12. Basic Principles of Chromatography

12.2.1 Batch Extraction

In batch extraction the solute is extracted from one solvent by shaking it with a second immiscible solvent. The solute partitions, or distributes, itself between the two phases, and, when equilibrium has been reached, the partition coefficient, K, is a constant:

(12.1)

After shaking, the phases are allowed to separate, and the layer containing the desired constituent is removed, for example, in a separatory funnel. In batch extraction, it is often difficult to obtain a clean separation of phases, owing to emulsion formation. Moreover, partition implies that a single batch extraction is usually incomplete.

12.2.2 Continuous Extraction

Continuous extraction requires special apparatus, but is more efficient than batch separation. One example is the use of a Soxhlet extractor (Chap. 17, Sect. 17.​2.​5) for extracting fat from solids using organic solvents. Solvent is recycled so that the solid is repeatedly extracted with fresh solvent. Other types of equipment have been designed for the continuous extraction of substances from liquids and/or solids, and different extractors are used for solvents that are heavier or lighter than water.

12.2.3 Countercurrent Extraction

Countercurrent distribution refers to a serial extraction process. It separates two or more solutes with different partition coefficients from each other by a series of partitions between two immiscible liquid phases. Liquid-liquid partition chromatography (Sect. 12.4.2), also known as countercurrent chromatography, is a direct extension of countercurrent extraction. Years ago the countercurrent extraction was done with a “Craig apparatus” consisting of a series of glass tubes designed such that the lighter liquid phase (mobile phase) was transferred from one tube to the next, while the heavy phase (stationary phase) remained in the first tube [5]. The liquid-liquid extractions took place simultaneously in all tubes of the apparatus, which was usually driven electromechanically. Each tube in which a complete equilibration took place corresponded to one theoretical plate of the chromatographic column (refer to Sect. 12.5.1.2.1). The greater the difference in the partition coefficients of various substances, the better was the separation. A much larger number of tubes were required to separate mixtures of substances with close partition coefficients, which made this type of countercurrent extraction very tedious. Modern liquid-liquid partition chromatography (Sect. 12.4.2) is much more efficient and convenient.

12.3.1 Historical Perspective

Modern chromatography originated in the late nineteenth and early twentieth centuries from independent work by David T. Day, a distinguished American geologist and mining engineer, and Mikhail Tsvet, a Russian botanist. Day developed procedures for fractionating crude petroleum by passing it through Fuller’s earth, and Tsvet used a column packed with chalk to separate leaf pigments into colored bands. Because Tsvet recognized and correctly interpreted the chromatographic processes and named the phenomenon chromatography, he is generally credited with its discovery.

After languishing in oblivion for years, chromatography began to evolve in the 1940s due to the development of column partition chromatography by Martin and Synge and the invention of paper chromatography. The first publication on GC appeared in 1952. By the late 1960s, GC, because of its importance to the petroleum industry, had developed into a sophisticated instrumental technique, which was the first instrumental chromatography to be available commercially. Since early applications in the mid-1960s, HPLC, profiting from the theoretical and instrumental advances of GC, has extended the area of LC into an equally sophisticated and useful method. Supercritical fluid chromatography (SFC), first demonstrated in 1962, has been gaining popularity in food analysis [7]. Efficient chromatographic techniques, including automated systems, continue to be developed for utilization in the characterization and quality control of food ingredients and products [4, 7–13].

12.3.2 General Terminology

Chromatography is a general term applied to a wide variety of separation techniques based on the partitioning or distribution of a sample (solute) between a moving or mobile phase and a fixed or stationary phase. Chromatography may be viewed as a series of equilibrations between the mobile and stationary phase. The relative interaction of a solute with these two phases is described by the partition (K) or distribution (D) coefficient (ratio of concentration of solute in stationary phase to concentration of solute in mobile phase). The mobile phase may be either a gas (for GC) or liquid (for LC) or a supercritical fluid (for SFC). The stationary phase may be a liquid or a solid. The field of chromatography can be subdivided according to the various techniques applied (Fig. 12.1) or according to the physicochemical principles involved in the separation. Table 12.1 summarizes some of the chromatographic procedures or methods that have been developed on the basis of different mobile-stationary phase combinations. Inasmuch as the nature of interactions between solute molecules and the mobile or stationary phases differ, these methods have the ability to separate different kinds of molecules. (The reader is urged to review Table 12.1 again after having read this chapter.)

table 12.1

Characteristics of different chromatographic methods

Method	Mobile/phase	Stationary phase	Retention varies with
Gas-liquid chromatography	Gas	Liquid	Molecular size/polarity
Gas-solid chromatography	Gas	Solid	Molecular size/polarity
Supercritical fluid chromatography	Supercritical fluid	Solid	Molecular size/polarity
Reversed-phase chromatography	Polar liquid	Nonpolar liquid or solid	Molecular size/polarity
Normal-phase chromatography	Less polar liquid	More polar liquid or solid	Molecular size/polarity
Ion-exchange chromatography	Polar liquid-Ionic solid	Ionic solid	Molecular charge
Size-exclusion chromatography	Liquid	Solid	Molecular size
Hydrophobic interaction chromatography	Polar liquid	Nonpolar liquid or solid	Molecular size/polarity
Affinity chromatography	Water	Binding sites	Specific structure

Reprinted from Heftmann [1], p. A21, with kind permission from Elsevier Science-NL, Sara Burgerhartstraat 25, 1055 KV Amsterdam, The Netherlands

12.3.3 Gas Chromatography

Gas chromatography is a column chromatography technique, in which the mobile phase is gas and the stationary phase is mostly an immobilized liquid on an inert solid support in either a packed or capillary-type column. GC is used to separate thermally stable volatile components of a mixture. Gas chromatography, specifically gas-liquid chromatography, involves vaporizing a sample and injecting it onto the head of the column. Under a controlled temperature gradient, the sample is transported through the column by the flow of an inert, gaseous mobile phase. Volatiles are then separated based on several properties, including boiling point, molecular size, and polarity. Physiochemical principles of separation are covered in Sect. 12.4. However, details of the chromatographic theory of separation as it applies specifically to GC, as well as detection and instrumentation of GC, are detailed in Chap. 14.

12.3.4 Liquid Chromatography

There are several liquid chromatography techniques applied in food analysis, namely, planar chromatography (both paper and thin-layer chromatography) and column liquid chromatography, all of which involve a liquid mobile phase and either a solid or a liquid stationary phase. However, the physical form of the stationary phase is quite different in each case. Separation of the solutes is based on their physicochemical interactions with the two phases, which is discussed in Sect. 12.4.

12.3.4.1 Planar Chromatography

12.3.4.1.1 Paper Chromatography

Paper chromatography was introduced in 1944, and today it is mostly used as a teaching tool. In paper chromatography the stationary phase (water) and the mobile phase (organic solvent) are both liquid (partition chromatography, see Sect. 12.4.2), with paper (usually cellulose) serving as a support for the liquid stationary phase. The support also may be impregnated with a nonpolar organic solvent and developed with water or other polar solvents (reversed-phase paper chromatography). The dissolved sample is applied as a small spot or streak about 1.5 cm from the edge of a strip or square of the paper, which is then allowed to dry. The dry strip is suspended in a closed container in which the atmosphere is saturated with the developing solvent (mobile phase) and the paper chromatogram is developed. The end closer to the sample is placed in contact with the solvent, which then travels up or down the paper by capillary action (depending on whether ascending or descending development is used), separating the sample components in the process. When the solvent front has traveled the length of the paper, the strip is removed from the developing chamber, and the separated zones are detected by an appropriate method.

In the case of complex sample mixtures, a two-dimensional technique may be used. The sample is spotted in one corner of a square sheet of paper, and one solvent is used to develop the paper in one direction. The chromatogram is then dried, turned 90°, and developed again, using a second solvent of different polarity. Another means of improving resolution is the use of ion-exchange (Sect. 12.4.4) papers, i.e., paper that has been impregnated with ion-exchange resin or paper, with derivatized cellulose hydroxyl groups (with acidic or basic moieties).

In planar chromatography, components of a mixture are often characterized by their relative mobility (R _f) value, where:

(12.2)

Unfortunately, R _f values are not always constant for a given solute/sorbent/solvent but depend on many factors, such as the quality of the stationary phase, layer thickness, humidity, development distance, and temperature.

12.3.4.1.2 Thin-Layer Chromatography

Thin-layer chromatography (TLC), first described in 1938, has largely replaced paper chromatography because it is faster, more sensitive, and more reproducible. The resolution in TLC is greater than in paper chromatography because the particles on the plate are smaller and more regular than paper fibers. Experimental conditions can be easily varied to achieve separation and can be scaled up for use in column chromatography, although thin-layer and column procedures are not necessarily interchangeable, due to differences such as the use of binders with TLC plates, vapor phase equilibria in a TLC tank, etc. There are several distinct advantages to TLC over paper chromatography and in some instances over column chromatography: high sample throughput, separations of complex mixtures, low cost, analysis of several samples and standards simultaneously, minimal sample preparation, and possibility to store the plate for later identification and quantification. Advances in TLC led to the development of high-performance thin-layer chromatography (HPTLC), which simply refers to TLC performed using plates coated with smaller, more uniform particles. This permits better separations in shorter times.

TLC, more so HPTLC, is applied in many fields, including environmental, clinical, forensic, pharmaceutical, food, flavors, and cosmetics. Within the food industry, TLC may be used for quality control. For example, corn and peanuts are tested for aflatoxins/mycotoxins prior to their processing into corn meal and peanut butter, respectively. Applications of TLC to the analysis of a variety of compounds, including lipids, carbohydrates, vitamins, amino acids, natural pigments, and sugar substitutes, are discussed in references [14, 17, 18].

1. 1.

   TLC General Procedures. TLC utilizes a thin (ca. 250 μm thick) layer of sorbent or stationary phase bound to an inert support. The support is often a glass plate (traditionally, 20 × 20 cm), but plastic sheets and aluminum foil also are used. Pre-coated plates, of different layer thicknesses, are commercially available in a wide variety of sorbents, including chemically modified silica. Four frequently used TLC sorbents are silica gel, alumina, diatomaceous earth, and cellulose. Modified silica for TLC may contain polar or nonpolar groups, so both normal and reversed-phase (see Sect. 12.4.2.1) thin-layer separations may be carried out.

   If adsorption TLC is to be performed, the sorbent is first activated by drying for a specified time and temperature. As in paper chromatography, the sample (in carrier solvent) is applied as a spot or streak about 1.5 cm from one end of the plate. After evaporation of the carrier solvent, the TLC plate is placed in a closed developing chamber, solvent migrates up the plate (ascending development) by capillary action, and sample components are separated. After the TLC plate has been removed from the chamber and solvent allowed to evaporate, the separated bands are made visible or detected by other means. Specific chemical reactions (derivatization), which may be carried out either before or after chromatography, often are used for this purpose. Two examples are reaction with sulfuric acid to produce a dark charred area (a destructive chemical method) and the use of iodine vapor to form a colored complex (a nondestructive method inasmuch as the colored complex is usually not permanent). Common physical detection methods include the measurement of absorbed or emitted electromagnetic radiation, such as measuring fluorescence when stained with 2,7-dichlorofluorescein, and measurement of β-radiation from radioactively labeled compounds. Different reagents that can react selectively to generate colored products also are used [17]. Biological methods or biochemical inhibition tests can be used to detect toxicologically active substances. An example is measuring the inhibition of cholinesterase activity by organophosphate pesticides.

   Quantitative evaluation of thin-layer chromatograms may be performed [17]: (1) in situ (directly on the layer) by using a densitometer [18], or (2) scraping a zone off the plate, eluting compound from the sorbent, and then analyzing the resultant solution (e.g., by liquid scintillation counting).

    
2. 2.
   Factors Affecting Thin-Layer Separations. In both planar and column liquid chromatography, the nature of the compounds to be separated determines what type of stationary phase is used. Separation can occur by adsorption, partition, ion-exchange, size-exclusion, or multiple mechanisms (Sect. 12.4). Table 12.2 lists the separation mechanisms involved in some typical applications on common TLC sorbents.

   table 12.2

   Thin-layer chromatography sorbents and mode of separation

   Sorbent	Chromatographic mechanism	Typical application
   Silica gel	Adsorption	Steroids, amino acids, alcohols, hydrocarbons, lipids, aflatoxins, bile acids, vitamins, alkaloids
   Silica gel RP	Reversed phase	Fatty acids, vitamins, steroids, hormones, carotenoids
   Cellulose, kieselguhr	Partition	Carbohydrates, sugars, alcohols, amino acids, carboxylic acids, fatty acids
   Aluminum oxide	Adsorption	Amines, alcohols, steroids, lipids, aflatoxins, bile acids, vitamins, alkaloids
   PEI cellulosea	Ion exchange	Nucleic acids, nucleotides, nucleosides, purines, pyrimidines
   Magnesium silicate	Adsorption	Steroids, pesticides, lipids, alkaloids

   Reprinted from Touchstone [19] by permission of Wiley, New York

   ^aPEI cellulose refers to cellulose derivatized with polyethyleneimine (PEI)

    

Solvents for TLC separations are selected for specific chemical characteristics and solvent strength (a measure of interaction between solvent and sorbent; see Sect. 12.4.1). In simple adsorption TLC, the higher the solvent strength, the greater the R _f value of the solute. An R _f value of 0.3–0.7 is typical. Mobile phases have been developed for the separation of various compound classes on the different sorbents (see Table 7.​1 in reference [19]).

In addition to the sorbent and solvent, several other factors must be considered when performing planar chromatography. These include the type of developing chamber used, vapor phase conditions (saturated vs. unsaturated), development mode (ascending, descending, horizontal, radial, etc.), and development distance. For additional reading refer to references [14–18].

12.3.4.2 Column Liquid Chromatography

Column liquid chromatography generally has enhanced resolution of solutes in a mixture and enables precise analysis compared to planar chromatography. Fractionation of solutes occurs as a result of differential migration through a closed tube of stationary phase, and analytes can be monitored while the separation is in progress. In column liquid chromatography, the mobile phase is liquid, and the stationary phase can be either solid or liquid supported by an inert solid. A system for low-pressure (i.e., performed at or near atmospheric pressure) column liquid chromatography is illustrated in Fig. 12.2.

Having selected a stationary and mobile phase suitable for the separation problem at hand, the analyst must first prepare the stationary phase (resin, gel, or packing material) for use according to the supplier’s instructions (e.g., the stationary phase often must be hydrated or preswelled in the mobile phase). The prepared stationary phase then is packed into a column (usually glass), the length and diameter of which are determined by the amount of sample to be loaded, the separation mode to be used, and the degree of resolution required. Longer and narrower columns usually enhance resolution and separation (Sect. 12.5.1). Adsorption columns may be either dry or wet packed; other types of columns are wet packed. The most common technique for wet packing involves making a slurry of the adsorbent with the solvent and pouring it into the column to the desired bed height. Pouring uniform columns is an art that is mastered with practice. If the packing solvent is different from the initial eluting solvent, the column must be thoroughly washed ( equilibrated with 2–3 column volumes) with the starting mobile phase.

The sample to be fractionated is dissolved in a minimum volume of the starting mobile phase, injected through a sample injection port, and carried by the mobile phase onto the column. Low-pressure chromatography utilizes only gravity flow or a high-precision peristaltic pump to maintain a constant flow of mobile phase (eluent or eluting solvent) through the column. If eluent is fed to the column by a peristaltic pump (see Fig. 12.2), then the flow rate is determined by the pump speed. Depending on the dimensions of the column, the flow rate is adjusted not to exceed the max pressure sustained by the pump.

The process of passing the mobile phase through the column is called elution, and the portion that emerges from the outlet end of the column is called the eluate (or effluent). Elution may be isocratic (constant mobile phase composition) or a gradient (changing the mobile phase, e.g., increasing solvent strength or pH). Gradient elution enhances resolution and decreases analysis time (see also Sect. 12.5.1). As elution proceeds, components of the sample are selectively retarded by the stationary phase based on the strength of interaction with the stationary phase and thus are eluted at different times.

The column eluate may be directed through a detector and then into tubes, changed at intervals by a fraction collector. The detector response, in the form of an electrical signal, may be recorded (the chromatogram) using a computerized software. Signals are then integrated for either qualitative or quantitative analysis (Sects. 12.5.2 and 12.5.3). The fraction collector may be set to collect eluate at specified time intervals or after a certain volume or number of drops have been collected. Components of the sample that have been chromatographically separated and collected can be further analyzed as needed.

12.3.5 Supercritical Fluid Chromatography

Supercritical fluid chromatography is performed above the critical pressure (P _c) and critical temperature (T _c) of the mobile phase. A supercritical fluid (or compressed gas) is neither a liquid nor a typical gas. The combination of P _c and T _c is known as the critical point. A supercritical fluid can be formed from a conventional gas by increasing the pressure or from a conventional liquid by raising the temperature. Carbon dioxide frequently is used as a mobile phase for SFC; however, it is not a good solvent for polar and high-molecular-weight compounds. A small amount of a polar, organic solvent such as methanol can be added to a nonpolar supercritical fluid to enhance solute solubility, improve peak shape, and alter selectivity. Other supercritical fluids that have been used in food applications include nitrous oxide, trifluoromethane, sulfur hexafluoride, pentane, and ammonia.

Supercritical fluids confer chromatographic properties intermediate to LC and GC. The high diffusivity and low viscosity of supercritical fluids mean decreased analysis times and improved resolution compared to LC. An additional benefit of short analysis time is the reduced solvent consumption. SFC offers a wide range of selectivity (Sect. 12.5.2) adjustment, by changes in pressure and temperature as well as changes in mobile phase composition and the stationary phase. Compared to HPLC (Chap. 13), SFC is better for separating compounds with a broader range of polarities. In addition, SFC makes possible the separation of nonvolatile, thermally labile compounds, which cannot be analyzed by GC without derivatization. In fact, SFC today is mainly used for the analysis of nonvolatiles.

Either packed or capillary columns can be used in SFC. Packed column materials are similar to those used for HPLC. Small particle, porous, high surface area, hydrated silica, either bare or bonded silica, is commonly used as the column packing material (Sect. 12.4.2.3 and Chap. 13). Capillary columns are generally coated with a polysiloxane (-Si-O-Si) film containing different functional groups, which is then cross-linked to form a polymeric stationary phase that cannot be washed off by the mobile phase. Polysiloxanes containing different functional groups, such as methyl, phenyl, pyridine, or cyano, may be used to vary the polarity of this stationary phase.

The development of SFC benefited from advancement in the instrumentation for HPLC and GC. One major difference in instrumentation is the presence of a back pressure regulator to control the outlet pressure of the system. Without this device, the fluid would expand to a low-pressure, low-density gas. Detectors used in GC and HPLC also can be used with SFC, including coupling with mass spectrometry (MS) (Chap. 11).

SFC has been used for the analysis of a wide range of compounds. Fats, oils, and other lipids are compounds to which SFC is increasingly applied. For example, the noncaloric fat substitute, Olestra^®, was characterized by SFC-MS. Other researchers have used SFC to detect pesticide residues, study thermally labile compounds from members of the Allium genus, fractionate citrus essential oils, and characterize compounds extracted from microwave packaging [20]. Bernal et al. [11] highlighted the use of packed column and capillary SFC for the analysis of food and natural products, namely, lipids and their derivatives, carotenoids, fat-soluble vitamins, polyphenols, carbohydrates, and food adulterants such as Sudan dyes.

12.4.1 Adsorption (Liquid-Solid) Chromatography

Sites available for interaction with any given substance are heterogeneous. Binding sites with greater affinities, the most active sites, tend to be populated first, so that additional solutes are less firmly bound. The net result is that adsorption is a concentration-dependent process, and the adsorption coefficient is not a constant (in contrast to the partition coefficient). Sample loads exceeding the adsorptive capacity of the stationary phase will result in relatively poor separation (Sect. 12.5.1).

Adsorption chromatography can be used to separate aromatic or aliphatic nonpolar compounds, based primarily on the type and number of functional groups present. The labile, fat-soluble chlorophyll and carotenoid pigments from plants have been studied extensively by adsorption column chromatography. Adsorption chromatography also has been used for the analysis of fat-solu is also another ble vitamins. Frequently, it is used as a batch procedure for the removal of impurities from samples prior to other analyses. For example, disposable solid-phase extraction cartridges (see Chap. 14) containing silica have been used for analyses of lipids in soybean oil, carotenoids in citrus fruit, and vitamin E in grain. Adsorption chromatography is also applied in specialized forms for the analysis of a wide range of compounds. Several of the chromatographic separation techniques described in the following sections are forms of specialized adsorption chromatography.

12.4.2 Partition (Liquid-Liquid) Chromatography

12.4.3 Hydrophobic Interaction Chromatography

Hydrophobic interaction chromatography (HIC) has gained popularity over recent years for the purification of compounds on a preparative and semi-analytical scale. In HIC biomolecules adsorb to a weak hydrophobic surface at high salt concentration. Elution of adsorbed molecules is achieved by decreasing the salt concentration of the mobile phase over time. This technique takes advantage of hydrophobic moieties on the surface of a compound. Accordingly, HIC is very commonly used for the purification of food proteins, enzymes, and antibodies while offering high resolution. The high salt concentration allows biomolecules with high surface charge to adsorb to the hydrophobic ligands by shielding the charges. Salt precipitation of proteins, in particular, onto the column does not cause denaturation, since the interaction with the hydrophobic ligands is weak.

The stationary phase in HIC consists of hydrophilic support bonded to hydrophobic ligands. Several polymeric materials can be used as support, including cellulose, agarose, chitosan, polymethacrylate, or silica. The support must be hydrophilic so as not to contribute additional hydrophobicity and hence strong interactions that may cause denaturation of the protein. Often the polymer used has a high degree of cross-linking to provide rigidity and high surface area. Most commonly used ligands, chemically bonded to the support polymer, are linear chain alkanes. The size of the alkyl chains used depends on the surface hydrophobicity of the biomolecules to be separated, with the length of the chain being higher for more hydrophobic biomolecules. Sometimes phenyl and other aromatic groups are also used. Often times though, ligands with intermediate hydrophobicity are used to avoid strong interactions. Butyl-Sepharose^® and phenyl-Sepharose^® are among the most commonly used stationary phases.

Different salts are used in HIC depending on their effects on protein precipitation. The most commonly used salt is ammonium sulfate. Concentration of the salt is also a determining factor for the precipitation of the protein on the column. Often 1 M ammonium sulfate is used for initial screening. It is important to prepare the sample in the same salt solution/buffer as the mobile phase. This, however, necessitates care in loading the sample to avoid precipitation of the protein prior to reaching the column. A wash step generally precedes loading the sample to allow for washing out impurities. Depending on the sample characteristics, elution can be performed gradually by decreasing salt concentration over time, which may allow for isolation of different proteins in the sample in less time. Changing of salt concentration over time requires optimization for best resolution and shortest analysis time. Switching directly to water after the salt wash also may be performed to elute all bound protein with minimal separation. This is dependent on the level of fractionation and purification required. If a compound resists elution even after reducing salt concentration to zero, other HIC ligands should be tried. Cleaning and regeneration of the column are required after several runs. Often 0.1–1 M NaOH is used to prevent fouling of the column, and sometimes detergents and alcohol washes are used.

The pH of the mobile phase also can influence retention on the column and elution. However, oftentimes buffers with pH 7 are used. Additives, such as water-miscible alcohols and detergents, are sometimes used to help elute the protein faster. The hydrophobic parts of these additives will compete with the protein for binding to the ligand, causing desorption of the protein. Finally, temperature may play a role in HIC. As temperature increases, hydrophobic interaction increases allowing for better retention on the column; lowering the temperature aids in elution. Control of temperature during a run may enhance resolution and reduce analysis time (see also Sect. 2.​5.​1). For further details on HIC and applications, readers are directed to reference [22].

12.4.4 Ion-Exchange Chromatography

Separation by ion exchange can be categorized into three types: (1) ionic from nonionic, (2) cationic from anionic, and (3) mixtures of similarly charged species. In the first two cases, one substance binds to the ion-exchange medium, whereas the other substance does not. Batch extraction methods can be used for these two separations; however, chromatography is needed for the third category.

Ion-exchange chromatography is viewed as a specialized type of adsorption chromatography in which interactions between solute and stationary phase are primarily electrostatic in nature. The stationary phase (ion exchanger) contains fixed functional groups that are either negatively or positively charged (Fig. 12.4a). Exchangeable counterions preserve charge neutrality. A sample ion (or charged sites on large molecules) can exchange with the counterion for the interaction with the charged functional group. Ionic equilibrium is established as depicted in Fig. 12.4b. The functional group of the stationary phase determines whether cations or anions are exchanged. Cation exchangers possess covalently bound negatively charged functional groups, whereas anion exchangers possess bound positively charged groups. The chemical nature of these acidic or basic residues determines how stationary phase ionization is affected by the mobile phase pH.

The strongly acidic sulfonic acid moieties (RSO3-) of “strong” cation exchangers are completely ionized at all pH values above 2. Strongly basic quaternary amine groups (RNR′3+) on “strong” anion exchangers are ionized at all pH values below 10. Since maximum negative or positive charge is maintained over a broad pH range, the exchange or binding capacity of these stationary phases is essentially constant, regardless of mobile phase pH. “Weak” cation exchangers contain weakly acidic carboxylic acid functional groups (RCO2−); consequently, their exchange capacity varies considerably between ca. pH 4 and 10. Weakly basic anion exchangers possess primary, secondary, or tertiary amine residues (R-NHR′2+) that are deprotonated in moderately basic solution, thereby losing their positive charge and the ability to bind anions. Thus, one way of eluting solutes bound to weak ion-exchange medium is to change the mobile phase pH. Changing of the pH in the case of weak ion exchangers allows for enhanced separation and better selectivity compared to strong ion exchangers when separating compounds with very similar charge densities. While strong ion exchangers are generally used for initial screening and optimization of the separation, oftentimes, weak ion exchangers are used to separate compounds that have similar adsorption coefficient, making use of a pH gradient. Changing the pH, however, may present a challenge when separating proteins. One must avoid the isoelectric point of the proteins during chromatographic separation; otherwise they will fall out of solution. A second way to elute bound solutes from either strong or weak ion exchangers is to increase the ionic strength (e.g., use NaCl) of the mobile phase, to weaken the electrostatic interactions.

Chromatographic separations by ion exchange are based upon differences in affinity of the exchangers for the ions (or charged species) to be separated. The factors that govern selectivity of an exchanger for a particular ion include: (1) ionic valence, radius, and concentration, (2) nature of the exchanger (including its displaceable counterion), and (3) composition and pH of the mobile phase. To be useful as an ion exchanger, a material must be both ionic in nature and highly permeable. Synthetic ion exchangers are thus cross-linked polyelectrolytes, and they may be inorganic (e.g., aluminosilicates) or, more commonly, organic compounds. Two commonly used types of organic compound-based synthetic ion exchangers are polystyrene and polysaccharide.

Polystyrene, made by cross-linking styrene with divinylbenzene (DVB), may be modified to produce either anion- or cation-exchange resins (Fig. 12.5). Polymeric resins such as these are commercially available in a wide range of particle sizes and with different degrees of cross-linking (expressed as weight percent of DVB in the mixture). The extent of cross-linking controls the rigidity and porosity of the resin, which, in turn, determines its optimal use. Lightly cross-linked resins permit rapid equilibration of solute, but particles swell in water, thereby decreasing charge density and selectivity (relative affinity) of the resin for different ions. More highly cross-linked resins exhibit less swelling, higher exchange capacity, and selectivity, but longer equilibration times. The small pore size, high charge density, and inherent hydrophobicity of the older ion-exchange resins have limited their use to small molecules [molecular weight (MW) <500].

Ion exchangers based on polysaccharides, such as cellulose, dextran, or agarose, have proven very useful for the separation and purification of large molecules, such as proteins and nucleic acids. These materials, called gels, are much softer than polystyrene resins, thus may be derivatized with strong or with weak acidic or basic groups via OH moieties on the polysaccharide backbone (Fig. 12.6). They have much larger pore sizes and lower charge densities than the older synthetic resins.

Food-related applications of ion-exchange chromatography include the separation of amino acids, sugars, alkaloids, and proteins. Fractionation of amino acids in protein hydrolysates was initially carried out by ion-exchange chromatography; automation of this process led to the development of commercially produced amino acid analyzers (see Chap. 24). Many drugs, fatty acids, and organic acids, being ionizable compounds, may be chromatographed in the ion-exchange mode. For additional details on the principles and applications of ion chromatography, please refer to reference [23].

12.4.5 Affinity Chromatography

Affinity chromatography is also another specialized form of adsorption chromatography, with the separation being based on very specific, reversible interaction between a solute molecule and a ligand immobilized on the chromatographic stationary phase (types of interactions are listed in Sect. 12.4.1). Affinity chromatography involves immobilized biological ligands as the stationary phase. These ligands can be antibodies, enzyme inhibitors, lectins, or other molecules that selectively and reversibly bind to complementary analyte molecules in the sample. Although both ligands and the species to be isolated are usually biological macromolecules, the term affinity chromatography also encompasses other systems, such as separation of small molecules containing cis-diol groups via phenylboronic acid moieties on the stationary phase.

The principles of affinity chromatography are illustrated in Fig. 12.7. A ligand, chosen based on its specificity/selectivity and strength of interaction with the analyte, is immobilized on a suitable support material. As the sample is passed through the column, analytes that are complementary to the bound ligand are adsorbed, while other sample components are eluted. Bound analytes are subsequently eluted via a change in the mobile phase composition as will be discussed below. After reequilibration with the initial mobile phase, the stationary phase is ready to be used again. The ideal support for affinity chromatography should be a porous, stable, high-surface-area material that does not adsorb anything itself. Thus, polymers such as agarose, cellulose, dextran, and polyacrylamide are used, as well as controlled pore glass.

Affinity ligands are usually attached to the support or matrix by covalent bond formation, and optimum reaction conditions often must be found empirically. Immobilization generally consists of two steps: activation and coupling. During the activation step, a reagent reacts with functional groups on the support, such as hydroxyl moieties, to produce an activated matrix. After removal of excess reagent, the ligand is coupled to the activated matrix. (Preactivated supports are commercially available, and their availability has greatly increased the use of affinity chromatography.) The coupling reaction most often involves free amino groups on the ligand, although other functional groups can be used. When small molecules such as phenylboronic acid are immobilized, a spacer arm (containing at least four to six methylene groups) is used to hold the ligand away from the support surface, enabling it to reach into the binding site of the analyte.

Elution methods for affinity chromatography may be divided into nonspecific and ( bio)specific methods. Nonspecific elution involves disrupting ligand analyte binding by changing the mobile phase pH, ionic strength, dielectric constant, or temperature. If additional selectivity in elution is desired, for example, in the case of immobilized general ligands, a biospecific elution technique is used. Free ligand, either identical to or different from the matrix-bound ligand, is added to the mobile phase. This free ligand competes for binding sites on the analyte. For example, glycoproteins bound to a concanavalin A (lectin) column can be eluted by using buffer containing an excess of lectin. In general, the eluent ligand should display greater affinity for the analyte of interest than the immobilized ligand.

In addition to protein purification, affinity chromatography may be used to separate supramolecular structures such as cells, organelles, and viruses; concentrate dilute protein solutions; investigate binding mechanisms; and determine equilibrium constants. Affinity chromatography has been useful especially in the separation and purification of enzymes and glycoproteins. In the case of the latter, carbohydrate-derivatized adsorbents are used to isolate specific lectins, such as concanavalin A, and lentil or wheat germ lectin. The lectin then agarose may be coupled to agarose, such as concanavalin A-agarose or lentil lectin-agarose, to provide a stationary phase for the purification of specific glycoproteins, glycolipids, or polysaccharides. Other applications of affinity chromatography include purification and quantification of mycotoxins [24] and pesticide residues/metabolites [25] in food/crops. For additional details on affinity chromatography, refer to references [26, 27].

12.4.6 Size-Exclusion Chromatography

Size-exclusion chromatography (SEC), also known as molecular exclusion, gel permeation (GPC), and gel filtration chromatography (GFC), is probably the easiest mode of chromatography to perform and to understand. It is widely used in the biological sciences for the resolution of macromolecules, such as proteins and carbohydrates, and also is used for the fractionation and characterization of synthetic polymers. Unfortunately, nomenclature associated with this separation mode developed independently in the literature of the life sciences and in the field of polymer chemistry, resulting in inconsistencies.

In the ideal SEC system, molecules are separated solely on the basis of their size; no interaction occurs between solutes and the stationary phase. In the event that solute/support interactions do occur, the separation mode is termed nonideal SEC. The stationary phase in SEC consists of a column packing material that contains pores comparable in size to the molecules to be fractionated. Solutes too large to enter the pores travel with the mobile phase in the interstitial space (between particles) outside the pores. Thus, the largest molecules are eluted first from an SEC column. The volume of the mobile phase in the column, termed the column void volume, Vo, can be measured by chromatographing a very large (totally excluded) species, such as Blue Dextran, a dye of molecular weight (MW) two million.

As solute dimensions decrease, approaching those of the packing pores, molecules begin to diffuse into the packing particles and, consequently, are slowed down. Solutes of low MW (e.g., glycyltyrosine) that have free access to all the available pore volume are eluted in the volume referred to as Vt, the total permeation volume of the column. The Vt is equal to the column void volume, Vo, plus the volume of liquid inside the sorbent pores (internal pore volume), Vi (Vt = Vo + Vi). These relationships are illustrated in Fig. 12.8. Solutes are ideally eluted between the void volume and the total permeation volume of the column. Because this volume is limited, only a relatively small number of solutes (ca. 10), of a particular size range, can be completely resolved by SEC under ordinary conditions.

The behavior of a molecule in a size-exclusion column may be characterized in several different ways. Each solute exhibits an elution volume, Ve, as illustrated in Fig. 12.9. However, Ve depends on column dimensions, the way in which the column was packed, and the sorbent pore size. The available partition coefficient, Kav, is used to define solute behavior independent of these variables:

(12.3)

where:

• Kav = available partition coefficient

• Ve = elution volume of solute

• Vo = column void volume

• Vt = total permeation volume of column

The value of Kav calculated from experimental data for a solute chromatographed on a given SEC column defines the proportion of pores that can be occupied by that molecule. For a large, totally excluded species, such as Blue Dextran or DNA, Ve = Vo and Kav = 0. For a small molecule with complete access to the internal pore volume, such as glycyltyrosine, Ve = Vt and Kav = 1.

For each size-exclusion packing material, a plot of Kav vs. the logarithm of the MW for a series of solutes, similar in molecular shape and density, will give an S-shaped curve (Fig. 12.9). In the case of proteins, Kav is actually better related to the Stokes radius, the average radius of the protein in solution. The central, linear portion of this curve describes the fractionation range of the matrix, wherein maximum separation among solutes of similar MW is achieved. This correlation between solute elution behavior and MW (or size) forms the basis for a widely used method for characterizing large molecules such as proteins and polysaccharides. A size-exclusion column is calibrated with a series of solutes of known MW (or Stokes radius) to obtain a curve similar to that shown in Fig. 12.9. The value of Kav for the unknown is then determined, and an estimate of MW (or size) of the unknown is made by interpolation of the calibration curve.

Column packing materials for SEC can be divided into two groups: semirigid, hydrophobic media and soft, hydrophilic gels. The former are usually derived from polystyrene and are used with organic mobile phases (GPC or nonaqueous SEC) for the separation of polymers, such as rubbers and plastics. Soft gels, polysaccharide-based packings, are typified by Sephadex, a cross-linked dextran (see Fig. 12.6a). These materials are available in a wide range of pore sizes and are useful for the separation of water-soluble substances in the MW range 1–2.5 × 10⁷ Da. In selecting an SEC column packing, both the purpose of the experiment and size of the molecules to be separated must be considered. If the purpose of the experiment is group separation, for which molecules of widely different molecular sizes need to be separated, a matrix is chosen such that the larger molecules, e.g., proteins, are eluted in the void volume of the column, whereas small molecules are retained in the total volume. When SEC is used for separation of macromolecules of different sizes, the molecular sizes of all the components must fall within the fractionation range of the gel.

As discussed previously, SEC can be used, directly, to fractionate mixtures, to purify, or, indirectly, to obtain MW/size information about a dissolved species. In addition to MW estimations, SEC is used to determine the MW distribution of natural and synthetic polymers, such as dextrans and gelatin preparations. Fractionation of biopolymer mixtures is probably the most widespread use of SEC, since the mild elution conditions employed rarely cause denaturation or degradation. Often times SEC is used as an early chromatographic separation step in a multidimensional chromatographic approach toward purification. It is also a fast, efficient alternative to dialysis for desalting solutions of macromolecules, such as proteins.

12.5.1 Separation and Resolution

This section will discuss separation and resolution as it pertains mainly to LC; separation and resolution optimization as it pertains specifically to GC will be discussed in Chap. 14.

12.5.1.1 Developing a Separation

There may be numerous ways to accomplish a chromatographic separation for a particular compound. In many cases, the analyst will follow a standard laboratory procedure or published methods. In the case of a sample that has not been previously analyzed, the analyst begins by evaluating what is known about the sample and defines the goals of the separation. How many components need to be resolved? What degree of resolution is needed? Is qualitative or quantitative information needed? Molecular weight (or MW range)/size, polarity, and ionic character of the sample will guide the choice of chromatographic separation mechanism (separation mode). Figure 12.10 shows that more than one correct choice may be possible. For example, small ionic compounds may be separated by ion-exchange, ion-pair reversed-phase (see Sect. 13.​3.​2.​1), or reversed-phase LC. In this case, the analyst’s choice may be based on convenience, experience, and personal preference.

Having chosen a separation mode for the sample at hand, one must select an appropriate stationary phase, elution conditions, and a detection method. Trial experimental conditions may be based on the results of a literature search, the analyst’s previous experience with similar samples, or the general recommendations from chromatography experts.

To achieve separation of sample components by all modes except SEC, one may utilize either isocratic or gradient elution. Isocratic elution is the simplest technique, in which solvent composition and flow rate are held constant throughout the run. Gradient elution involves reproducibly varying mobile phase composition or flow rate (flow programming) during the LC analysis. Gradient elution is used when sample components vary a lot in inherent characteristics such as polarity and/or charge density, so that an isocratic mobile phase does not elute all components within a reasonable time. The change may be continuous or stepwise. Gradients of increasing ionic strength, or changing pH, are extremely valuable in ion-exchange chromatography (see Sect. 12.4.4), whereas gradients of increasing or decreasing polarity are valuable in normal or reversed-phase chromatography, respectively (Sect. 12.4.2). Increasing the “strength” of the mobile phase (Sect. 12.4.1), either gradually (continuously) or in a stepwise fashion, shortens the analysis time.

Method development may begin with an isocratic mobile phase, possibly of intermediate solvent strength; however, using gradient elution for the initial separation may ensure that some level of separation is achieved within a reasonable time period and nothing is likely to remain on the column. Data from this initial run allows one to determine if isocratic or gradient elution is needed, and to estimate optimal isocratic mobile phase composition or gradient range.

Once an initial separation has been achieved, the analyst can proceed to optimize resolution (separation of closely related analytes). This generally involves changing mobile phase variables, including the following: (1) nature and percentage of organic solvents (and type of organic solvents), (2) pH, (3) ionic strength, (4) nature and concentration of additives (such as ion-pairing agents), (5) flow rate, and (6) temperature. In the case of gradient elution, gradient steepness (slope, i.e., rate of change) is another variable to be optimized. However, the analyst must be aware of the principles of chromatographic resolution as will be discussed in the following section.

12.5.1.2 Chromatographic Resolution

12.5.1.2.1 Introduction

The main goal of chromatography is to segregate components of a sample into separate bands or peaks as they migrate through the column. A chromatographic peak is defined by several parameters including retention time (Fig. 12.11), peak width, and peak height (Fig. 12.12). The volume of the mobile phase required to elute a compound from an LC column is called the retention volume, V _R. The associated time is the retention time, t _R. Shifts in retention time and changes in peak width greatly influence chromatographic resolution.

Differences in column dimensions, loading, temperature, mobile phase flow rate, system dead volume, and detector geometry may lead to discrepancies in retention time. By subtracting the time required for the mobile phase or a non-retained solute (t _m or t _o) to travel through the column to the detector, one obtains an adjusted retention time, t′_R (or volume), as depicted in Fig. 12.11. The adjusted retention time (or volume) corrects for differences in system dead volume and signifies the time the analyte spends adsorbed on the stationary phase.

The resolution of two peaks from each other is related to the separation factor, α. Values for α (Fig. 12.11) depend on temperature, flow rate, stationary phase, and mobile phase used. Resolution is defined as follows:

(12.4)

where:

• R _s = resolution

• Δt = difference between retention times of peaks 1 and 2

• w ₂ = width of peak 2 at baseline

• w ₁ = width of peak 1 at baseline

Figure 12.12 illustrates the measurement of peak width [part (a)] and the values necessary for calculating resolution [part (b)]. (Retention and peak or band width must be expressed in the same units, i.e., time or volume).

Chromatographic resolution is a function of column efficiency, selectivity, and the capacity factor. Mathematically, this relationship is expressed as follows:

(12.5)

where:

• a = column efficiency term

• b = column selectivity term

• c = capacity term

These terms, and factors that contribute to them, will be discussed in the following sections.

12.5.1.2.2 Column Efficiency

If faced with the problem of improving resolution, a chromatographer should first examine the efficiency of the column. An efficient column keeps the solutes from spreading widely on the column and hence gives narrow peaks. Column efficiency can be calculated by:

(12.6)

where:

• N = number of theoretical plates

• t _R = retention time

• σ = standard deviation for a Gaussian peak

• w = peak width at baseline (w = 4σ)

• w _½ = peak width at half height

The measurement of t _R, w, and w _½ is illustrated in Fig. 12.12. (Retention volume may be used instead of t _R; in this case, peak width is also measured in units of volume.) Although some peaks are not actually Gaussian in shape, normal practice is to treat them as if they were. In the case of peaks that are incompletely resolved or slightly asymmetric, peak width at half height is more accurate than peak width at baseline.

The value N calculated from the above equation is called the number of theoretical plates. The theoretical plate concept, borrowed from distillation theory, can best be understood by viewing chromatography as a series of equilibrations of solutes between mobile and stationary phases, analogous to countercurrent distribution. Thus, a column would consist of N segments (theoretical plates) with one equilibration occurring in each. As a first approximation, N is independent of retention time and is therefore a useful measure of column performance. One method of monitoring column performance over time is to chromatograph a standard compound periodically, under constant conditions, and to compare the values of N obtained. It is important to note that columns often behave as if they have a different number of plates for different solutes in a mixture. Different solutes have different partition coefficient and thus have distinctive series of equilibrations between mobile and stationary phases. Peak broadening due to column deterioration will result in a decrease of N for a particular solute. Peak broadening is a result of an extended time for a solute to reach equilibrium between mobile and stationary phases (causing them to spread over a wider area on the column).

The number of theoretical plates is generally proportional to column length. The longer the column, the higher the number of theoretical plates, and hence the better is the resolution. Because columns are available in various lengths, it is useful to have a measure of column efficiency that is independent of column length. This may be expressed as:

(12.7)

where:

• HETP = height equivalent to a theoretical plate

• L = column length

• N = number of theoretical plates

The so-called HETP, height equivalent to a theoretical plate, is sometimes more simply described as plate height (H). If a column consisted of discrete segments, HETP would be the height of each imaginary segment. Small plate height values (a large number of plates) indicate good efficiency of separation (minimal spread of solute within the column, resulting in sharp peaks). Conversely, reduced number of plates results in poor separation due to the extended equilibrium time (spread of solutes on the column, resulting in wide peaks) in a deteriorating column.

In reality, columns are not divided into discrete segments and equilibration is not infinitely fast. The plate theory is used to simplify the equilibration concept. The movement of solutes through a chromatography column takes into account the finite rate at which a solute can equilibrate itself between stationary and mobile phases. Thus, peak shape depends on the rate of elution and is affected by solute diffusion. Any mechanism that causes a peak of solute to broaden will increase HETP and decrease column efficiency. The various factors that contribute to plate height are expressed by the Van Deemter equation:

(12.8)

where:

• HETP = height equivalent to a theoretical plate

• A, B, C = constants

• u = mobile phase rate

The constants A, B, and C are characteristic for a given column, mobile phase, and temperature. The A term represents the eddy diffusion or multiple flowpaths. Eddy diffusion refers to the different microscopic flowstreams that the mobile phase can take between particles in the column (analogous to eddy streams around rocks in a brook). Sample molecules can thus take different paths as well, depending on which flowstreams they follow. As a result, solute molecules spread from an initially narrow band to a broader area within the column. The larger is the particle size of the packing material, the more paths a solute might take. Eddy diffusion may be minimized by good column packing techniques and the use of small diameter particles of narrow particle size distribution.

The B term of the Van Deemter equation, sometimes called the longitudinal diffusion term, exists because all solutes diffuse from an area of high concentration (the center of a chromatographic peak) to one of low concentration (the leading or trailing edge of a chromatographic peak). In LC, the contribution of this term to HETP is small except at low flow rate of the mobile phase. With slow flow rates there will be more time for a solute to spend on the column; thus, its diffusion will be greater.

The C (mass transfer) term arises from the finite time required for solute to equilibrate between the mobile and stationary phases. Mass transfer is practically the partitioning of the solute into the stationary phase, which does not occur instantaneously and depends on the solute’s partition and diffusion coefficients. If the stationary phase consists of porous particles (see Chap. 13, Sect. 13.​2.​3.​2, Fig. 13.​3), a sample molecule entering a pore ceases to be transported by the solvent flow and moves by diffusion only. Subsequently, this solute molecule may diffuse back to the mobile phase flow or it may interact with the stationary phase. In either case, solute molecules inside the pores are slowed down relative to those outside the pores, and peak broadening occurs. Contributions to HETP from the C term can be minimized by using porous particles of small diameter or pellicular packing materials (Chap. 13, Sect. 13.​2.​3.​2.​2). Also, using a narrower column with a smaller inner diameter reduces the C value, given that equilibrium time will be reduced since there is less packing material.

As expressed by the Van Deemter equation, mobile phase flow rate, u, contributes to plate height in opposing ways – increasing the flow rate increases the equilibration point (Cu), but decreases longitudinal diffusion of the solute particles (B/u). A Van Deemter plot (Fig. 12.13) may be used to determine the mobile phase flow rate at which plate height is minimized and column efficiency is maximized. Flow rates above the optimum may be used to decrease analysis time if adequate resolution is still obtained. However, at very high flow rates, there will be less time to approach equilibrium, which will lead to a shorter retention time and co-elution of closely related solutes.

In addition to flow rate, temperature can affect the longitudinal diffusion and the mass transfer. Increasing the temperature causes enhanced movement of the solute between the mobile phase and the stationary phase and within the column, thus leading to faster elution and narrower peaks.

12.5.1.2.3 Column Selectivity

Chromatographic resolution depends on column selectivity as well as efficiency. Column selectivity refers to the distance, or relative separation, between two peaks and is given by:

(12.9)

where:

• α = separation factor

• t _R1 and t _R2 = retention times of components l and 2, respectively

• t _o (or t _m) = retention time of unretained components (solvent front)

• t′ _Rl and t′ _R2 = adjusted retention times of components l and 2, respectively

• K _l and K ₂ = distribution coefficients of components l and 2, respectively

Retention times (or volumes) are measured as shown in Fig. 12.11. The time, to, can be measured by chromatographing a solute that is not retained under the separation conditions (i.e., travels with the solvent front). When this parameter is expressed in units of volume, V _o or V _m, it is known as the dead volume of the system. Selectivity is a function of the stationary and/or mobile phase. For example, selectivity in ion-exchange chromatography is influenced by the nature and number of ionic groups on the matrix but also can be affected by pH and ionic strength of the mobile phase. Changing the pH or the ionic strength of the mobile phase will impact the extent of interaction of the solutes with the column. Selectivity can be either static as in isocratic elution (the choice of the mobile phase from the beginning impact selectivity) or dynamic as in gradient elution (selectivity changing over time). Good selectivity is probably more important to a given separation than high efficiency (Fig. 12.14), since resolution is directly related to selectivity but is quadratically related to efficiency; thus, a fourfold increase in N is needed to double Rs (Eq. 12.5).

12.5.1.2.4 Column Capacity Factor

The capacity or retention factor, k′, is a measure of the amount of time a chromatographed species (solute) spends in/on the stationary phase relative to the mobile phase. The relationship between capacity factors and chromatographic retention (which may be expressed in units of either volume or time) is shown below:

(12.10)

where:

• k′ = capacity factor

• K = distribution coefficient of the solute

• V _s = volume of stationary phase in column

• V _m = volume of mobile phase

• V _R = retention volume of solute

• t _R = retention time of solute

• t _o = retention time of unretained components (solvent front)

Small values of k′ indicate little retention, and components will be eluted close to the solvent front, resulting in poor separations. Overuse or misuse of the column may lead to the loss of some functional groups, thus resulting in small k′ values. Large values of k′ result in improved separation but also can lead to broad peaks and long analysis times. On a practical basis, k′ values within the range of 1–15 are generally used. (In the equation for R _s, k′ is actually the average of k ₁′ and k ₂′ for the two components separated.)

12.5.2 Qualitative Analysis

12.5.3 Quantitative Analysis

Assuming that good chromatographic resolution and identification of sample components have been achieved, quantification involves mostly measuring peak area and comparing the data with those for standards of known concentration. Nowadays all chromatography systems use data analysis software that recognizes the start, maximum, and end of each chromatographic peak, even when not fully resolved from other peaks. These values then are used to determine retention times, peak height, and peak areas. At the end of each run, a report is generated that lists these data and post-run calculations, such as relative peak areas, areas as percentages of the total area, and relative retention times. If the system has been standardized, data from external or internal standards can be used to calculate analyte concentrations (discussed below). In low-pressure, preparative chromatography, post-chromatography analysis of collected fractions is sometimes used to identify samples eluted. Examples of post-chromatography analysis include the BCA (bicinchoninic acid) protein assay (Chap. 18, Sect. 18.​4.​2.​3) and the phenol-sulfuric acid assay for carbohydrate (Chap. 19, Sect. 19.​3). After obtaining the absorbance reading on a spectrophotometer for such assays, the results are plotted as fraction number on the x-axis and absorbance on the y-axis to determine which fractions contain protein and/or carbohydrate.

Having determined peak areas, one must compare these data with appropriate standards of known concentration to determine sample concentrations. Comparisons may be by means of external or internal standards. Comparison of peak area of unknown samples with standards injected separately (i.e., external standards) is common practice. Standard solutions covering the desired concentration range (preferably diluted from one stock solution) are chromatographed, and the peak area is plotted versus concentration to obtain a standard curve. An identical volume of sample is then chromatographed, and an area of the sample peak is used to determine sample concentration via the standard curve (Fig. 12.15a). This absolute calibration method requires precise analytical technique and requires that detector sensitivity be constant from day to day if the calibration curve is to remain valid.

The use of an internal standard can minimize errors due to losses incurred during sample preparation, apparatus noise (inherent apparatus error), and systematic errors in the operator’s technique (including the volume of sample injected). In other words, the internal standard, upon addition to the sample, will be subjected to the same sequence of events that may introduce changes to the actual content of the compounds of interest and the standard. In this technique, compared to the compounds of interest in the sample, the internal standard compound must: (1) be chemically/structurally related to the compounds of interest, (2) elute at a different time, and (3) not be naturally present in the sample. Basically, the amount of each component in the sample is determined by comparing the area of that component peak to the area of the internal standard peak. However, variation in detector response between compounds of different chemical structure must be taken into account. One way to do this is by first preparing a set of standard solutions containing varying concentrations of the compound(s) of interest. Each of these solutions is made to contain a known and constant amount of the internal standard. These standard solutions are chromatographed, and peak area is measured. Ratios of peak area (compound of interest/internal standard) are calculated and plotted against concentration to obtain calibration curves such as those shown in Fig. 12.15b. A separate response curve must be plotted for each sample component to be quantified. Next, a known amount of internal standard is added to the unknown sample, and the sample is chromatographed. Peak area ratios (compound of interest/internal standard) are calculated and used to calculate the concentration of each relevant component using the appropriate calibration curve. The advantages of using internal standards are that injection volumes need not be accurately measured and the detector response need not remain constant since any change will not alter ratios.

Standards should be included during each analytical session, since detector response may vary from day to day. Analyte recovery should be checked periodically. This involves addition of a known quantity of standard to a sample (usually before extraction) and determination of how much is recovered during subsequent analysis. During routine analyses, it is highly desirable to include a control or check sample, a material of known composition. This material is analyzed parallel to unknown samples. When the concentration of analyte measured in the control falls outside an acceptable range, data from other samples analyzed during the same period should be considered suspect. Carefully analyzed food samples and other substances are available from the National Institute of Standards and Technology (formerly the National Bureau of Standards) for use in this manner.