An introduction to chromatography

---

G. M. Gaucher university of Calgary Calgary, Alberta, Canada

.

An Introduction to Chromatography

clearly for the practicing chemist and particularly for the practicing biochemist an uuderstanding of basic chromatographic principles along with a working knowledge of the various types of chromatography are indispensable. Thus the object of this presentation is to serve as a preliminary practical introduction and guide to chromatography. Historical Development ( I )

The separation technique known as chromatography was extensively used and described for the first time by Tswett (2, S), a Russian chemist-botanist in 1903. His interest in the green and yellow pigments of chloroplasts resulted in one of the first descriptions of what is now known as adsorption column chromatography. When a. petroleum ether solution is filtered through a column of adsorbent (I use mainly cdeium carbonate which is tamped firmly into a narrow glass tube), the pigments are resolved, sccording to the adsorption sequence, fmm top to bottom into various colored zones, since the more strongly adsorbed pigments displace the more weakly adsorbed ones and force them farther downward. This separation becomes practically complete when, after the pigment solution has flowed through, one passes a stream of pure solvent through the adsorbent column. Like light rays in the spectrum, the different components of s. pigment mixture, obeying a law, separate on the oalcium carbonate column and can thus be qualitatively and quantitatively determined. I call such a preparation a chromatogram, m d the corresponding method the chromatographic method. .. . I t is self-evident that the adsorption phenomena described are not restricted to the chlorophyll pigments and one must assume that all kinds of colored and colorless chemical compounds are subject to the same laws.

Perhaps because the need wae not yet great enough, this powerful new separation technique went virtually unused until 1931 when Kuhn, Winterstein, and Lederer (4) separated the isomers of such polyene pigments as carotene on a "Tswett" column. The chromatographic separation of many natural pigments followed. While such colored hydrophobic compounds were admirably separated by this technique, the need to separate colorless, hydrophilic compounds, such as the constituents of the proteins, the polysaccharides, and the nucleic acids, was increasingly evident. Thus the next chromatographic advances utilized the phenomena of ion exchange and partition. While ion exchange was first recognized as a property of aluminum silicates in soils, the widespread use of ion-exchange chromatography awaited the synthesis of the first ion-exchange resins by Adams and Holmes (6) in 1935. Then in 1941, Martin and Synge (6) developed partition column chromatography. They found that a mixture of amino acids applied to the top of a silica gel column containing definite amounts of water could be separated by passing a suitable organic solvent through

the column. Soon after this in 1944 Consden, Gordon, and Martin (7) replaced the silica gel support by strips of paper and so initiated the wide-spread use of paper chromatography. I n the same year, Craig (8) established the usefulness of the non-chromatographic multiple partition separation technique of counter current distribution by developing the first practical apparatus. A subsequent design by Craig and coworkers (9,10) is still in use today. A further advance in partition chromatography occurred in 1952 when James and Martin (11) published the first account of gas-Ziguid chromatography in which the mobile phase is a gas rather than a liquid as in paper chromatography. Unnoticed during this period of partition chromatography's rapid development was the beginning of a dramatic improvement in adsorption chromatography which involved the utilization of an adsorbent in the form of a thin layer on a glass plate. Thus the description by Stahl (12) in 1958 of an ingenious device for preparing these "open columns" initiated the rapid adoption of thin-layer chromatography. Finally the discovery by Porath and Flodin (IS) in 1959 that gels of cross-linked dextran (Sephadex) possessed the properties of an almost perfect "molecular sieve" led to the development of the unique chromatographic procedure called gel filtration which fractionates molecules such as proteins by virtue of differences in their size. Stimulated particuIar1y by the biochemist's continuing interest in separating the complex molecular components of the cell, advances in chromatographic methods will undoubtedly continue. The extraordinary advances in our understanding of the biological world a t the molecular level which have resulted from the use of chromatographic methods are innumerable and unparalleled. Adsorption a n d Pattition

Chromatography may be defined as an analytical technique (14) for separating compounds on the basis of differences in affinity for a stationary and mobile phase. These differences in affinity involve the processes of either adsorption or partition. Adsorption iuvolves the binding of a compound to the surface of a solid phase. For example, the purification of a compound by solution in a n organic solvent followed by treatment with activated charcoal is dependent upon the impurity being adsorbed preferentially onto the charcoal. Partitioa on the other hand involves the relative solubility of a compound in two phases resulting in the "partition" of that compound between the two phases. For exampIe, the extraction of an organic compound from an aqueous solution with ether is dependent upon the organic compound being dissolved preferentially in the Volume 46, Number 1 I , November 1969

/

729

ether phase. Thus, as indicated in the table, the various types of chromatography may he classified as either partition chromatography or adsorption chromatography depending upon whether the stationary phase is a liquid or a solid, respectively. As indicated the mobile phase may he either a liquid or a gas.

NUMBER OF TRbNSFERS

TUBE

NuMsia

Counter Current Distribution

As previously indicated, the possibility of developing a practical separation technique based on the partition of a solute between two immiscible solvents led in turn to the development of partition chromatography by Martin and Synge (6) and counter current distribution by Craig (8). I n each case the respective authors provided a theoretical treatment of their technique. Thus Martin and Synge (6), in analogy to the "theoretical plate" description of fractional distillation, presented one of the first theoretical treatments of column chromatography and Williamson and Craig (15) derived a mathematical treatment of counter current distrihution capable of accurately predicting their experimental results. The theoretical plate treatment of chromatography and the mathematical basis of counter current distribution have been aptly summarized by Johnson (16) and Clark (17), respectively. I n addition the more fundamental physico-chemical basis of chromatography has been discussed by Giddings (18). Because of its theoretical simplicity and mathematical predictability, counter current distribution represents an ideal model of partition chromatography. Thus an understanding of this technique leads most readily to an appreciation of basic chromatographic principles. The usual counter current distribution apparatus (19) consists of a series of interconnected glass tubes (from l(t1000) so designed that in each tube an upper (mobile) phase can he shaken with a lower (stationary) phase, allowed to separate from this lower phase, and then automatically transferred to the next tube leaving the lower phase behind. A counter current distrihution experiment begins with a solute partitioned between two phases in the first tube of a series. A measure of the relative affinity of a solute for these two phases is given by that solute's partition coefficient K where C, and Cz are the concentrations of solute in the upper and lower phases respectively. If the volumes of the two phases are equal, the partition coefficient may he redefined as K = z/y where x is the fraction of the-solute in the upper phase and y is the fraction of the solute in the lower phase such that x y = 1. Thus as illustrated in Figure 1, with zero transfers having taken place, the solute is distributed in tube 0 according to this partition coefficient. One "transfer" then consists of

+

moving this upper phase(s) to the next tube(s) containing fresh lower phase and adding fresh upper phase to the first tube, after which all contacting phases are mixed and allowed to separate

After one such transfer the fraction of solute remaining in tube 0 (i.e., y) is partitioned so that y of this fraction 730 / lournol of Chemicol Educofion

Figure 1.

A two transfer counter current distribution xheme.

(i.e., yZ) remains in the lower phase while x of this fraction (i.e., xy) goes into the upper phase. A similar partitioning of the fraction of solute transferred to tube 1yields the distrihution x2and yx in the upper and lower phases, respectively. Finally the fraction of the total solute found in each tube (both phases!) is given below the tubes (i.e., y and x) while the sum of these fractions is given to the right as its binomial equivalent (ie., (x y)'). Upon following the results of the second transfer the pattern which emerges is such that the distrihution of solute in each tube after n transfers is seen to be given by the terms of the binomial expansion of (x y)"

+

+

Thus given the partition coefficient for two compounds and the number of transfers effected, the distrihution of each compound in each tube can be accurately calculated. Such calculations readily demonstrate that for two solutes originally mixed in tube 0,their degree of separation increases the greater the difference between their respective partition coefficients and the greater the number of transfers. I n addition as an increased number of transfers tends to increase the resolution between these solute peaks, each peak hecomes shorter and wider. Partition chromatography can then he looked upon as an extension of this technique in which the contiuuous motion of a mobile phase over a stationary phase adhering to a solid support replaces the step-wise transfer of an upper phase to succeeding individual lower phases. Chromatography is superior to counter current distrihution in many instances, since with much simpler equipment a much smaller amount of material may be separated into its components with better resolution. For example, chromatography on a small 20-cm long silica gel column can yield a resolution equivalent to approximately 1000 transfers. Such chromatography columns are common and in gasliquid chromatography special capillary columns equivalent to greater than 500,000 transfers may he prepared. A greater equivalent number of transfers and therefore better separations are commonly obtained by

increasing the length of the column using slower flow rates, and thus operating closer to equilibrium conditions

The Basic Types of Chromatography

Having examined the partition of a solute between a stationary and mobile phase, a brief description of the major types of chromatography is possible. I n addition to consulting the literature cited for practical detials of theory and procedure, a number of journals and general texts are of interest (80-28). I n addition to the above mentioned references, a great wealth of information on biochemical applications of chromatography is to be found in the many multi-volume works which deal with biochemical techniques (e.g., Colowick and Kaplan (29)). I n partition column chromatography, a column is packed with a porous solid of high surface area (e.g., silica gel, cellulose) which has been coated with the liquid stationary phase (usually water). The components of a mixture are then separated by passing a mobile phase (usually an organic solvent mixture) through the column. Thus depending on the relative solubility of the components in the stationary phase, they are partitioned between the two phases and move down the column a t different rates (e.g., components which are more soluble in the stationary phase move more slowly down the column). A variety of both hydrophobic and hydrophilic compounds (from fatty acids to proteins) have been separated on such columns. In paper chromatography (SO), a paper sheet takes the place of a packed column, and the components of a mixture are partitioned between cellulose-bound water (stationary phase) and a n organic solvent mixture (mobile phase) which moves through the paper by capillary action. This method has been extensively employed in virtually all fields and has the specific advantages of simplicity in equipment and technique along with the remarkable ability to resolve very small amounts of material particularly when two dimensions are used. It is less useful for preparative or quantitative separations than is partition column chromatography. I n zone electrophoresis (31, SZ), a solid support such as paper (paper electrophoresis) or cellulose acetate strips, or starch and agar gels (gel electrophoresis) contains an electrolyte (a buffer solution) in which the components of a mixture move by virtue of their charge and the electric current passed through the electrolyte. The direction and distance moved depends on the charge on the particle (pH of buffer) and the applied field (voltage and current). Low-voltage electrophoresis (5-20 V/cm) is of general value for protein separations, while high-voltage electrophoresis (50-200 V/cm) is of particular value for separating amino acids and peptides. This technique is not chromatography hut is closely related and is particularly useful when used in conjunction with chromatography. I n gel filtration (SS), or molecular sieving a hydrophilic gel (e.g., Sephadex, a cross-linked deMran; BioGel P, a polyacrylamide; or Agarose, a non-ionic galactan from agar) in the form of porous beads contains the aqueous stationary phase which is only distinguished from the mobile phase by its immobilization within the

bead. The ability of a particular molecule to penetrate the gel particle is determined solely by the molecule's size and shape and the porosity of the particle. Thus the components of a mixture are separated according to their size by virtue of a differential distribution between easily displaceable water present in the inter-bead space (mobile phase) and water immobilized inside the beads (stationary phase). The value of gel filtration in the desalting (an alternative to dialysis), separation, purification, and molecular weight determination of biopolymers, (particularly proteins and en~ymes) has resulted in its extensive use in biochemistry. I n gas-liquid chromatography (34, S5), a column is packed with a porous inert solid (e.g., Celite or crushed firebrick) coated with a thin layer of an involatile liquid as the stationary phase. Components of a mixture are separated by being partitioned between this stationary phase and a gaseous mobile phase which is usually helium. Of greatest value in separating organic compounds, glc (vpc) is limited to those compounds which are a t least somewhat volatile. Thus biopolymers such as proteins which are not readily volatilized cannot be gas-chromatographed. However, manv essentiallv non-volatile com~oundssuch solving power of gas chromatography results from the' ability to maintain fast flow rates in long columns due to the mobile gas phase's low viscosity and ability to rapidly attain equilibrium. Gas chromatography is also favored by its ease of automation. I n adsorption column chromatography, the separation of a mixture's components is determined by the differential adsorption of these components on an "active" solid such as alumina, silica gel, or charcoal (the stationary phase) as an organic solvent (the mobile phase) containing them passes over it. Thus weakly adsorbed components will travel down the column more rapidly than strongly adsorbed ones. The tendency of components to "tail," the difficulty in preparing adsorbents of uniform activity, and the tendency of these adsorbents to catalyze chemical changes often result in partition chromatography being preferred over adsorption chromatography. I n thin layer chromatography (86, S7), rather than packing the adsorbent into a column, the adsorbent (stationary phase) is spread over a glass plate in a thin film of even thickness. The chromatographic separation then takes place by allowing a solvent (mobile phase) to move up the plate by capillary action as in paper chromatography. TLC combines the capability of adsorption chromatography in the separation of non-polar compounds with the practical simplicity and sensitivity of paper chromatography. I n addition tlc is much faster (30 rnin/l0 cm versus 4 hr/lO cm) than paper chromatography and usually yields superior resolution (spot sizes remain approximately the same as those applied to the origin as opposed to the substantial enlargement found in paper chromatography). The variety of different adsorbents available and the ease with which adsorbed water may be retained or eliminated (by drying) makes both partition and adsorption thin layer chromatography possible. Hence tlc is rapidly replacing many forms of chromatography including paper chromatography as Volume 46, Number 7 7 , November 7969

/ 731

Chromatography Type Partition chromatographya

Adsorption chromatogmphy

Types of Chromatography Based on the Stationary Phase Phases Mobile Stationary Examples Partition column chromatography liquid liquid* Paper chromatography Thin layer chromatography (tlc) Gel filtration (Sephadex chromatograqhy) gas liquid Gas-hqmd chromatography (glc) or Vapor phase chromatography (TO) s liquid solid Adsorption column chromatography Thin layer chromatography electrolyte ionic polymer Ion exchange chromatography solution

The technique of electrophoresis often uses the physical set-ups of liquid-liquid partiti~nchromatography but is not chromatography as such. Counter current distribution represents a. ml~themsticsllypredictable model of partition chromatography.

the method of choice for analytical separations and some preparative separations. I n ibn exchange chromatography (38, .99), ionized compounds in aqueous solution (mobile phase) are separated by virtue of their differences in affinity for ionized groups which are a n integral part of a n insoluble solid phase (stationary phase). These ionic solids are usually synthetic polystyrene resins in bead form which contain either acidic groups (cation exchangers) or basic groups (anion exchangers). Thus a mixture of cations in aqueous solution applied to the top of a column packed with a cation exchange resin in the hydrogen ion form will displace the hydrogen ions and form a narrow adsorbed band a t the top of the column. Upon adding a solution of some other appropriate cation, the cations to be separated will be displaced, moving down the column a t different rates according to their relative affinity for the resin. Because the exchange capacity and therefore the effectiveness of synthetic ion exchange resins is dependent upon penetration of the resin bead by the ionic solutes to be separated, the pore size of these beads determines the size of the solute which can be effectively separated. Thus while available ion exchange resins are of great value in separating small molecules such as amino acids, they exclude large macromolecules such as proteins. However cellulose and dextran (Sephadex) ion exchangers e . , DEAE (diethyl amino ethyl) cellulose and DEAE Sephadex) &re commonly used to s e ~ a r a t esuch macromolecules on the basis of ionic affinities. Having classified and described the various types of chromatography it is noteworthy that the simple divi-

Figure 2.

732

/

Types of chmmotogrophy based on the physic01 set-up.

Journal of Chemical Education

sion of all types into either partition or adsorption chromatography as in the table is not rigid. Virtually all the types of chromatography mentioned exhibit a combination of physical effects. Thus ion exchange resins exhibit molecular sieve properties, while dextran gels exhibit both ion exchange and adsorption properties. Both tlc and adsorption column chromatography require some adsorbed water for optimum adsorption chromatography. And paper chromatography separations can be achieved using water as both the mobile and stationary phases! This situation exemplifies the still very empirical nature of chromatography and the fact that experience rather than theory is still of greatest value. Chromalographic Methodology

It is evident from the various types of chromatography described, that chromatographic methodology or in other words the physical set-up used, may vary considerably. This is indicated in order of increasing complexity in Figure 2. The resolution, simplicity, and rapidity of any separation is often dependent upon this physical set-up. Thus while "complete" chromatography is simplest and most rapid, "flowing" or gradient elution chromatography is capable of greater resolution. This is because flowing chromatography effectively increases the length of a chromatographic set-up, while gradient elution chromatography decreases peak broadening and therefore increases resolution. While non-equilibrium conditions (caused for example by high flow rates) may result in "tailing" and "fronting," i t is the existence of non-linear partition or adsorption isotherms which are the major cause of such peak broadening effects. Thus the amount of a substance taken up by the stationary phase depends upon this substance's concentration in the mobile phase. A plot of "amount in stationary phase" (A,) versus "concentration in the mobile phase" (C,) a t constant temperature illustrates this relationship. As illustrated in Figure 3 these adsorption or partition isotherms determine whether a ~ h r o m a t i g r a ~ hpeak y is symmetrical or exhibits "tailing" or "fronting." For example in case B, as the band moves down the column, the coucentratiou of the compound in the mobile phase decreases a t the back of the band while increasing a t the front. Simultaneously the amount of compound held by the stationary phase increases a t the back and decreases a t the front, resulting in "tailing."

Figure 3. T h e effed of adsorption or portition isotherms on peak rhope (281.

(42). Nore complex gradients of virtually any shape may be produced using the gradient device described by Peterson and Sober (44) and marketed by Buchler Instruments, Inc., Fort Lee, New Jersey. Finally it should be noted that in the method described above the solutions in the reservoir and in the mixing chamber should not differ in density by more than 5y0. However a modified method of value for adsorption column chromatography, where for example a methanol (density = 0.79): chloroform (density = 1.49) gradient is used, has been described by Wren (45). Literature Cited

Tailing is almost inevitably found to some extent in adsorption chromatography. Symmetrical peaks are more common in partition chromatography. Fronting is relatively rare. Since the phenomenon of tailing is often the cause of Door resolution (LO). an invaluable techniaue first described by ~ l m , ' ~ i l l i a mand s , Tiselius (41) is that of gradient elution. This technique involves the use of an eluant solution in which the concentration of the most powerful eluting agent gradually increases, thus forming a concentration gradient down the column. The tail or rear portion of a peak is therefore always in contact with a stronger eluant solution than is the front of the peak. The tail therefore moves more rapidly than the peak itself and is eliminated. A variety of continuous elution gradients may be produced by using two eluant containers of different size and/or shape (43,43). As indicated in Figure 4,

(1) ZmnMersTsn. L., "Progress i n Chr~matot~graphy 1938-1947," Chapman and Hall, London, 1960. (2) R o a l ~ a o aT.. , J. CXEM.EDUO.,36. 144 (1959). (3) STRAIN, H. H., AND SRERMA. I.,J. CXEU.EDUC..44, 235 (1967). (4) K u x s . R.. WINTERSTEIN.A.. A N D LEDERER.E., Hopp6Seglds 2. Physiol. Cham.. 197, 141 (1931). (5) A o m a , B. A., A N D HOIIMEB, E. L., J . SDC.Chem. Ind. (London), 64,

- \.""",.

1 IIOPF,

(6) M A ~ T I NA., J. F., A N D R r s a ~R. , L. M., Biochcm. J . , 35, 1358 (1941). (7) Coasosn. R., Gonnox, A. H.. AND MARTIN.A. J. P.. Biochcm. J . , 38, 224 (1944). (8) Cn*la, L. C., J.Biol. Chcm.,166,519 (1944). (9) C n ~ oL., C.. A N D POBT,0.. Anal. Cham., 91,500 (1949). CMO. L. n.. H A U ~ M A W N . w.. A ~ R E N ~P.. . H ~ R P B N ~E. ~ Ta,. . .

A simple rot-up for producing e l v ~ n gradients. t

variations in the size of cylindrical containers (43) is a simple method of obtaining eluant gradients. Since the two containers are open to the atmosphere the liquid levels in each drop a t the same rate. Thus as indicated in Figure 4 the rate of solution addition to the column (Ra) is equal to twice the rate of solution transfer from container a to b (R.) for a linear gradient. If the crosssectional areas ( A , A ) , the initial concentrations (C&Ca)and the volnmes (V = T/. Va) are known the concentration of the solution produced (c) after any given volume (v) has been added to the column, may be readily calculated using the following equation

+

I n general linear and concave gradients have been found to yield the best chromatographic separations

.

.

PORATHJ., AND PLODIN,P.. Nature. 183, 1657 (1959). C ~ s s m rH. . G.. J. CXEM.EDUC.,38, 482 (1956). WILIIIAMBON, B., A N D CRAIQ,L. C.. J. Bid. Cham., 168,687 (1947). JOHNBON, M. J., "Design of Chiomhtographic Procedurss" i n U ~ n n ~ l r . W. W.. B u n n ~ s , RH..AND . STAUFER, J . F.,"Manometric Teohniques." (4th Ed.), Burgess Puhliahinp Co., Minneapolis, 1964, p. 233. (17) C u n x , J. M. Jn., (Edilov), "Experimental Biochemistry;' W. H. Freeman and Co.. Sen Francisco, 1964. D. 190. (18) GIDDINGB, J. c..J. CREM.EDVC.,44,704 i1967). (19) C n ~ o L. , C.. "Counter Current Distribution" i n FLoRrnr, M., A N D Slmrz. E. H. IEdilor8l. "Comorehensive Biochemistrv." " . American ~ i & i k r . New 1969, vol. k , 1. (20) . . Anal. Cham.. Amerioan Chemied Soeietv. Washinston. . . various reviews each April. (21) HEFTMANN,E. (Editor), "Chrom~tography;' (2nd Ed.), Reinhold. Nera York, 1967. (22) Jmms. A. T.. nlro Monnrs, L. J., "New Biochemical Separations." Van Nostrand. London. 1964. (23) LEoenEn. M. (Ediloi), J . Chiomatour., Elsevier. Amsterdam. (24) L m ~ n e n M. , (Editor). Chromatour. Em.. Elsevier. Amsterdam. (25) L m m m , E., A N D LEDEILEII, M., i n F ~ o n r r M.. ~ , A N D STOTZ, E. H.. (Editorsi. "Comorehensive Bioohemistrv." . . American Elsevier. New ~ o r k1669, , vo1.b.p. 32. (26) MmeZ. 0.. (Edilor). "Laboratory Handbook of Chromatogrsphic Methods," Van Nostrand. London. 1966. (27) M o n m s , C. J. 0. R., A N D M o n ~ x s P., , ''Separation Methods in Bioohemistry." Pitman and Sons. London. 1968. (28) S ~ a o a ,R., AND R ~ c E C, , B. F., "Chromatographi Methods" (2nd Ed.), Chapman and Hall, London. 1967. (29) C o ~ o w r c x .S. P., nno KAPLAN,N. 0. (Editon). "Methods in Enzymology." Academic Press. Nerv York. 1 9 5 5 6 9 Vol. 1-16. (30) H A I ~I., M.. A N D M ~ c n n .I