LC Separation

Now, it will be shown later in this book, the following approximation is valid for most LC separations ... 

The solvent used was 5 %v/v ethyl acetate in n-hexane at a flow rate of 0.5 ml/min. Each solute was dissolved in the mobile phase at a concentration appropriate to its extinction coefficient. Each determination was carried out in triplicate and, if any individual measurement differed by more than 3% from either or both replicates, then further replicate samples were injected. All peaks were symmetrical (i.e., the asymmetry ratio was less than 1.1). The efficiency of each solute peak was taken as four times the square of the ratio of the retention time in seconds to the peak width in seconds measured at 0.6065 of the peak height. The diffusivities obtained for 69 different solutes are included with other physical and chromatographic properties in table 1. The diffusivity values are included here as they can be useful in many theoretical studies and there is a dearth of such data available in the literature (particularly for the type of solutes and solvents commonly used in LC separations). 

In the previous two chapters, equations were developed to provide the optimum column dimensions and operating conditions to achieve a particular separation in the minimum time for both packed columns and open tubular columns. In practice, the vast majority of LC separations are carried out on packed columns, whereas in GC, the greater part of all analyses are performed with open tubular columns. As a consequence, in this chapter the equations for packed LC columns will first be examined and the factors that have the major impact of each optimized parameter discussed. Subsequently open tubular GC columns will be considered in a similar manner. 

The interactions between solute and the pha.ses are exactly the same as those present in LC separations, namely, dispersive, polar and ionic interactions. At one extreme, the plate coating might be silica gel, which would offer predominately polar and induced polar interactions with the solute and, con.sequently, the separation order would follow that of the solute polarity. To confine the polar selectivity to the stationai y phase, the mobile phase might be -hexane which would offer only dispersive interactions to the solute. The separation of aromatic hydrocarbons by induced polar selectivity could be achieved, for example, with such a system. 

Multidimensional LC separation has been defined as a teehnique whieh is mainly eharaeterized by two distinet eriteria, as follows (1). The first eriterion for a multidimensional system is that sample eomponents must be displaeed by two or more separation teehniques involving orthogonal separation meehanisms (2), while the seeond eriterion is that eomponents that are separated by any single separation dimension must not be reeombined in any further separation dimension. 

Trace enrichment and sample clean-up are probably the most important applications of LC-LC separation methods. The interest in these LC-LC techniques has increased rapidly in recent years, particularly in environmental analysis and clean-up and/or trace analysis in biological matrices which demands accurate determinations of compounds at very low concentration levels present in complex matrices (12-24). Both sample clean-up and trace enrichment are frequently employed in the same LC-LC scheme of course, if the concentration of the analytes of interest are Sufficient for detection then only the removal of interfering substances by sample clean-up is necessary for analysis. 

Figure 9.5 The generic setup for two-dimensional liquid chromatography-capillary zone electrophoresis as used by Jorgenson s group. The LC separation was performed in hours, while the CZE runs were on a time scale of seconds.

In 1990, Bushey and Jorgenson developed the first automated system that eoupled HPLC with CZE (19). This orthogonal separation teehnique used differenees in hydrophobieity in the first dimension and moleeular eharge in the seeond dimension for the analysis of peptide mixtures. The LC separation employed a gradient at 20 p.L/min volumetrie flow rate, with a eolumn of 1.0 mm ID. The effluent from the ehromatographie eolumn filled a 10 p.L loop on a eomputer-eontrolled, six-port miero valve. At fixed intervals, the loop material was flushed over the anode end of the CZE eapillary, allowing eleetrokinetie injeetions to be made into the seeond dimension from the first. 

Gianesello et al. (120) described the determination of the bronchodilator brox-aterol in plasma by on-line LC-GC. After deproteination and extraction, the LC separation was carried out by using a mixture of -pentane and diethyl ether (55 45 (vol/vol) as mobile phase. A small cut of the LC chromatogram (shown in Figure 11.9(a)) was introduced at 85 °C into the GC via so-called concurrent solvent evaporation. Figure 11.9(b) demonstrates that a detection limit of about 0.03 ng/ml was obtained. A fully automated LC-GC instrument was described by Munari and Grob (121) and its applicability was demonstrated by the determination of heroin metabo-... 

Pigure 13.16 shows the LC separation of this extract and the GC/PID chromatogram of the oxy-PAC fraction. The different oxy-PACs can be quantified without interfering compounds with a non-selective detector such as an PID. 

Figure 13.16 LC separation of urban air particulate exrtact (a), along with the GC/FID cliro-matogram (b) of an oxy-PAC fraction (transfeired via a loop-type interface). Reprinted from Environmental Science and Technology, 29, A. C. Lewis et al., On-line coupled LC-GC-ITD/MS for the identification of alkylated, oxygenated and nirtated polycyclic aromatic compounds in urban air particulate exti acts , pp. 1977-1981, copyright 1995, with permission from the American Chemical Society.

For the LC separation, 10 ml of sample was injeeted through a loop. The LC flow-rate was 1000 p.1 min and at the end of the enriehment proeess this was redueed to 100 p.1 min . The mobile phase eomposition was optimized to eliminate matrix polar eompounds and elute the phthalates in a small fraetion. 

Chirobiotie Handbook, Guide to using maeroeyelie glyeopeptide bonded phases for ehiral LC separations, Advaneed Separation Teehnologies Ine. (ASTEC), 2 Ed. Whippany, New York... 

SFC has been performed with either open capillary columns similar to those used in GC or packed columns transferred from LC, and the instrumentation requirements differ for these two approaches [12]. This chapter will focus on the use of packed column technology because of its dominance in the area of pharmaceutical compound separations. Current commercial instrumentation for packed column SFC utilizes many of the same components as traditional LC instruments, including pumps, injection valves, and detectors. In fact, most modem packed column SFC instm-ments can also be used to perform LC separations, and many of the same stationary phases can be used in both LC and SFC [9]. 

The difference in resolution between LC and SFC can be significant enough to turn a marginal LC separation into a viable chromatographic method in SFC. String-ham et al. reported that packed column SFC yielded satisfactory chiral analysis... 

The vast majority of modem liquid chromatography systems involve the use of silica gel or a derivative of silica gel, such as a bonded phase, as a stationary phase. Thus, it would appear that most LC separations are carried out by liquid-solid chromatography. Owing to the adsorption of solvent on the surface of both silica and bonded phases, however, the physical chemical characteristics of the separation are more akin to a liquid-liquid distribution system than that of a liquid-solid system. As a consequence, although most modern stationary phases are in fact solids, solute distribution is usually treated theoretically as a liquid-liquid system. 

The separation was carried out on a bonded phase LC-PCN column carrying cyanopropylmethyl moieties on the surface. Thus, in contrast to the extraction process, which appears to be based on ionic interactions with the weak ion exchange material, the LC separation appears to be based on a mixture of interactions. There will be dispersive interactions of the drugs with the hydrocarbon chains of the bonded moiety and also weakly polar interactions with the cyano group. It is seen that the extraction procedures are very efficient and all the tricyclic antidepressant drugs are eluted discretely. 

In the analysis of complex PAH mixtures obtained from environmental samples, reversed-phase LC-FL typically provides reliable results for only 8-12 major PAHs (Wise et al. 1993a). To increase the number of PAHs determined by LC-FL, a multidimensional LC procedure is used to isolate and enrich specific isomeric PAHs that could not be measured easily in the total PAH fraction because of interferences, low concentrations, and/or low fluorescence sensitivity or selectivity. This multi-dimensional procedure, which has been described previously (Wise et al. 1977 May and Wise 1984 Wise et al. 1993a, 1993b), consists of a normal-phase LC separation of the PAHs based on the number of aromatic carbon atoms in the PAH, thereby providing fractions containing only isomeric PAHs and their alkyl-substituted isomers. These isomeric fractions are then analyzed by reversed-phase LC-FL to separate and quantify the various isomers. 

Once the analyte has been separated from the matrix in LC, the best approach to the detection of the molecule must be determined. This section will discuss the detection techniques of ultraviolet/visible (UVA IS), fluorescence (FL), and electrochemical (EC) detection, with MS being addressed separately in Section 4.2. When deciding on the most appropriate detector for an LC separation, the appropriate chemical data on the analyte should be collected by using a spectrophotometer, fluorimeter, and potentiometer. 

In the case of the low abundance of some compounds, there are difficulties with signal overlap. To overcome these difficulties, there have been developments involving NMR hyphenation with techniques such as HPLC and mass spectrometry. In LC/NMR methods of analysis, NMR is used as the detector following LC separation and this technique is capable of detecting low concentrations in the nanogram range. This technique has been reported for the detection and identification of flavanoids in fruit juices and the characterization of sugars in wine [17]. 

As the vast majority of LC separations are carried out by means of gradient-elution RPLC, solvent-elimination RPLC-FUR interfaces suitable for the elimination of aqueous eluent contents are of considerable use. RPLC-FTTR systems based on TSP, PB and ultrasonic nebulisa-tion can handle relatively high flows of aqueous eluents (0.3-1 ml.min 1) and allow the use of conventional-size LC. However, due to diffuse spray characteristics and poor efficiency of analyte transfer to the substrate, their applicability is limited, with moderate (100 ng) to unfavourable (l-10pg) identification limits (mass injected). Better results (0.5-5 ng injected) are obtained with pneumatic and electrospray nebulisers, especially in combination with ZnSe substrates. Pneumatic LC-FI1R interfaces combine rapid solvent elimination with a relatively narrow spray. This allows deposition of analytes in narrow spots, so that FUR transmission microscopy achieves mass sensitivities in the low- or even sub-ng range. The flow-rates that can be handled directly by these systems are 2-50 pLmin-1, which means that micro- or narrow-bore LC (i.d. 0.2-1 mm) has to be applied. 

Figure 7.27 Constructed Gram-Schmidt chromatogram of a temperature-programmed packed-capillary LC separation of 4.8 ig Irgafos P-EPQ dissolved in DMF temperature programme 50°C for 8min, 4°Cmin-1 up to 140°C. Legend 1, mono-P-EPQ 2, Irgafos 168 3, 4,3 -P-EPQ 4, oxidised 4,4 -P-EPQ 5, 4,4 -P-EPQ. After Bruheim et al. [511]. From I. Bruheim et al., Journal of High Resolution Chromatography, 23, 525-530 (2000). Wiley-VCH, 2000. Reproduced by permission of Wiley-VCH.

Whereas the components of (known) test mixtures can be attributed on the basis of APCI+/, spectra, it is quite doubtful that this is equally feasible for unknown (real-life) extracts. Data acquisition conditions of LC-APCI-MS need to be optimised for existing universal LC separation protocols. User-specific databases of reference spectra need to be generated, and knowledge about the fragmentation rules of APCI-MS needs to be developed for the identification of unknown additives in polymers. Method development requires validation by comparison with established analytical tools. Extension to a quantitative method appears feasible. Despite the current wide spread of LC-API-MS equipment, relatively few industrial users, such as ICI, Sumitomo, Ford, GE, Solvay and DSM, appear to be somehow committed to this technique for (routine) polymer/additive analysis. 

In HPLC-TLC coupling, the crucial aspect is the maintenance of the chromatographic integrity during the deposition process. The chromatogram is preserved after LC separation, and is available for further separation and/or investigation. LC-TLC coupling increases the separation efficiency, and allows detection modes which are incompatible with LC (e.g. spectroscopic techniques... 

SFE-GC and SFC-GC may replace normal-phase LC-GC. SFC-SFC may be adopted for a wide range of applications, particularly for solutes that are not amenable to GC and as a replacement for some coupled LC-LC separations. However, no applications to additives in polymers can be mentioned. 

---