Anal. Chem. 2004, 76, 2525-2530
Comprehensive Two-Dimensional Normal-Phase (Adsorption)-Reversed-Phase Liquid Chromatography Paola Dugo,† Olinda Favoino,‡ Rosario Luppino,‡ Giovanni Dugo,‡ and Luigi Mondello*,‡,§
Dipartimento di Chimica Organica e Biologica, Facolta` di Scienze, Universita` di Messina, salita Sperone 31, 98166 Messina, Italy, Dipartimento Farmaco-chimico, Facolta` di Farmacia, Universita` di Messina, viale Annunziata, 98168 Messina, Italy, and Universita` “Campus Biomedico” di Roma, Via Emilio Longoni, 83-00155 Roma, Italy
A comprehensive two-dimensional HPLC system has been developed. It is based on the use of a microbore silica column operated in normal-phase (adsorption) mode (NP) in the first dimension and a monolithic type C18 column operated in reversed-phase (RP) mode in the second dimension. The interface was a 10-port, 2-position valve equipped with two storage loops. The first column was operated at a flow rate of 20 µL/min in isocratic mode, while the monolithic column flow rate was 4 mL/ min and was operated in gradient mode. The sample loops had a volume of 20 µL each, and the analysis time in the second dimension was 1 min. In this way, every fraction from the first dimension was transferred on-line to the second dimension switching the automated valve every minute. A photodiode array detector has been used after the secondary column. The use of normal- and reversedphase mode in the two dimensions can be helpful in the separation of complex mixtures of a natural origin that contain uncharged molecules of comparable dimension, different in polarity and hydrophobicity. The use of a microbore column in the first dimension permits the injection of a small volume in the secondary column, making the transfer of incompatible solvents from the first to the second dimension possible. Since the mobile phase in the NP separation is always stronger than the mobile phase at the head of the secondary column operated in RP mode, the initial eluent strength is important in order to obtain an effective focusing of the sample. The use of a monolithic type column in the second dimension permits the performance of very fast analysis operating at higher flow rates without loss of resolution, due to a higher permeability and increased mass-transfer properties in comparison to conventional particulate columns. Due to the brief reconditioning time necessary for monolithic columns, repetitive gradients can be carried out, extending the field of application to mixtures that contain components with different polarities. The utility of the system has been demonstrated in the analysis of the oxygen heterocyclic fraction of cold-pressed lemon oil, made up of coumarins and psoralens. These components may contain hydroxyl, methoxyl, isopentenyl, isopentenyloxyl, and geranyloxyl groups and oxygen-containing modi10.1021/ac0352981 CCC: $27.50 Published on Web 03/27/2004
© 2004 American Chemical Society
fication of the terpenoid side-chain groups, such as epoxides or vicinal diol groups. The relative location of the components in the 2D plane varied in relation to their chemical structure and allowed positive peak identification. The UV spectra recorded with the photodiode array detector supplied additional information that was used for the characterization of the studied sample. Complex samples require analytical methods characterized by an extremely high resolving power in order to provide thorough analysis of the sample components. Multidimensional (MD) chromatography is an approach capable of providing greater resolution. The most common use of MD separation is the pretreatment of a complex matrix in an off-line mode. The offline approach is very easy but presents several disadvantages: it is time-consuming, operationally intensive, and difficult to automate and to reproduce. Moreover, sample contamination or formation of artifacts can occur. On-line MD chromatography offers the advantages of ease of automation and greater reproducibility in a shorter analysis time. On the other hand, on-line systems are more difficult to operate and need specific interfaces. In liquid chromatography (LC), on-line MDLC is achieved through the coupling of a second column by means of a highpressure switching valve, which traps a defined volume of collected sample, usually in a loop, and directs it to the second column (“heart-cutting”). Heart-cutting is suitable for the characterization of specific parts of a sample but does not permit a complete (“comprehensive”) two-dimensional separation of the entire sample. In 1978, Erni and Frei1 were probably the first to introduce comprehensive two-dimensional liquid chromatography, soon followed by Bushey and Jorgenson.2 A comprehensive two-dimensional system should possess the following features: (1) all components in a sample mixture are subjected to two separations in which their displacement depends on different factors; (2) any two components separated in the first * To whom correspondence should be addressed. Phone: +39-090-6766536. E-mail:
[email protected]. † Dipartimento di Chimica Organica e Biologica, Universita` di Messina. ‡ Dipartimento Farmaco-chimico, Universita` di Messina. § Universita` “Campus Biomedico” di Roma. (1) Erni, F.; Frei, R. W. J. Chromatogr. 1978, 149, 561. (2) Bushey, M.; Jorgenson, J. W. Anal. Chem. 1990, 62, 161.
Analytical Chemistry, Vol. 76, No. 9, May 1, 2004 2525
dimension must remain separated when they are subjected to the second dimension; (3) the elution profiles from both dimensions are preserved.3 If compared to single-dimension chromatography, the separation power of comprehensive two-dimensional chromatography is greatly increased. Giddings4-6 has shown theoretically that the peak capacity can be greatly enhanced by coupling HPLC columns that separate according to different (orthogonal) retention mechanisms. In MDLC, the use of a microbore LC column for the firstdimension separation, is a good solution for different reasons:7,8 (1) the small column i.d. helps to ensure a minimum of dilution and provides flow rates that are compatible with the sample volume for the secondary column; (2) there is no need for a preconcentration step at the head of the secondary column, and solvent incompatibility between different separation modes is avoided. The possible disadvantage could be the lower sample capacity of microbore LC columns. However, in MDLC, a sensitivity enhancement can be obtained if the formation of compressed solute bands at the head of the secondary column is achieved during the transfer from the first to the second dimension. In the case of a large injection volume, in MDLC, if the LC microcolumn is used as a highly efficient preseparation step, a limited decrease in efficiency due to a large injection volume can be tolerated. The combination of a microbore column in the first dimension and a conventional column in the second dimension, connected by a multiport switching valve equipped with two sample loops, has been used for the comprehensive two-dimensional chromatography of proteins by Jorgenson and co-workers2,9 and synthetic polymers by Schoenmakers.10 When a conventional column is used as a first-dimensional column, a different LC/LC interface has been developed, using two fast secondary columns in parallel rather than storage loops. With this kind of interface, the mobile phase from the first column should have a very low strength so that analytes can be trapped at the head of the secondary columns during the loading step. Such a system has been used by Opiteck et al.,11 Wagner et al.,12 and Unger et al.,13 for the analysis of peptides and proteins. All the methods cited used SEC or IEX in one of the two dimensions, while reversed-phase (RP)-LC is normally used in the other one. Recently, a simple, automated 2D-LC system equipped with an electronically controlled, 12-port valve has been used for the analysis of aromatic amines and non-amines, operating both dimensions under comparable reversed-phase conditions.14 In this (3) Schoenmakers, P.; Marriott, P.; Beens, J. LC-GC Eur. 2003, 16, 335. (4) Giddings, J. C. Anal. Chem. 1984, 56, 1258A. (5) Giddings, J. C. HRC & CC, High Resolut. Chromatogr. Chromatogr. Commun. 1987, 10, 319. (6) Giddings, J. C. In Multidimensional Chromatography: Techniques and Applications; Cortes, H. J., Ed.; Marcel Dekker: New York, 1990; pp 1-27. (7) Cortes, H. J. J. Chromatogr. 1992, 626, 3. (8) Ko ¨hne, A. P.; Welsch, T. J. Chromatogr., A 1999, 845, 463. (9) Opiteck, G. J.; Lewis, K. C.; Jorgenson, J. W.; Anderegg, R. J. Anal. Chem. 1997, 69, 1518. (10) Van der Horst, A.; Schoenmakers, P. J. J. Chromatogr., A 2003, 1000, 693. (11) Opiteck, G. J.; Jorgenson, J. W.; Anderegg, R. J. Anal. Chem. 1997, 69, 2283. (12) Wagner, K.; Racaityte, K.; Unger, K. K.; Miliotis, T.; Edholm, L. E.; Bischoff, R.; Marko-Varga, G. J. Chromatogr., A 2000, 893, 293. (13) Unger, K. K.; Racaityte, K.; Wagner, K.; Miliotis, T.; Edholm, L. E.; Bischoff, R.; Marko-Varga, G. J. High Resolut. Cromatogr. 2000, 23, 259.
2526 Analytical Chemistry, Vol. 76, No. 9, May 1, 2004
Figure 1. Structures of oxygen heterocyclic components present in the nonvolatile residue of cold-pressed citrus oils.16 In coumarins and psoralens, the substituents may be present in the numbered positions. Isomers and homologues are possible.
case, the orthogonal separation was achieved by tuning the operating parameters, such as mobile-phase strength, temperature, and buffer strength, in conjunction with column selectivity. The interfacing of normal-phase (NP) and reversed-phase systems is particularly difficult, due to the mobile-phase immiscibility. The combination of normal (silica) and reversed (C18) phase HPLC in a comprehensive 2D-LC system was used for the analysis of alcohol ethoxylates,15 but the normal phase was run using aqueous solvents, so the mobile phases used in the two dimensions were miscible, resulting in the easy injection of the entire first-dimension effluent onto the second-dimension column. The system developed in this study uses a microbore silica column operated in normal-phase mode in the first dimension, and a monolithic type column operated in reversed-phase mode in the second dimension. The interface was a 10-port, 2-position valve equipped with two storage loops. The use of normal- and reversed-phase mode in the two dimensions can be helpful in the separation of a complex mixture of natural origin, that contains uncharged molecules of comparable dimension, different in polarity and hydrophobicity. In the present study, the comprehensive 2D-LC system was evaluated by analyzing the oxygen heterocyclic components of a cold-pressed lemon oil. Oxygen heterocyclic components (coumarins, psoralens, polymethoxylated components) represent the main part of the nonvolatile fraction of cold-pressed citrus oils. Their structures and substituents are reported in Figure 1. These components have an important role in the characterization of coldpressed citrus oils, since their qualitative and quantitative composition of the fraction is characteristic of each oil.16 The analysis of these components is usually carried out by HPLC, using both normal- and reversed-phase modes. However, with both methods, some coelutions may occur. The use of a two-dimensional system can achieve the complete resolution of all the components of the fraction. EXPERIMENTAL SECTION Reagents. Most of the coumarins and psoralens used in this study were isolated in the laboratory from a sample of lemon oil by classical column chromatography and semipreparative HPLC.17 (14) Venkatramani, C. J.; Zelechonok, Y. Anal. Chem. 2003, 75, 3484. (15) Murphy, R. E.; Schure, M. R.; Foley, J. P. Anal. Chem. 1998, 70, 4353. (16) Dugo, P.; McHale, D. In Citrus; Dugo, G., Di Giacomo, A., Eds.; Taylor and Francis: London, 2002; pp 355-390.
Figure 2. Schematic of 2D-LC system.
All the solvents were HPLC grade and were purchased from Carlo Erba (Milan, Italy). Stock solution of 5-geranyloxy-7-methoxycoumarin, bergamottin, citropten, and bergapten were prepared in ethyl acetate and used for the optimization of the method. A genuine Sicilian coldpressed lemon essential oil was diluted 1:10 v/v in a mixture of n-hexane/ethyl acetate, 80:20. A 2-µL sample of the standard solution or of the solution of lemon oil was injected into the HPLC. Instrumentation and Chromatographic Conditions. The instrumentation for comprehensive 2D-LC is shown in Figure 2. The key component of the system is an electronically controlled 2-position, 10-port valve that enables continuous, alternate sampling of the primary column eluent onto the secondary column through two equivalent sample loops. In one of the two positions, the eluent from the primary column fills one of the two 20-µL sample loops. At the same time, the content of the other 20-µL loop from the previous cycle is sampled onto the secondary column. When the valve is switched to the second position, the effluent from the primary column flows through the other loop, while the loop filled during the previous cycle is sampled onto the secondary column. All instruments used in this study are commercially available. The first dimension isocratic LC system consisted of a Shimadzu LC-10AD vp solvent delivery unit, a Rheodyne twoposition, six-port injection valve model 7725i equipped with a 2-µL loop, and a Shimadzu SPD-10A vp UV detector provided with a microcell U-Z view (LC Packings, Amsterdam, The Netherlands). A primary SupelcoSil LC-SI column (Supelco, Milan, Italy) (300 × 1-mm i.d., 5-µm particle diameter) was used. A flow rate of 20 µL/min of n-hexane/acetonitrile (75:25) was used. Pressure was 14 bar. UV absorbance was measured at a wavelength of 315 nm; the acquisition frequency was of 1.666 67 Hz. The second-dimension gradient LC system consisted of two Shimadzu LC-10AD vp solvent delivery units connected in parallel to a gradient mixer, a Shimadzu SPD-M10A vp photodiode array detector, and a Shimadzu SCL 10A vp controller. The secondary column was a Merck Chromolith Flash (Merck KGaA, Darmstadt, Germany) (25 × 4.6 mm i.d.) equipped with a Merck Chromolith guard column (5 × 4.6 mm i.d.). The mobile phase used was water and acetonitrile. A gradient run was from 0 to 12 s, 50% acetonitrile, (17) Dugo, P.; Mondello, L.; Cogliandro, E.; Cavazza, A.; Dugo, G. Flavour Fragrance J. 1998, 13, 329.
increasing to 100% acetonitrile in 36 s, and followed by a 12-s reequilibration step to the initial conditions. Flow rate was 4 mL/ min. Pressure was 170 bar. The UV spectra of eluting peaks were monitored in the range 240-360 nm, and the chromatogram was acquired at 315 nm; sampling frequency was 12.5 Hz; time constant was 0.32 s. Total run time was 55 min. Data acquisition of both UV and photodiode array detectors was by Shimadzu Class vp 5.0 software. Column switching was performed using an electronically controlled 10-port, 2-position Supelpro valve purchased from Supelco and controlled by a method editor software; the valve was operated with two injection loops of 20 µL. The valve was switched every 60 s by the Class vp programmed external events. By using the export function of the Class vp software, the ASCII data were converted into a matrix with rows corresponding to a 60-s duration and data columns covering all successive seconddimension 60-s chromatograms using the 2D GC converter 2.0 (Chromatography Concepts, Doncaster, Australia). Contour representation of the 2D chromatograms was through Transform version 3.3 software (Fortner Software, VA). RESULTS AND DISCUSSION Figure 3 shows the comprehensive two-dimensional LC/LC chromatogram of the oxygen heterocyclic components of a lemon essential oil. The flow rate of 4 mL/min through the secondary column resulted in 1-min RP-LC chromatograms. The application of a 20 µL/min flow rate through column 1 and the use of 20-µL loops in the 10-port valve enabled 1 injection/min onto the secondary column. In this way, all the effluent from the first column was analyzed in the second dimension. Separation occurring in the first dimension (NP-LC) is shown along the x axis, with a time scale of 50 min, while separations occurring in the second dimension (RP-LC) are shown along the y axis, with a time scale of 0-1 min. Flow rates, sampling time, and eluent strength, in both the first and second dimensions, were optimized to obtain the analysis of the whole sample in both dimensions. A series of analytical conditions were tried in order to obtain a sufficient number of samplings of each first-dimension peak for a complete bidimensional characterization of the sample in the shortest amount of time, in accordance to Murphy et al.18 This was mainly important for peaks 2 and 5, whose effective separation is achieved in the first dimension. Under our conditions, two samplings for peak 2 and three for peak 5 were performed. These values are below the three to four samplings per peaks reported to obtain a high-fidelity separation.3,18 To increase the number of samplings per peak, a reduction of the velocity in the firstdimension separation was performed, but a decrease in firstdimension analysis speed was unsuitable for the later-eluting components. However, the separation between peaks 2 and 5 in the 2D-LC plane is clearly visible (Figure 3). The main limitation in the development of a MDLC system, where the two dimensions are operated in NP and RP mode, is mobile-phase incompatibility. It has been demonstrated that the introduction of large volumes of an incompatible solvent yields broadened and distorted peaks.7 The use of a microbore column in the first dimension permits the injection of a small volume onto (18) Murphy, R. E.; Schure, M. R.; Foley, J. P. Anal. Chem. 1998, 70, 1585.
Analytical Chemistry, Vol. 76, No. 9, May 1, 2004
2527
Figure 3. Comprehensive 2D normal-phase (adsorption)-reversed-phase LC separation of the oxygen heterocyclic fraction of a lemon oil sample.
the secondary column, making the transfer of incompatible solvents from the first to the second dimension possible without peak shape deterioration or resolution losses. When normal-phase LC is used as first dimension, the mobile phase sampled onto the secondary column operated in RP-LC mode is always stronger than the mobile phase at the head of the secondary column. In this case, to obtain an effective focusing of the sample in the secondary column, the initial eluent strength was maintained low. In addition, the transfer of the loop content from the first to the second dimension was very fast. In fact, it took only 0.3 s. A repetitive gradient in the second dimension was then necessary to elute all the components of the fraction within the 1-min analysis. A gradient program in the second dimension is necessary when the differences in polarity and hydrophobicity of the components present in the matrix are very large; hence, suitable isocratic conditions for their separation in a very short time are very difficult to find. However, the use of a monolithic column in the fast dimension permits the performance of successive cycles with a very brief equilibration time because of the higher permeability in contrast to conventional particulate columns. Moreover, the use of the monolithic column allows the use of high flow rates without loss of resolution, due to the better mass-transfer properties of a monolithic skeleton over particlepacked columns,19 thus reducing the analysis time. Figure 4 shows four consecutive 1-min second-dimension runs with fractions
containing peaks 6-9. As can be seen, the 2D difference in retention time for corresponding peaks, in successive secondary chromatograms, is exactly 1 min. This, is a confirmation of the complete reconditioning of the monolithic column in the 12-s step between two consecutive gradient cycles of 48 s each. For example, peak 6, identified as citropten, elutes at 24.171, 25.163, 26.171, and 27.168 min in the four runs. This means that the average citropten retention time in the four 2D-LC chromatograms is 0.168 min, with a standard deviation of (0.004. As we can see from Figure 3, 11 components have been separated with the comprehensive 2D-LC analysis. Identification of these components was achieved by comparison of retention times in the single dimensions, and the relative location in the 2D plane, with those of standard components, when available. Identification was also supported by the observation of the UV spectra relative to each component and by using literature data. In fact, coumarins and psoralens present UV spectra, with characteristic absorption maximums in relation to the substituted positions.20 On the basis of these data, peaks were identified as two coumarins and eight psoralens, as specified in Table 1. The remaining component showed a UV spectrum in discordance with those presented by coumarins and psoralens and was not identified. The different intensities of the spots in the 2D-LC chromatogram are due to strong differences in concentration of the components in the lemon oil sample.16
(19) Lubda, D.; Cabrera, K.; Kraas, W.; Schafer, C.; Cunningham, D. LC-GC Eur. 2001, 14, 730.
(20) Murray, R. D. H.; Me´ndez, J.; Brown, S. A. The Natural Coumarins; John Wiley and Sons: Chichester, U.K., 1982.
2528
Analytical Chemistry, Vol. 76, No. 9, May 1, 2004
Figure 4. Enlargment of the raw second-dimension chromatogram from 24 to 28 min. Elution of peaks 6-9 in four consecutive 1-min cycles.
Table 1. Oxygen Heterocyclic Compounds Separated in the Comprehensive 2D-LC Analysis of the Lemon Oil compound identification 1 2 3 4 5 6 7 8 9 10 11
unknown 5-geranyloxypsoralen 5-isopentenyloxypsoralen 5-geranyloxy-8-methoxypsoralen 5-geranyloxy-7-methoxycoumarin 5,7-dimethoxycoumarin 5-methoxy-8-isopentenyloxypsoralen 8-geranyloxypsoralen 5-isopentenyloxy-8-epoxyisopentyloxypsoralen 5-epoxyisopentyloxypsoralen 5-methoxy-8-(2,3-epoxyisopentyloxy)psoralen
trivial name bergamottin isoimperatorin citropten phellopterin oxypeucedanin byakangelicol
As can be seen from the position of the peaks in the 2D space, NP and RP separation modes are partially correlated techniques. Oxygen heterocyclic components of lemon oil represent an example of a real sample where components can be separated on the basis of differences in aromatic moieties and position of the aliphatic chain in normal-phase mode and then separated on the basis of differences in the aliphatic moieties in the RP mode. The oxygen heterocyclic components identified in lemon oil present different substituents that modify substantially the hydrophobicity and polarity of the molecules. Citropten is the only one with two methoxyl groups, four components present a geranyloxy chain, and the other five contain at least one isopentenyloxy chain or a modification of this chain with an epoxide group. If we observe the elution order in the two dimensions, some considerations can be made, on the basis of the structures of the identified components and on the basis of the retention factors in normal- (adsorption) and reversed-phase LC.21,22 Under normal-phase (adsorption) conditions (1) psoralens elute before the corresponding coumarins. This can be seen by ob(21) Snyder, L. R.; Kirkland, J. J. Introduction to Modern Liquid Chromatography; John Wiley and Sons: New York, 1979. (22) Yost, R. W.; Ettre, L. S.; Conlon, R. D. Practical Liquid Chromatography; Perkin-Elmer: Norwalk, CT, 1980.
serving the elution order of bergamottin (peak 2) and 5-geranyloxy-7-methoxycoumarin (peak 5) and (2) psoralens substituted in position 5 elute before those substituted in position 8. In fact, bergamottin (peak 2) elutes before 8-geranyloxypsoralen (peak 8). In accordance with the mechanism of retention in normal-phase adsorption chromatography, isomers are usually well separated, as can be observed for bergamottin and 8-geranyloxypsoralen, while homologues are not very well separated, since hydrocarbon substituents contribute little to sample retention. Bergamottin (peak 2) and isoimperatorin (peak 3), which differ in the length of the alkyl chain, elute very closely. Polar functional groups are strongly attracted to the adsorbent surface, so compounds with substituents of different polarity are readily separated. As an example, isoimperatorin (peak 3) and oxypeucedanin (peak 10), or phellopterin (peak 7) and byakangelicol (peak 11), which differ respectively for the presence of an epoxy substituent in the side chain, are well-resolved. Under reversed-phase conditions, a combination of hydrophobicity and polarity dominates retention. Components with a long geranyloxy side chain are much more retained than the other components and elute in the same zone (see peaks 2, 4, 5, and 8). In this case, the position occupied by the geranyloxy group discriminates less than in normal-phase LC, and components such as bergamottin and 8-geranyloxypsoralen elute very closely. Components with a shorter alkyl chain are eluted earlier and were found in an other zone of the 2D plane (see peaks 3, 7, and 9). More polar components, containing an epoxyisopentenyloxy chain, elute first (see peaks 10 and 11). The different retention mechanisms of the two columns permit the resolution of some coelutions obtained in the first-dimension separation. For example, 5-geranyloxy-7-methoxycoumarin coelutes or it is not well separated from isoimperatorin and another small peak here identified as 5-geranyloxy-8-methoxypsoralen. These three peaks are well-resolved on the C18 column used in the second dimension. Information for the identification of these Analytical Chemistry, Vol. 76, No. 9, May 1, 2004
2529
of peak 4 corresponds to a psoralen substituted at positions C-5 and C-8, and the spectrum of peak 5 matches that of a 5,7disubstituted coumarin. Peak identification is in accordance with all this combined information. By coupling LC × LC to PDA, a third dimension was added that greatly supported the identification of the analyzed components. When dealing with natural matrixes, difficulties in peak identification are mainly due to the lack of standards and the difficulty in isolating sufficient amounts of them from the matrix. The use of a spectroscopic technique in combination with the separation system can represent a rapid method to obtain on-line structural information, very useful for peak identification or confirmation, especially when no commercial standards are available. The PDA detector, while being less useful than the MS detector, both for the amount of information that can be obtained and for their universality, presents the advantage of a lower cost and a very wide diffusion in analytical laboratories.
Figure 5. UV spectra of (A) isoimperatorin, (B) 5-geranyloxy-8methoxypsoralen, and (C) 5-geranyloxy-7-methoxycoumarin.
peaks can be obtained from the specific position occupied in the 2D-LC plane. Peak 4, for example, falls within a specific 2D zone pinpointed by a RP and NP coordinate. The former shows the presence of a geranyloxy chain, while the latter shows that the chain is in position 5. The UV spectra recorded with the PDA detector can supply additional information. Figure 5 illustrates the UV spectra of components 3-5, showing characteristic differences. In accordance with data from the literature and from data obtained from the analysis of standard components, the UV spectrum of peak 3 is typical for a psoralen substituted at C-5, the spectrum
2530
Analytical Chemistry, Vol. 76, No. 9, May 1, 2004
CONCLUSIONS The comprehensive 2D-LC system presented in this work is simple and easy to use. It is completely automated and was obtained using only commercially available equipment. When the 2D-LC system is not in use, the two LC systems can be used independently with minimal change in plumbing. The application of this system to the analysis of a real sample has demonstrated that the quantity of information obtained is higher if compared to single-column chromatography. In fact, the 2D space chromatogram derived from LC × LC analysis has a great potential for identification, because the contour plot positions give characteristic patterns for specific classes of compounds. Moreover, some couples of peaks that are critical pairs in one separation mode can be easily separated using the 2D approach. The system is very versatile and can be used for the analysis of samples covering a wide range of polarity. The use of a PDA as detector represents a third dimension that can greatly help in the unequivocal identification of the analytes. ACKNOWLEDGMENT Support was obtained from the COM-CHROM research training network HPRN-CT-2001-00180 (COM-CHROM).
Received for review November 3, 2003. Accepted February 22, 2004. AC0352981
