored until a stable absorbance reading was attained, at which point the column was flushed with PBS for 5 min. The column was then connected to the coupled-column system and subjected to the desorption cycle as described in the preceding section. Analysis of Surface Runoff Water by IAC Coupled-Column LC/MS. A sample of surface runoff water was collected near Ithaca, NY. The water was filtered through a 0.2-pm filter disk by vacuum filtration and used to make a PBS solution. The resulting solution, of which 90% was sample water, was then spiked at levels of 40, 200, and 1000 pg/mL of sample water with carbofuran in methanol. The equivalent of 50 m L of sample water (55.6 mL total) was then pumped through the IAC column a t 4 mL/min. The remainder of each analysis was carried out as described in Table 1. Analysis of Potato Digest by IAC Coupled-Column LC/ MS. Preparation of potato extracts was performed in a manner 232
Analytical Chemistry, Vol. 66, No. 2, January 15, 1994
utilized by the manufacturer of carbofuran (personal communication with Dr. G. Singer, FMC Corp., Princeton, NJ). This involved cutting-up potato and boiling it in aqueous 0.25 N hydrochloric acid for 1 h. Upon cooling, the mixture was filtered through glass wool and an aliquot taken for analysis. The aliquot was again filtered, using a 0.2-pm filter disk attached toa Luer-lock syringe, and diluted to the appropriate concentration with PBS. The resulting solution was adjusted to pH 7.4 using 2 N sodium hydroxide solution before spiking with carbofuran and introduction onto the IAC column. RESULTS AND DISCUSSION The NHS column was evaluated in this research because of its relative inexpensiveness. These columns can tolerate a pressure of 1000 psi, according to manufacturer's literature, although the recommended operating pressure is 300 psi. The column matrix consists of poly(methy1 methacrylate) beads that are stable from pH 1 to 12. The aldehyde-activated column is based on macroporous silica beads that are stable at p H 2-10.5. This column is designed for high-pressure separations. A column based on this same matrix showed excellent properties in previous studies,6 including very little nonspecific binding. An N H S column prepared with anti-carbofuran IgG did not show desirable binding characteristics, since two other compounds, fluometuron and propoxur, also bound to the column. This indicated that retention of these analytes was taking place primarily through nonspecific adsorption to the polymeric matrix. Because of nonspecific binding, no further work was conducted with the NHS column. Instead, anticarbofuran IgG was covalently coupled to a column based on silica beads. Approximately 700 pg of IgG remained on the column after the coupling procedure and one desorption cycle. After the first desorption, the column was connected to the coupled-column system. It had been previously established by us (data not shown) that carbofuran would bind to a C18 trapping column under conditions of low pH (2.6) for a period of at least 5 min at 3 mL/min. Two separate capacity determinations for carbofuran indicated retention of 26 and 28 ng by the IAC column. Based on the assumption that 10% of the total antiserum IgG is analyte specific, and with two binding sites per IgG molecule, this represents 13% retention of binding activity. An evaluation of the IAC column specificity toward carbofuran was made by pumping 10 mL of PBS containing a mixture of carbofuran and fluometuron, each at a level of 2 ng/mL, through the column and subsequently performing the desorption and column-switching procedure (Figure 1). The analysis performed by direct injection of a standard solution onto the trapping column, without the IAC column, indicates the retention times of these two compounds (Figure 1A). As can be seen in Figure l B , carbofuran was retained by the IAC column after the PBS washing step, while fluometuron was not. This indicates the high degree of selectivity of the anti-carbofuran antibody, as well as the absence of nonspecific binding on this column matrix. The column was next used for the determination of carbofuran that had been spiked into surface runoff water with on-line mass spectrometry as the means of detection. Determinations were made at levels of 40, 200, and 1000
Carbofuran
I n
i'
b'31i
21 123
Flgure 3. Analysis of 50 mL of surface runoff water, spiked with carbofuran at a level of 200 pg/mL, by IAC coupied-coiumn LC/MS: (A, top) background-subtracted full-scan mass spectrum averaged over peak at retentiontime 4.3 min; (B, bottom) summed ion current profile for m/z 222, 165, and 123.
TO1
tU+Hl+
1 !\
165 123
I
/I
' I ' / ' I ' I / I ' l '
i
'
I
'
I
I
'
I, I
,
'
l.#lll I
'
I
'
~
I,
' I lY ~'
I
~
I '
-
50 1:48
!,
II I
., : , : ,
,
, , 58 1:48
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,,, , ,
, I , , ,
188
150
288
3:2a
5:ea
6:48
,
1
,
I
,
,I
scA)( II (nln)
TIME
Figure 2. Direct injectionof 20-ng carbofuranstandard onto trapping column, with subsequent back-flushing to the analytical column and ion trap mass spectrometer (coupled-column LCIMS): (A, top) fuilscan mass spectrum: (E, bottom) summed Ion current profile of mlz 222, 165, and 123.
pg/mL carbofuran in water. A full-scan mass spectrum, and a summed ion current profile including three ions from carbofuran, were obtained by direct injection of a 20-ng carbofuran standard onto the trapping column (Figure 2). This full-scan ( m / z 100-300) mass spectrum from the API ion trap system shows abundant (M + H)+molecular weight information at m / z 222 plus two fragment ions at m / z 165 and 123. In Figure 3 the mass spectrum and summed ion current profile for the same three ions are shown following IAC coupled-column LC/MS analysis of surface runoff water. In this example, the water was spiked at a level of 200 pg/mL. Examination and comparison of these mass spectra provide confirmation of analyte identity. The limit of detection was determined to be approximately 40 pg/mL based on a 5:l signal-to-noise ratio for the summed ion current profile of m/z 222, 165, and 123 (Figure 4). This is 1000 times below the current regulatory level of 40 ng/mL for carbofuran in water.'6 Comparison of the recovery obtained by IAC analysis of spiked water samples with direct injections of carbofuran
188
150
3:28
5:BB
288
SCAN
6:48 TIME
x
bin)
Figure 4. Analysis of 50 mL of surface runoff water, spiked with carbofuran at a level of 40 pg/mL, by IAC coupled-column LC/MS: (A) total ion current profile, (6)summed Ion current profile for mlz222, 165. and 123.
standards onto the trapping column indicated that recovery is lower using IAC. In addition, the percentage recovery with IAC decreases as the concentration of analyte increases. These results were anticipated because of the relatively low binding capacity of the column and are similar to those observed previously.6 The ease with which a reversed-phsae trapping column can concentrate analytes from a clean aqeuous solution makes it questionable whether there is any advantage to utilizing IAC extraction. This is of particular concern when one considers that multiple analytes can be trapped by a reversed-phase column whereas an IAC column is useful only for the analytes for which it was produced. The real advantage of IAC might lie in performing on-line extraction of more complex sample matrices. To explore this possibility, potato was chosen, because of its availability and potential interest with regard to carbofuran. The UV chromatogram obtained by direct injection of a 20-ng carbofuran standard onto the IAC column, followed by coupled-column LC, is shown in Figure 5A. This chromatogram shows the retention time behavior and appearance of carbofuran. The arrow indicates the carbofuran peak at the expected retention time for these experimental conditions. Figure 5B shows the UV chromatogram obtained by IAC/ LC/UV analysis of a crude potato extract. This analysis was (16) Code of Federal Regulations, 4 0
CFR 141.61, July 1, 1992; p 687.
Analytical Chemistry, Vol. 66, No. 2, January 15, 1994
233
1885:
123
58 1:48
158 5:88
188 3:ze
288 6:48
SCA)(
TINE
a
bin)
Figure 7. Coupled-column LC/MS analysis of potato (0.25 g equiv in 20 mL of PBS) obtained without IAC by pumping extract directly onto trapping column. Extract spked wlth carbofwan at level of 100 ng/g. After washing and "desorptlon" ste , the trapping column was backTotal ion current profile; (8-D) flushed to the analytical column: extracted ion current profiles for m/z 222,165, and 123, respectively.
K
Figure6. IAC cwpled-cdumn LClMSanalysis of potatoextract (0.25 g equiv in 20 mL of PBS). Extract spiked with carbofuran at level of 100 ng/g and pumped through IAC column at 2 mL/min: (A) total ion current profile; (B-D) extracted ion current profiles for mlz 222, 165, and 123, respectively.
obtained by pumping the extract equivalent of 0.25 g of potato through the IAC column, and according to the timed events outlined in Table 1. The extract was spiked with carbofuran at a level of 100 ng/g. While the peak in the UV chromatogram (denoted by arrow) is small, this was sufficient to produce a noticeable peak in the total ion current profile generated by the ion trap mass spectrometer (Figure 6A). Integration of the extracted ion current peaks (Figure 6B-D) yielded area ratios that were within 20%of those obtained by the standard injection shown in Figure 2. These data provide additional confirmation of analyte identity. The virtues of IAC are especially apparent in Figure 5C and Figure 7, where the same spiked potato extract (0.25 g 284
~ n a ~ chemistry, c a ~ vd. 66, NO. 2, January 15, 1994
equiv, 100 ng of carbofuran/g) was pumped directly onto the trapping column without the sample cleanup benefits of IAC. The extracted ion current profiles (Figure 7B-D) show a very weak peak only for m l z 222 at the expected retention time for carbofuran. This suggests that either carbofuran was not retained on the trapping column or coeluting material has greatly suppressed the electrospray ionization of carbofuran. Examination of the corresponding UV chromatogram (Figure 5C) suggests either explanation may be a possibility. It is evident from Figure 5C that there is an enormous amount of material that is loaded onto the trapping column, which is subsequently eluted onto the analytical column upon backflushing. After more than 30 min without a return of the UV absorbance baseline, the analytical column was disconnected from the system and washed with an ACN/MeOH/H*O mixture overnight. Figure 5C suggests that the trapping column was rapidly saturated with starch and/or other components of the potato extract, thereby rendering thecolumn incapable of retaining the analyte. Alternatively, the large amount of material eluting from these columns to our ion spray interface may have interfered with the ionization process. Because potato is known to be approximately 20%starch (wet weight),17 it is assumed that this material consists primarily of long-chain oligo- and polysaccharides. These results serve to demonstrate the degree of sample purification that can be obtained using IAC. The on-line mass spectrometric detection (Figure 7A-D) also points out that complex samples require prior purification and that IAC can perform this for ion spray LC/MS. By pumping a 2 g equiv of potato extract through the IAC column it was possible to determine the presence of carbofuran in potato at a level of 2.5 ng/g (Figure 8; estimated limit of detection). These levels are far below current U.S.Environmental Protection Agency regulatory levels for potatoes (1 ppm for carbofuran and its metabolite)I8 and comparable to determinations made (17) Parson, D. The Chemicul Analysis of Foods; Churchill Livingstone: Edinburgh, 1976; p 164. (18) Code of Federal Regulationr, 40 CFR 180.254, July 1, 1992; p 341. (19) Stan,H.-J.; Klaffenbach, P. Freseniw J. Anal. Chem. 1991, 339, 151-157. (20) Pleasance, S.; Anacleto, J. F.;Bailey, M.R.; North, D. H.J. Am. Soc. Mass Spectrom. 1992.3, 378-397.
obtained using IAC with coupled-column LC. By combining an atmospheric pressure ionization ion trap mass spectrometer with this technology, full-scan mass spectra may be obtained to provide confirmation of target analytes with minimal sample preparation. Future work in this area may focus on quantitation and the dynamic range over which analyses can be conducted. It is believed that developments in antibody production technology, such as recombinant production of antibody F, fragments, should make this technology even more attractive in the future. 123
58
188
1:40
3:28
158 5:m
zm
6:18
m
a
TIHE h l n )
Flguro 8. IAC coupled-column LC/MS analysis of potato extract (2 g equhr In 20 mL of PBS). Extract spiked wlth carbofuran at level of 2.5 ng/g and pumped through IAC column at 2 mL/mln: (A) total ion current proflle; (8-D) extracted ion current profiles for mlz 222, 165, and 123, respecthrely.
by others using gas chromatography/mass spectrometry'9 and LC/MSZ0
CONCLUSIONS This work demonstrates the feasibility of utilizing automated IAC as an on-line method of sample preparation for ion spray mass spectrometric analysis of a complex sample. When carbofuran is spiked into a crude potato extract, results show that a considerable degree of sample cleanup can be
ACKNOWLEDGMENT The authors thank Chromatochem Inc., for the donation of several activated IAC columns, the Dionex Corp. for use of the HPLC gradient pump, the Eastman Kodak Co. for financial support for the construction of the benchtop API ion trap mass spectrometer system, and Dr. A. Mordehai for his assistance with using the ion trap system. This work was also supported in part by a grant from the Cornel1 Center for Advanced Technology (CAT) in biotechnology, which is sponsored by the New York State Science and Technology Foundation, a consortium of industries and the National Science Foundation. Received for review July 8, 1993. Accepted October 22, 1993.' Abstract published in Advance ACS Absrrocrs, December 1, 1993.
Analytical CkmisttY, Vol. 86, No. 2, January 15, 1994
235
