On-Line Derivatization into Precolumns for the Determination of Drugs

Rosa Herra´ ez-Herna´ ndez, Pilar Campıns-Falco´ ,* and Adela Sevillano-Cabeza. Department of Analytical Chemistry, University of Vale`ncia, Dr. M...
0 downloads 0 Views 301KB Size
Anal. Chem. 1996, 68, 734-739

On-Line Derivatization into Precolumns for the Determination of Drugs by Liquid Chromatography and Column Switching: Determination of Amphetamines in Urine Rosa Herra´ez-Herna´ndez, Pilar Campı´ns-Falco´,* and Adela Sevillano-Cabeza

Department of Analytical Chemistry, University of Vale` ncia, Dr. Moliner 50, 46100 Burjassot, Vale` ncia, Spain

A chromatographic system for the on-line derivatization of drugs using column switching is described. The system uses a 20 mm × 2.1 mm i.d. precolumn packed with a unmodified ODS stationary phase. This column is used for sample cleanup and enrichment of the analytes. Next, the trapped analytes are derivatized by injection of the derivatization reagent into the precolumn. Finally, the derivatives are transferred to the analytical column for their separation under reversed-phase conditions. The influence of several parameters such as the reaction time, the amount of derivatization reagent, or the system design has been studied using some amphetamines as model compounds and three derivatization reagents: sodium 1,2-naphthoquinone-4-sulfonate, o-phthaldialdehyde, and 9-fluorenylmethyl chloroformate. The potential of the described approach is illustrated by determining amphetamine and methamphetamine in untreated urine at ambient temperature. In the determination of traces of analytes in biological fluids by column liquid chromatography (LC), two types of problems are commonly encountered. First, sample conditioning is required before injection into the LC system. Otherwise, column clogging and rapid deterioration in column performance (caused by irreversible adsorption of the proteins) occur, resulting in a limited lifetime of the column. Second, since analyte concentration is usually low, some kind of enrichment and/or derivatization is often required to improve analyte detectability. Many strategies have been developed to overcome laborious sample cleanup, most of them based on precolumn techniques. In this way, sample cleanup and analytical separation can be performed on-line in the same system. However, on-line derivatization of anlaytes is not so widespread, probably because of the inherent difficulties of incorporating a chemical reaction into the flow scheme of a liquid chromatograph. This difficulties mainly concern the compatibility of the derivatization reagent and derivatized products with the mobile phase used for separation.1,2 Ideally, sample cleanup and derivatization should be integrated in the same process. In this respect, one of the most attractive approaches proposed is based on immobilizing detection-sensitive reagents on solid supports (mainly silica, alumina, and organic polymer based).3-7 Solid-phase reagents can be used as precol(1) Turnell, D. C.; Cooper, J. D. H. J. Chromatogr. 1989, 492, 59. (2) Krull, I. S.; Deyl, Z.; Lingeman, H. J. Chromatogr. 1994, 659, 1. (3) Gao, C.-X.; Krull, I. S. BioChromatography 1989, 4, 222.

734 Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

umn packings and integrated into the LC system. Since the choice of pore size of the solid support (or percentage of cross-linking when using a polymer network) will effect size exclusion, thus enabling derivatization of small molecules but hindering the large ones, these precolumns can be used for sample cleanup. At the same time or somewhat later, the reagent part of the solid-phase material can be made to react with the adsorbed analytes. This methodology has been successfully applied to the derivatization of a variety of drugs.8 However, the synthesis of the solid-phase reagent is often difficult, and periodic regeneration of the reagent may be required to obtain high yield, fewer side products, and reproducible results. In addition, these reagents have a limited stability in aqueous solutions, especially at elevated temperatures and pH, which are often required to obtain satisfactory reaction rates.4,7 In a previous study we demonstrated the possibility of performing derivatization with sodium 1,2-naphthoquinone-4sulfonate (NQS) into C18 solid-phase extraction cartridges.9 The described procedure was applied to the determination of amphetamine in pharmaceuticals and in urine samples. In this work, we describe a very simple procedure for the online derivatization of drugs based on the employment of precolumns packed with a conventional ODS stationary phase. The precolumn is used to purify the sample and concentrate the analytes and, next, to retain the derivatization agent and/or the derivatized analytes. Finally, the derivatives are transferred to the analytical column by means of a switching arrangement. This method offers some advantages over derivatization into solid-phase extraction cartridges. First, sample handling is minimized since untreated samples are directly injected into the system. Second, the analytes are transferred to the analytical column for separation immediately after derivatization, which avoids stability problems.10 Moreover, parameters affecting the reaction rates, especially the reaction time, can be better controlled. This is very important if the derivatization reaction exhibits slow kinetics but suitable sensitivity is achieved without quantitative conversion. The described system has been applied to the derivatization of three amphetamines, the primary amines β-phenylethylamine (4) Gao, C.-X.; Chou, T.-Y.; Krull, I. S. Anal. Chem. 1989, 61, 1538. (5) Bourque, A. J.; Krull, I. S. J. Chromatogr. 1991, 537, 123. (6) Zhou, F.-X.; Krull, I. S.; Feibush, B. J. Chromatogr. 1992, 609, 103. (7) Zhou, F.-X.; Krull, I. S.; Feibush, B. J. Chromatogr. 1993, 648, 357. (8) Krull, I. S.; Szulc, M. E.; Bourque, A. J.; Zhou, F.-X.; Yu, J.; Strong, R. J. Chromatogr. 1994, 659, 19. (9) Campı´ns-Falco´, P.; Molins-Legua, C.; Sevillano-Cabeza, A.; Kohlman, M. J. Chromatogr., submitted. (10) Farrel, B. M.; Jefferies, T. M. J. Chromatogr. 1983, 272, 111. 0003-2700/96/0368-0734$12.00/0

© 1996 American Chemical Society

Table 1. Summary of the Chromatographic Conditions Used for the Separation of Amphetamine Derivatives elution program derivtzn reagent

analytical column

NQS

Hypersil ODS C18, 5 µm, 250 mm × 4 mm i.d.

OPA

LiChrospher 100 RP 18, 5 µm 125 mm × 4 mm i.d.

FMOC

LiChrospher 100 RP 18, 5 µm 125 mm × 4 mm i.d.

a

time (min) 0 2.5 3.5 8 0 2.5 5 10 0 2.5 5 10

mobile-phase comptn (%) acetonitrile aq solventa 40 50 70 70 40 50 70 70 40 50 70 100

60 50 30 30 60 50 30 30 60 50 30 0

flow (mL/min)

detection

1

UV 280 nm

0.8

fluorescence λexc ) 345 nm λem ) 445 nm

1.5

fluorescence λexc ) 264 nm λem ) 313 nm

Propylamine hydrochloride (0.5%) in water (v/v) for the NQS and OPA methods; water for the FMOC method.

and amphetamine and the secondary amine methamphetamine, because most UV and fluorescent methods proposed for the analysis of these drugs involve their chemical transformation.11 We have evaluated the possibility of extending this technique to different derivatization reagents: NQS, o-phthaldialdehyde (OPA), and 9-fluorenylmethyl chloroformate (FMOC), which have been widely used for the off-line derivatization of primary and secondary amines (only primary amines when using OPA.)2,11-14 Since the determination of amphetamine and methamphetamine is of great interest in toxicologic and forensic fields, the potential of the described approach has been tested by analyzing these drugs in urine at their therapeutic concentrations. EXPERIMENTAL SECTION Apparatus. The chromatographic system used consisted of two quaternary pumps (Hewlett-Packard, 1050 Series, Palo Alto, CA), an automatic sample injector (Hewlett-Packard, 1050 series), and two high-pressure six-port valves (Rheodyne Model 7000). A diode array (Hewlett-Packard, 1040 series) or a fluorescence (Hewlett-Packard, 1046 series) detector linked to a data system (Hewlett-Packard HPLC Chem Station, Dos Series) was used for data acquisition and storage. Reagents. All the reagents were of analytical grade. Acetonitrile, ethyl acetate (Scharlau, Barcelona, Spain), and n-hexane (Panreac, Barcelona, Spain) were of HPLC grade. Sodium 1,2naphthoquinone-4-sulfonate, β-phenylethylamine hydrochloride, methamphetamine hydrochloride, and amphetamine sulfate were obtained from Sigma (St. Louis, MO). o-Phthaldialdehyde, and propylamine hydrochloride (Fluka, Buchs, Switzerland), 9-fluorenylmethyl chloroformate (Aldrich, Steinheim, Germany), sodium hydrogen carbonate (Probus, Badalona, Spain), sodium hydroxide, boric acid (Panreac), and 2-mercaptoethanol (Merck, Hoenbrunn, Germany) were also used. Preparation of Solutions. Amphetamine Solutions. Stock solutions (1000 µg/mL) of each compound were prepared in water. Working solutions were prepared from the stock solutions by dilution with water. These solutions were stored in the dark at 2 °C. (11) Campı´ns-Falco´, P.; Sevillano-Cabeza, A.; Molins-Legua, C. J. Liq. Chromatogr. 1994, 17, 731. (12) Kinberger, B. J. Chromatogr. 1981, 213, 166. (13) Maeder, G.; Pelletier, M.; Haerdi, W. J. Chromatogr. 1992, 593, 9. (14) Campı´ns-Falco´, P.; Bosch-Reig, F.; Sevillano-Cabeza, A.; Molins-Legua, C. Anal. Chim. Acta 1994, 287, 41.

Derivatization Reagents. Concentrations of the derivatization reagents were selected from the literature,12-15 the ratio of concentration of derivatization reagent to the concentration of analyte in the samples being in the 60-300 range. NQS solutions (0.5%, w/v) were prepared daily by dissolving the pure compound in water. OPA solutions were prepared by dissolving 0.05 g of the pure compound in 0.5 mL of methanol and 0.1 mL of 2-mercaptoethanol. This solution was further diluted to 10 mL with to 15 mM boric acid solution. The boric acid solution was adjusted to pH 10 with 10% NaOH. This reagent is stable at least for 3 days when protected from light and stored at 2 °C. FMOC solutions (20 mM) were prepared by dissolving the pure compound in acetonitrile. For the NQS and FMOC determinations, a 4% bicarbonate buffer (adjusted to pH 10 with 10% NaOH) was also used. Columns and Mobile Phases. The precolumn (20 mm × 2.1 mm i.d.) was dry-packed with a Hypersil ODS-C18 (30 µm) or a SynChropack C18 (30-70 µm) stationary phase (HewlettPackard). Purified water was used for washing the precolumn in both sample cleanup and derivatization steps. A Hypersil ODSC18, 250 mm × 4 mm i.d. (5 µm) (Hewlett-Packard) or a LiChrospher 100 PR 18, 125 mm × 4 m i.d. (5 µm) (Merck, Darmstat, Germany) column was used as an analytical column. Chromatographic conditions used for separation of the amphetamine derivatives are summarized in Table 1. In all instances, the mobile phases were prepared daily, filtered with a nylon membrane, (0.45 µm; Teknokroma, Barcelona, Spain), and degassed with helium before use. Off-Line Derivatization. For the NQS method, 0.5 mL of reagent and 0.5 mL of buffer were added to 1 mL of sample. After a given reaction time, the mixture was subjected to liquid-liquid extractions with three 2 mL aliquots of n-hexane-ethyl acetate (1:1). The organic phase was evaporated to dryness by heating at 80 °C, and the residue was reconstituted in 2 mL of acetonitrilewater (1:1). For the OPA method, 0.5 mL of sample was added to 0.25 mL of water and 0.25 mL of derivatization reagent; for derivatizations with FMOC, 0.5 mL of sample was added to 0.25 mL of carbonate buffer and 0.25 mL of FMOC reagent. For the OPA and FMOC methods, 1 mL of acetonitrile was added to the reaction mixture after the reaction time. (15) Campı´ns-Falco´, P.; Molins-Legua, C.; Herra´ez-Herna´ndez, R.; SevillanoCabeza, A. J. Chromatogr. 1995, 663, 235.

Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

735

Figure 1. Schematic representation of the system used for the online cleanup plus derivatization of amphetamines.

In all instances, derivatizations were carried out at ambient temperature and a 50 µL aliquot of the final solutions was directly injected into the analytical column. Each sample was assayed in triplicate. On-Line derivatization. The setup used for the on-line derivatization of the samples is shown in Figure 1. A 50 mL aliquot of sample was injected into the precolumn; next, the precolumn was flushed with 2 mL of water (at a flow rate of 1 mL/min) for sample cleanup. Meanwhile, the autosampler was programmed to take variable volumes of derivatization reagent and 25 µL of buffer (this latter solution only for the NQS and FMOC methods) from different vials and to mix them in the needle. The reagent mixture was injected into the precolumn immediately after sample cleanup. After injection, pump 1 was stopped for a given period of time (reaction time). For the NQS method, pump 1 was activated after the reaction time to eliminate the excess of reagent by flushing the precolumn with water (at a flow rate of 1 mL/min). Finally, V1 was rotated, so the derivatives were transferred to the analytical column by means of the eluent from pump 2. When urine samples were derivatized with NQS or OPA, the precolumn was flushed after the analytical separation with 1 mL of ethyl acetate, 2 mL of n-hexane, and again 1 mL of ethyl acetate. In that case, an additional valve (V2) was incorporated to the system, so the eluent was sent directly to waste. At the end of each assay, valves were turned to the original positions to regenerate and reequilibrate both the precolumn and the analytical column. Rotation of the valves was manually performed. Recovery Studies. The efficiency of the on-line sample cleanup and derivatization steps was calculated by comparing the peak areas obtained for the injection of 50 µL of standard solution samples (10 µg/mL) in the described system, with the values obtained for direct injections of the off-line derivatized extracts containing the equivalent amount of analytes. Each concentration was assayed in triplicate. Urine Samples. Untreated urine samples were spiked with amphetamine or methamphetamine reproducing concentrations in the 0.4-4.0 and 0.5-10.0 µg/mL intervals, respectively. Volumes of 1 mL of these samples were placed into glass injection vials, and 50 µL was injected onto the chromatographic system. RESULTS AND DISCUSSION On-Line vs Off-Line Derivatization. The chromatograms obtained for an aqueous mixture of the tested compounds after off-line and on-line derivatization are compared in Figure 2. As can be see in this figure, peak broadening introduced by the described setup is negligible. Various unknown peaks are observed in the chromatograms when analyzing blank (water) and standard amphetamine solutions 736 Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

Figure 2. UV chromatograms obtained for aqueous solutions of amphetamines after off-line (a) and on-line (b) derivatization with NQS. Fluorescence chromatograms obtained for aqueous solutions of amphetamines after off-line (c) and on-line (d) derivatization with OPA, and after off-line (e) and on-line (f) derivatization with FMOC. Concentration of each compound, 10 µg/mL; reaction time, 10 min; derivatization reagent volume used for the on-line derivatization, 25 µL. (β-PEA ) β-phenylethylamine; AMP ) amphetamine; MET ) methamphetamine). For experimental details, see text.

and derivatizations are performed off-line. These peaks are most probably due to partial hydrolysation of the derivatization reagents at pH 10, which is necessary to obtain satisfactory derivatization rates. Hydrolysis of derivatization reagents is one of the most important obstacles for derivatization in aqueous solutions, particularly at low analyte concentrations.7,8,13,16 As can be seen in Figure 2, this effect is more important when the reaction is carried out in the precolumn. In that case, additional peaks are observed, even when pure water is processed; the number and intensity of these peaks depend on the stationary phase used in the precolumn. Thus, they are most probably due to the presence of impurities in the packing material or to degradation of the packing itself at those basic pH. The efficiency of the derivatization is also dependent on the packing material. This is illustrated in Table 2, which shows the (16) Kwakman, P. J. M.; Koelewijn, H.; Kool, I.; Brinkman, U. A. Th.; De Jong, G. J. J. Chromatogr. 1990, 511, 155.

Table 2. Recoveries of Amphetamines from Aqueous Solutions after On-Line Cleanup plus Derivatization Using Hypersil and SynChropack Packingsa recovery (n ) 3) (%) derivtzn reagent NQS OPA FMOC

Hypersil packing β-PEA AMP MET

SynChropack packing β-PEA AMP MET

97 ( 7 103 ( 6 27 ( 5 17 ( 2 12 ( 3 12 ( 1 129 ( 7 76 ( 1 26 ( 3 31 ( 2 96 ( 4 98 ( 3 96.6 ( 0.2 11.3 ( 0.8 11 ( 5 21 ( 2

a Concentration of analytes, 10 µg/mL; reaction time, 10 min; derivatization reagent volume used for the on-line derivatization, 25 µL. (β-PEA ) β-phenylethylamine; AMP ) amphetamine; MET ) methamphetamine).

percentage of drugs transformed for the two packings evaluated relative to the amount of drug transformed in the off-line procedure. Best results are obtained for the Hypersil packing, which exhibits the highest loading capacity. Similar behavior was observed when derivatizations were performed off-line in ODS solid-phase extraction cartridges (data not shown). ODS packings have been successfully used for the selective retention of underivatized amphetamines.15 Therefore, the poor recoveries observed are most probably due to the low retention of the derivatization reagent and/or the derivatives formed. For the OPA method, the recovery of β-phenylethylamine in a Hypersil material is unacceptable high. This can be explained by the interference produced by an hydrolysation byproduct. Therefore, a Synchropack stationary phase should be used for the determination of this compound. Following these results, in further experiments a Hypersil packing was used, and the OPA method was only applied to the derivatization of amphetamine. The time of reaction and the derivatization reagent volume were optimized for the on-line procedure. The results of these studies are shown in Figure 3. Derivatization with NQS. Total conversion of amphetamines into their NQS derivatives usually involves high temperatures and reaction times.14 However, we observed that at a pH of ∼10, reaction rates are satisfactory at ambient temperatures.9 Stable responses are obtained for β-phenylethylamine and amphetamine when reaction times equal or higher than 10 min are used (Figure 3a). The secondary amine methamphetamine exhibits reaction rates lower than those obtained for β-phenylethylamine and amphetamine (primary amines); however, if reaction times equal or higher than 10 min are used, the sensitivity observed is satisfactory for most applications concerning to the determination of this compound.11 The excess of NQS must be eliminated before linking the precolumn to the analytical column. Otherwise severe interference with interesting compounds is observed. Indeed, analyte responses can be increased by increasing the volumes of NQS injected (Figure 3b). However, reagent volumes higher than 25 µL are not recommended, because in such a case, the flushing stage must be prolonged to eliminate the excess of reagent. For reagent volumes in the 10-25 µL interval, satisfactory chromatograms were obtained when the precolumn was flushed with 5 mL of water (at a flow rate of 1 mL/min). A reaction time of 10 min in combination with a derivatization reagent volume of 25 µL was selected as the best option for derivatization of interesting compounds. Under these conditions, the responses obtained for

a mixture of amphetamines (concentration of each compound, 10 µg/mL) were comparable to those obtained for solutions containing a single drug. This means that the three amphetamines can be simultaneously quantified. Derivatization with OPA. OPA reacts rapidly at ambient temperature with amphetamine (Figure 3c), and its derivative can be transferred to the analytical column immediately after injection of the reagent, resulting in a very rapid procedure. No significant differences in the responses of this compound were observed for volumes of reagent in the 10-75 µL interval (Figure 3d). In order to minimize peak responses corresponding to hydrolysation byproducts, a reagent volume of 15 µL was selected for urine studies. Derivatization with FMOC. The reaction with FMOC is also very rapid, and there are no significant differences in the reaction rates within the 0-10 min reaction time interval (Figure 3e). However, recoveries lower than expected were observed for FMOC volumes higher than 20 µL (see Figure 3f). This can explained by losses of the amphetamines and/or their derivatives by breakthrough, due to the presence of acetonitrile in the derivatization reagent solution. At the present concentration (20 mM), FMOC is partially soluble in water. However, in order to avoid precipitation in the needle of the autosampler during dilution with the buffer, FMOC solutions were prepared in acetonitrile. Satisfactory results are obtained by injecting 15 µL of reagent. As for the NQS method, the three amphetamines tested can be simultaneously quantified. Analysis of Urine Samples. On the basis of the above studies, the determination of amphetamine and methamphetamine in urine at their therapeutical levels was performed, using a precolumn packed with a Hypersil material. Optimal conditions for each derivatization reagent are summarized in Table 3. For the NQS and (by minor extension) for the OPA methods, the responses drastically decrease after a few injections, when urine samples instead of water solutions are processed. In our experience, the same solid-phase cartridge or precolumn can be repeatedly used for trapping amphetamines in urine prior to their off-line derivatization. Therefore, the decrease in the reaction rates can be explained by the irreversible adsorption of urinary components to the precolumn, which makes the retention of the derivatization agent and/or the retention of the derivatives more difficult. To overcome this problem, several eluents of different polarities were tested for cleaning of the precolumn after every urine injection. Best results were obtained by using acetate-nhexane. Therefore, pump 2 was programmed to deliver 1 mL of ethyl acetate, 2 mL of n-hexane, and again 1 mL of ethyl acetate ethyl (at a flow rate of 1 mL/min) after the analytical separation. In order to protect the analytical column from the desorbed impurities, an additional valve (V2 in Figure 1) was incorporated in the setup, so the washing eleuent was sent directly to waste. Under the described configuration, the reequilibration of the analytical column after processing a sample takes a few minutes, but it is carried out during cleanup and derivatization of the next one. For the FMOC method, cleaning of the precolumn with 1 mL of acetonitrile provided reproducible responses, so the final procedure can be simplified. In Table 4 are shown relevant analytical data of the described procedures. Although rotation of the valves was manually performed, the reproducibility achieved is comparable to that of most other LC methods involving on-line derivatizations.4 HowAnalytical Chemistry, Vol. 68, No. 5, March 1, 1996

737

Figure 3. Effects of the reaction time and derivatization reagent volume on peak areas: (a, c, e) effect of the reaction time for NQS, OPA, and FMOC methods, respectively; (b, d, f) effect of the reagent volume for NQS, OPA, and FMOC methods, respectively. (β-PEA ) β-phenylethylamine; AMP ) amphetamine; MET ) methamphetamine). For experimental details, see text.

ever, the on-line to off-line responses obtained for methamphetamine in the NQS method are unacceptable low. This is in agreement with the fact that slow kinetic reactions of secondary amines into solid supports provide lower rates compared with the analogous solution derivatization.4,9 In this instance, the recovery can be improved by extending the reaction time or by performing the derivatization at higher temperatures. Utility of the System. The main advantage over previously published procedures is the possibility of performing sample cleanup, derivatization, and elimination of the excess of derivatizing agent (if necessary) within the same system and using 738

Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

conventional ODS packings. In the described assay, the limiting step is the chromatographic separation. However, separation of a sample can be performed during sample cleanup plus derivatization of the next one. Although untreated samples are directly injected, both the precolumn and the analytical column are effectively equilibrated after an analysis. In all instances, at least 40 samples can be processed without replacing the precolumn. Best sensitivity is achieved with FMOC. In addition, the procedure can be simplified when this latter reagent is used, and amphetamine and methamphetamine can be quantified in a single run, the total analysis time being 12 min. Therefore, FMOC would

Table 3. Conditions of the On-Line Cleanup plus Derivatization of Amphetamines in Urine derivtzn reagent

Sample cleanup

Vderivtzn reagent (µL)

Vbuffer (µL)

reaction time (min)

cleaning of the precolumn after derivtzn

cleaning of the precolumn after anal. sepn

NQS

2 mL of water

25

25

10

5 mL of water

OPA

2 mL of water

15

FMOC

2 mL of water

15

1 mL of ethyl acetate 2 mL of n-hexane 1 mL of ethyl acetate 1 mL of ethyl acetate 2 mL of n-hexane 1 mL of ethyl acetate 1 mL of acetonitrile

0 25

0

Table 4. Analytical Data for the On-Line Determination of Amphetamine and Methamphetamine in Urine linearity derivtzn reagent NQS

analyte

recovery (n ) 3) (%)

AMP

101 ( 3

MET

31 ( 3

OPA

AMP

77 ( 5

FMOC

AMP

98 ( 3

MET

86.4 ( 0.2

a

precision (%)

y ) a + bx

Sxy

n

intraday (n ) 6)

interday (n ) 15)

limit of detection (ng/mL)

total anal. time (min)

a ) -14.5 ( 3.6 b ) 73.7 ( 1.5 a ) -0.7 ( 3.0 b ) 19.4 ( 0.5 a ) -6.7 ( 4.6 b ) 79.8 ( 2.0 a ) 3.2 ( 1.1 b ) 15.2 ( 0.5 a ) 15.1 ( 1.3 b ) 15.0 ( 0.2

6.55

12

3

4

25

35

6.10

11

5

7

50

8.46

11

6

8

10

20

2.07

12

7

8

1

12

2.48

12

4

8

1

AMP ) amphetamine; MET ) methamphetamine. b Determined at half of highest concentration in tested range.

In principle, the described methodology could be used for the formation of diastereomeric derivatives, for example, by using a homochiral thiol in the OPA method or a chiral active chloroformate. Coupling of a chiral column to the derivatization precolumn could be also possible. Both possibilities are currently under investigation in our laboratory.

Figure 4. Fluorescence chromatograms obtained for blank (a) and spiked with amphetamine and methamphetamine (b) urine samples. Concentration of analytes: amphetamine, 2.5 µg/mL; methamphetamine, 5.0 µg/mL. (AMP ) amphetamine; MET ) methamphetamine). For experimental details, see text.

be the reagent of choice for routine determinations of amphetamine and methamphetamine in urine. As an example, typical chromatograms obtained for blank and spiked urine samples derivatized with FMOC are shown in Figure 4. Under normal conditions, methamphetamine is partially demethylated to amphetamine, the methamphetamine-to-amphetamine ratio being generally between 3 and 10 in urine samples.17-20 About 30-40% of amphetamine is excreted unchanged in urine.17 Since amphetamine and methamphetamine are the predominant forms in urine, the present method can be considered suitable for most applications concerning these compounds. On the other hand, enantiomeric analysis of amines is of interest since the pharmacological activities of the isomers differ. (17) Beckett, A. H.; Rowland, M. Nature 1964, 204, 1203. (18) Shinichi, S.; Takako, I.; Tetsukichi, N. J. Chromatogr. 1983, 267, 381. (19) Tsuchihashi, H.; Nakajima, K.; Nishikawa, M.; Shiomi, K.; Takahashi, S. J. Chromatogr. 1989, 467, 227. (20) Binder, S. R.; Regalia, M.; Biaggi-McEachern, M.; Mazhar, M. J. Chromatogr. 1989, 473, 325.

CONCLUSIONS This study illustrates the potential of conventional ODS precolumns for the on-line sample cleanup and derivatization of solutes in the analysis of biological samples prior to LC analysis. The described approach can be used with different derivatization reagents, but operating conditions should be optimized according to the kinetics of each reaction. Efficiency of the derivatization mainly depends on a proper selection of the precolumn stationary phase, the major limitation being the presence of hydrolysation byproducts. Compared with off-line derivatization methods, the described setup facilitates the analysis and provides better reproducibility, since no manipulation of the samples is involved. In contrast to solid-phase reagents, an excess of derivatization reagent is necessary; however, consumption of reagents can be reduced by using higher reaction times or higher tempratures. The described approach is fully automatable, only requiring electrically controlled switching valves. ACKNOWLEDGMENT The authors are grateful to the ClCyT for financial support received for the realization of Project SAF 95-0586.

Received for review May 25, 1995. Accepted December 4, 1995.X AC9505076 X

Abstract published in Advance ACS Abstracts, January 15, 1996.

Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

739