Introduction to Modern Liquid Chromatography, Third Edition part 82 pot

766 SAMPLE PREPARATION 16.5.1 Theory The Nernst distribution law states that any species will distribute between two immiscible solvents so that the ratio of the concentrations remains constant. K D = C o C aq (16.1) where K D is the distribution constant, C o is the concentration of the analyte in the organic phase, and C aq is the concentration of the analyte in the aqueous phase. A more useful expression is the fraction E of analyte extracted, given by E = C o V o C o V o + C aq V aq = K D ψ 1 + K D ψ (16.2) where V o is the volume of organic phase, V aq the volume of aqueous phase, and ψ is the phase ratio V o /V aq . Many LLE procedures are carried out in separatory funnels, and typically use tens or hundreds of milliliters of each phase. For one-step extractions K D (or 1/K D ) must be large (e.g., > 10) for the quantitative recovery of analyte in one of the two phases, since the phase-ratio ψ must be maintained within a practical range of values, such as 0.1 <ψ<10 (see Eq. 16.2). In most separatory-funnel LLE procedures, quantitative recoveries ( > 99%) require two or more extractions. For successive multiple extractions, with pooling of the analyte phases from each extraction, E = 1 −  1 1 + K D ψ  n (16.3) where n is the number of extractions. For example, if K D = 5 for an analyte, and the volumes of the two phases are equal (ψ = 1), three extractions (n = 3) would be required for > 99% recovery of the analyte. Several approaches can be used to increase the value of K D : • organic solvent can be changed to increase K D. • K D can be increased if the analyte is ionic or ionizable by suppressing its ionization so as to make it more soluble in the organic phase • the analyte can be extracted into the organic phase by ion pairing, provided that the analyte is ionized and an ion-pair reagent is added to the organic phase • ‘‘salting out’’ can be used to decrease an analyte’s concentration in the aqueous phase, by addition of an inert, neutral salt (e.g., sodium sulfate) to the aqueous phase. 16.5.2 Practice Table 16.5 provides examples of typical extraction solvents, as well as some unsuit- able (water-miscible) extraction solvents. Apart from miscibility considerations, the main selection criterion is the polarity P  of the solvent (Tables 2.3, I.4) in relation to that of the analyte. Maximum K D occurs when the polarity of the extraction solvent matches that of the analyte. For example, the extraction of a polar analyte 16.5 LIQUID –LIQUID EXTRACTION 767 Table 16.5 Extraction Solvents for Liquid—Extraction Aqueous Solvents Water-Immiscible Water-Miscible Organic Organic Solvents Solvents (Unsuitable for LLE) Pure water Acidic solution Basic solution High salt (salting-out effect) Complexing agents (ion pairing, chelating, chiral, etc.) Combination of two or more above Aliphatic hydrocarbons (hexane, isooctane, petroleum ether, etc.) Diethyl ether or other ethers Methylene chloride Chloroform Ethyl acetate and other esters Aliphatic ketones (C 6 and above) Aliphatic alcohols (C 6 and above) Toluene, xylenes (UV absorbance!) Combination of two or more above Alcohols (low molecular weight) Ketones (low molecular weight) Aldehydes (low molecular weight) Carboxylic acids (low molecular weight) Acetonitrile Dimethyl sulfoxide Dioxane Note: Any solvent from the ‘‘aqueous solvents’’ column can be matched with any solvent of the ‘‘water-immiscible organic solvents’’ column; water-miscible organic solvents should not be used with aqueous solvents to perform LLE. from an aqueous sample matrix would be best accomplished with a more polar (large P  ) organic solvent. An optimum-polarity organic solvent can be conveniently selected by blending two solvents of different polarity (e.g., hexane [P  = 0.1] and chloroform [P  = 4.1]), and measuring K D vs. the composition of the organic phase [11]. A solvent mixture that gives the largest value of K D is then used for the LLE procedure. Further changes in K D can be achieved, with improvement in the separation of analytes from interferences, by varying organic-solvent selectivity in addition to polarity. Solvents from different regions of the solvent-selectivity triangle (Fig 2.9) are expected to provide differences in selectivity; see also the discussion of [12]. In solvent extraction, ionizable organic analytes often can be transferred into either phase, depending on the selected conditions. For example, consider the extraction of an organic acid from an aqueous solution. If the aqueous phase is buffered at least 1.5 pH units above its pK a value, the analyte will be ionized and prefer the aqueous phase; less-polar interferences will be extracted into the organic phase. If the pH of the aqueous solution is lowered (pK a ), so that the analyte is no longer ionized, the analyte will be extracted into the organic phase, leaving more-polar interferences in the aqueous phase. Successive extractions at high pH followed by low pH are able to separate an acid from both more- and less-polar interferences. Equilibria involving pH are discussed further in Section 7.2. Note that the principles of acid-base extraction as a function of pH are the same for LLE and RPC. If the analyte K D is unfavorable, additional extractions may be required for improved recovery (Eq. 16.3). For the case of an organic-soluble analyte, a fresh portion of organic solvent is added to the aqueous phase in order to extract additional solute; all extracts are then combined. For a given volume of final extraction-solvent, multiple extractions are generally more efficient in removing a solute quantitatively, 768 SAMPLE PREPARATION as opposed to the use of a single extraction volume. Back-extraction can be used to further reduce interferences. For example, consider the example above of an organic-acid analyte. If the analyte is first extracted at low pH into the organic phase, polar interferences (e.g., hydrophilic neutrals, protonated bases) are left behind in the aqueous phase. If a fresh portion of high-pH aqueous buffer is used for the back-extraction of the organic phase, the ionized organic acid is transferred back into the aqueous phase, leaving less-polar interferences in the organic phase (the latter procedure is similar to successive extractions at high pH followed by low pH described above). Thus a two-step extraction with change of pH can allow the removal of both basic and neutral interferences, whereas a one-step extraction can eliminate one or the other of these interferences, but not both. If K D is not much greater than 1, or the required volume of sample is large, it may be impractical to carry out multiple extractions for quantitative recovery of the analyte—too many extractions are required, and the volume of total extract is too large (Eq. 16.3). If extraction is slow, a long time may also be required for the equilibrium to be established. In these cases continuous liquid–liquid extraction can be used, where fresh solvent is continually recycled through the aqueous sample. Continuous extractors that use heavier-than-water and lighter-than-water solvents have been described [13]. These extraction devices can run for extended periods (12–24 hr), and quantitative extractions ( > 99% recovery) can be achieved, even for less-favorable values of K D . For more efficient LLE, a countercurrent distribution apparatus can provide a thousand or more equilibration steps (but with more time and effort). This allows the recovery of analytes having K D values near unity; countercurrent distribution also provides a better separation of analytes from interferences. Small-scale labo- ratory units are commercially available. For further information on these devices, see [14]. In some cases LLE can enhance analyte concentration in the extract fraction relative to its concentration in the initial sample. According to Equation (16.2), by choosing a smaller volume of organic solvent, the analyte concentration can be increased by the volumetric ratio of organic-to-aqueous phases (assuming near-complete extraction into the organic phase or large K D ). For example, assume 100 mL of aqueous sample, 10 mL of organic solvent, and a very large K D (e.g., K D > 1000). The concentration of the analyte in the organic phase will then increase by a factor of 10. For large ratios of aqueous-to-organic, a slight solubility of the organic solvent in the aqueous phase can reduce the volume of the recovered organic solvent significantly; this problem can be avoided by presaturating the aqueous solvent with organic solvent. Note that when the solvent ratio V 0 /V aq is small, the physical manipulation of two phases (including recovery of the organic phase) becomes more difficult. 16.5.3 Problems Some practical problems associated with LLE include: • emulsion formation • analytes strongly adsorbed to particulates 16.5 LIQUID –LIQUID EXTRACTION 769 • analytes bound to high-molecular-weight compounds (e.g., protein-drug interactions) • mutual solubility of the two phases 16.5.3.1 Emulsion Formation Emulsions are a problem that can occur with some samples (e.g., fatty matrices) under certain solvent conditions. If emulsions do not ‘‘break,’’ with a sharp boundary between the aqueous and organic phases, analyte recovery can be adversely affected. Emulsions often can be broken by: • addition of salt to the aqueous phase • heating or cooling the extraction vessel • filtration through a glass wool plug • filtration through phase-separation filter paper • addition of a small amount of different organic solvent • centrifugation 16.5.3.2 Analyte Adsorption If particulates are present in a sample, adsorption onto these particulates can result in a low recovery of the analyte. In such cases, washing the particulates after filtration with a stronger solvent often will recover the adsorbed analyte; this extract should be combined with the analyte phase from LLE. A ‘‘stronger’’ solvent for recovering adsorbed analyte may involve a change in pH, increase in ionic strength, or the use of a more polar organic solvent. 16.5.3.3 Solute Binding Compounds that normally are recovered quantitatively in LLE may bind to proteins when plasma samples are processed, resulting in low recovery. Protein-binding is especially troublesome when measuring drugs and drug metabolites in physiological fluids. Techniques for disrupting protein binding in plasma samples include: • addition of detergent • addition of organic solvent, chaotropic agents, or strong acid • dilution with water • displacement with a more strongly binding compound 16.5.3.4 Mutual Phase-Solubility ‘‘Immiscible’’ solvents have a small, but finite, mutual solubility, and the dissolved solvent can change the relative volumes of the two phases. Therefore it is a good practice to saturate each phase with the other, so that the volume of phase containing the analyte can be known accurately, allowing an optimum determination of analyte recovery. For values of the solubility of a solvent in water (or of water in the solvent), see [15]. 770 SAMPLE PREPARATION 16.5.4 Special Approaches to Liquid–Liquid Extraction 16.5.4.1 Microextraction Extractions in this form of LLE are carried out with organic-aqueous ratios of 0.001 to 0.01. Analyte recovery may suffer, compared to conventional LLE, but the analyte concentration in the organic phase is significantly increased and solvent usage is greatly reduced. Such extractions are conveniently carried out in a volumetric flask. The organic extraction solvent is chosen to have a density less than that of water, so that the small volume of organic solvent accumulates in the neck of the flask for easy removal. For quantitative analysis, internal standards should be used and extractions of calibration standards carried out. Some modern autosamplers are capable of performing microextractions automatically on small volumes of aqueous samples in 2-mL vials provided that the position of the pickup needle is adjustable [16]. 16.5.4.2 Single-Drop Microextraction (SDME) This technique uses a 1- or 2-μL droplet of immiscible organic solvent, held at the end of a syringe needle, to extract and concentrate analytes from an aqueous (immiscible) sample [17]. Analytes diffuse into the droplet, resulting in a considerable increase in analyte concentration. When equilibrium is achieved, the microdrop is retracted into the syringe and then injected into an HPLC column or diluted in a microvial to achieve RPC mobile-phase compatibility. An example of the use of SDME is the RPC analysis of hypercins in plasma and urine [18]. Sometimes the use of a hollow-fiber membrane filled with a small volume of organic solvent is more useful in containing the solvent droplet. This procedure is often referred to as liquid-phase microextraction (LPME). 16.5.4.3 Solid-Supported Liquid–Liquid Extraction (SLE) SLE replaces the separatory funnel in LLE with a small column that contains an inert support such as diatomaceous earth. An aqueous sample is first applied to the column, so as to coat the support with sample. A buffered immiscible solvent is then passed through the column, with extraction of any hydrophobic analytes. Samples that have been extracted in this way include diluted plasma, urine, and milk. The solvent moves through the column by means of gravity flow or a gentle vacuum. Because there is no vigorous shaking of the sample and extraction solvent, as in conventional LLE, there is no possibility of emulsion formation. The packed tubes are disposable, and the entire process is amenable to automation. Packed 96-well plates with several hundred milligrams of packing per well are suitable for the extraction of 150 to 200 μL of aqueous sample. Examples of commercial products that perform SLE are Varian’s Hydromax (Palo Alto, CA), Biotage’s Isolute HM-N (Charlottesville, VA), and Merck’s Extrelut (Darmstadt, Germany). 16.5.4.4 Immobilized Liquid Extraction (ILE) ILE involves extraction of hydrophobic analytes from an aqueous sample into a polymeric film comprising a phase similar to the bonded liquid phases used in capillary GC. The polymeric film can be applied to the cap of a vial, the inner walls of a 96-well plate, or inside a micropipette tip. The sample is first exposed to the film for extraction of analytes, followed by an aqueous wash to remove polar interferences, and a final wash with organic solvent to recover the analyte. 16.6 SOLID-PHASE EXTRACTION (SPE) 771 The sample can be directly injected into the chromatograph or evaporated and redissolved in a more HPLC-compatible solvent. Devices for ILE are supplied by ILE Inc. (Ferndale, CA) and other suppliers. ILE can be automated and compares favorably to SPE [19]. 16.6 SOLID-PHASE EXTRACTION (SPE) Solid-phase extraction is the most important technique used in sample pretreatment for HPLC. SPE can be used in similar fashion as LLE, but whereas LLE usually is a one-stage separation process, SPE is a chromatographic procedure that resembles HPLC and has a number of potential advantages compared to LLE: • more complete extraction of the analyte • more efficient separation of interferences from analytes • reduced organic solvent consumption • easier collection of the total analyte fraction • more convenient manual procedures • removal of particulates • more easily automated Because SPE is a more efficient separation process than LLE, it is easier to obtain a higher recovery of the analyte. LLE procedures that require several successive extractions to recover 99+% of the analyte often can be replaced by one-step SPE methods. With SPE it is also possible to obtain a more complete removal of interferences from the analyte fraction. Reversed-phase SPE proce- dures are the most popular because only small amounts of organic solvent are required while maintaining a higher concentration of analyte. There is no need for phase separation (as in LLE), so the total analyte fraction is easily collected in SPE, eliminating errors associated with variable or inaccurately measured extract volumes. In SPE there is no chance of emulsion formation. Finally, larger partic- ulates are trapped by the SPE cartridge and do not pass through into the analyte fraction. Some disadvantages of SPE compared with LLE include: • potential variability of SPE packings • irreversible adsorption of some analytes on SPE cartridges • more-complex method development is required (up to 4 steps involved, Fig. 16.4) The solvents used in LLE are usually pure and well defined, so that LLE separations are quite reproducible. Although the cartridges used in the past for SPE sometimes varied from lot to lot, initiatives to improve production quality have led to major improvements in cartridge reproducibility. The surface area of an LLE device (e.g., separatory funnel) is quite small (and less active) compared to an SPE cartridge (with its high-surface-area packing), so irreversible binding of analyte (with lower recoveries) is less likely with LLE vs. SPE. 772 SAMPLE PREPARATION 16.6.1 SPE and HPLC Compared In its simplest form, SPE employs a small, plastic, disposable column or cartridge, often the barrel of a medical syringe packed with 0.1 to 1.0 g sorbent. The sorbent is commonly a reversed-phase material (e.g., C 18 -silica) that resembles RPC in its sepa- ration characteristics. In the following discussion we will assume reversed-phase SPE (RP-SPE) unless noted otherwise. Although silica-gel-based bonded-phase packings were introduced first, polymeric sorbents have become available in recent years and have been gaining in popularity. Compared to silica-based SPE packings, polymeric packings have several advantages: (1) higher surface area (thus higher capacity), (2) better wetability, (3) tolerance to partial drying after the conditioning step, without affecting recovery and reproducibility, (4) an absence of silanols (less chance of irreversible adsorption of highly basic compounds), and (5) a wide pH range (more flexibility in adjusting conditions). In its most popular configuration, the SPE packing, is held in a syringe barrel by frits, similar to an HPLC column (Fig. 16.2a).Theparticlesize(e.g.,40-μm average) typically is larger than that in HPLC (1.5–5-μm). Because shorter bed lengths, larger particles, and less well-packed beds are used, SPE cartridges are much less efficient than an HPLC column (N < 100). For cost reasons, irregularly shaped, type-A-silica packings (rather than spherical, type-B particles; see Section 5.2.2.2) usually are used in SPE. Recently spherical silicas for SPE have come on the market but have not impacted the sale of the most popular products. Polymeric sorbents, which generally are spherical, are more expensive than silica-based packings. Some SPE disks, however, use the more expensive spherical SPE packings with particle diameters in the 7-μm range. Overall, the principles of separation, selection of conditions, and method development are similar for both SPE and HPLC, except that SPE uses a series of isocratic steps during retention and elution of the analyte. One major difference between SPE and HPLC is that the SPE cartridge is usually used once and discarded, since potential interferences can remain on the cartridge, whereas HPLC columns are used many times. 16.6.2 Uses of SPE SPE is used for six main purposes in sample preparation: • removal of interferences and ‘‘column killers’’ • concentration or trace enrichment of the analyte •desalting • solvent exchange • in-situ derivatization • sample storage and transport 16.6.2.1 Interference Removal Interferences that overlap analyte peaks in the HPLC separation complicate method development and can adversely affect assay results. In some cases, especially for complex samples (e.g., natural products, protein digests), a large number of inter- ferences in the original sample can make it almost impossible to separate these from one or more analyte peaks by means of a single HPLC separation. SPE can be used 16.6 SOLID-PHASE EXTRACTION (SPE) 773 pipette tip sorbent coating cartridge body frit frit sorbent bed Luer tip holder holder sorbent dis k pre-filter (optional) reservoir (a) (b) (c) Figure 16.2 Different means for carrying out solid-phase extraction (SPE). (a) Disposable cartridge (syringe-barrel format); (b)disk;(c) micropipette tip (MPT). to reduce or eliminate those interferences prior to HPLC. Some samples contain components, such as hydrophobic substances (e.g., fats, oils, greases), proteins, polymeric materials, or particulates that can plug or deactivate the HPLC column. These ‘‘column killers’’ often can be removed by RP-SPE. 16.6.2.2 Analyte Enrichment SPE can be used to increase the concentration of a trace component. If an SPE cartridge can be selected so that k  1 for the analyte, a relatively large volume of sample (e.g., several mL) can be applied before the analyte saturates the cartridge and begins to elute from the cartridge. An increase in analyte concentration (trace enrichment) can then be achieved, provided that the cartridge is eluted with a small volume of strong solvent (k > 1). An example of trace enrichment is the use of SPE to concentrate sub-ppb of polynuclear aromatic hydrocarbons [20] or pesticides [21] from environmental water samples using a RP-SPE cartridge. A strong solvent (e.g., ACN or MeOH) elutes these analytes from the cartridge in a small volume, which saves on evaporation time. The sample can then be redissolved in a solvent compatible with the subsequent HPLC separation. Alternatively, the eluted sample can be diluted directly into a suitable injection solvent. 774 SAMPLE PREPARATION 16.6.2.3 Desalting RP-SPE can be used to desalt samples, especially prior to ion-exchange chromatog- raphy (IEC) where a low-ionic strength sample is desirable. Conditions of pH and %-organic are selected to retain the analyte initially so that the inorganic salts can be washed from the cartridge with water. The analyte can then be eluted (salt free) with organic solvent [22]. 16.6.2.4 Other Applications The remaining applications of SPE—solvent exchange, in-situ derivatization, and sample storage/transport—are either seldom used or are less relevant for HPLC; for details, see [21, 23, 24]. 16.6.3 SPE Devices Several SPE configurations are used (Fig. 16.2): • cartridge •disk • pipette tip • 96-well plate • coatedfiberorstirbar 16.6.3.1 Cartridges The most popular SPE configuration is the cartridge. A typical SPE disposable cartridge (syringe-barrel format) is depicted in Figure 16.2a. The syringe barrel is usually medical-grade polypropylene that is fitted with a Luer tip, so that a needle can be affixed to direct the effluent to a small container or vial. The frits that hold the particle bed in the cartridge are of made of PTFE, polypropylene, or stainless steel with a porosity of 10 to 20 μm, and thus offer little flow resistance. SPE cartridges may vary in design to fit an automated instrument or robotic system. SPE cartridges are relatively inexpensive, and they are discarded after a single use to avoid sample cross-contamination. To accommodate a wide range of SPE applications, cartridges are available with packing weights of 35 mg to 2 g, as well as with reservoir volumes (the volume above the packing in the cartridge) of 0.5 to 10 mL. For very large samples, ‘‘mega’’ cartridges contain up to 10 g of packing with a 60-mL reservoir. Cartridges with a larger amount of packing should be used for dirty samples that may overload a low-capacity cartridge. However, cartridges containing 100 mg of packing or less are preferred for relatively clean liquid samples where cartridge capacity is not an issue, as well as for small sample volumes. Because of the higher surface areas of polymeric SPE packings, less packing is needed than for silica-based particles (2- to 60-mg). In most cases it is desirable to collect the analyte in the smallest possible volume, which means that the SPE cartridge generally should also be as small as possible. 16.6 SOLID-PHASE EXTRACTION (SPE) 775 16.6.3.2 Disks The second most popular SPE configuration is the disk (Fig. 16.2b). SPE disks combine the advantages of membranes (see below) and solid-phase extraction. In their appearance, the disks closely resemble membrane filters: they are flat, usually ≤1-mm thick, and 4 to 96 mm in diameter. The physical construction of the SPE disks differs from membrane filters. SPE disks can be acquired in any of the following configurations: • flexible or expanded PTFE networks filled with silica-based or resin packings • rigid fiberglass disks with embedded packing material • packing-impregnated polyvinylchloride • derivatized membranes Filled Disks. The packing in these disks generally comprise 60-90% of the total membrane weight. Some disks are sold individually and must be installed in a reusable filter holder. Others are sold preloaded in disposable holders or cartridges with Luer fittings for easy connection to syringes. SPE disks and cartridges differ mainly in their length/diameter ratios (L/d): disks have L/d < 1 and cartridges have L/d > 1. Compared to SPE cartridges, this characteristic of the disk enables higher flow rates and faster extraction. ‘‘Dirty’’ water or water containing particulates, such as wastewater, can plug the porous disks, just as in the case of cartridges. So a prefilter is used prior to the SPE treatment. Some disk products come with a built-in prefilter. Channeling, which can cause uneven flow through poorly packed cartridges, is not a problem with disks. Due to the thinness of the disk (typically 0.5–2 mm), however, compounds with low k-values tend to have lower breakthrough volumes than for SPE cartridges. SPE disks are especially useful for environmental applications, such as the analysis of trace organics in surface water, which often require a large sample volume to obtain the necessary sensitivity. The EPA has approved SPE technology as an alternative for large-volume LLE methods [25] in the preparation of water samples for HPLC analysis. Examples of approved methods include procedures for phenols [26], haloacetic acids in drinking water [27], and pesticides and polychlorinated biphenyls (PCBs) [28]. Embedded Disks. Low-bed-mass, rigid fiberglass disks with 1.5 to 30 mg of embedded packing material are useful for pretreating small clinical samples (e.g., plasma or serum; [29]). Their reduced sorbent mass and small volume reduces solvent consumption (and any related sample contamination by solvent impurities). An advantage of this type of disk is an absence of frits that are a possible further source of contamination. Other Disks. Packing-impregnated polyvinylchloride (PVC) and derivatized membranes are used very little in SPE applications and are not discussed here. 16.6.3.3 Other SPE Formats The move toward miniaturization in analytical chemistry has prompted the devel- opment of new formats for SPE: . LLE include: • emulsion formation • analytes strongly adsorbed to particulates 16.5 LIQUID LIQUID EXTRACTION 769 • analytes bound to high-molecular-weight compounds (e.g., protein-drug interactions) •. Approaches to Liquid Liquid Extraction 16.5.4.1 Microextraction Extractions in this form of LLE are carried out with organic-aqueous ratios of 0.001 to 0.01. Analyte recovery may suffer, compared to. droplet. This procedure is often referred to as liquid- phase microextraction (LPME). 16.5.4.3 Solid-Supported Liquid Liquid Extraction (SLE) SLE replaces the separatory funnel in LLE with a small column