Multidimensional Planar Chromatography Szabolcs Nyiredy, Research Institute for Medicinal Plants, Budakalász, Hungary.
Modern planar chromatographic (PC) techniques offer high separation power, thanks to two-dimensional (2D) planar arrangements, chromatoplates with different properties, a variety of solvents combinations, various forced-flow techniques and multiple development modes. By combining these possibilities, multidimensional planar chromatography (MD-PC) can be performed in different ways. This article will discuss the potential of various MD-PC techniques, including comprehensive 2D PC (PCPC), targeted or selective 2D PC (PCPC), modulated 2D PC (nPC) and coupled-layer PC (PC–PC).
Introduction In planar chromatography (PC) the solvent system (mobile phase) moves through the stationary phase by capillary action or under the influence of forced flow in a planar arrangement. Relatively simple PC problems can be solved using capillary action on either chromatographic paper, thin layer chromatography (TLC) plates or high performance (HP) TLC plates. For more complicated separations the use of forced-flow techniques is necessary, because they permit optimal mobilephase velocity to be exploited over the whole separation distance without loss of resolution. Forced flow can be achieved either by application of an external pressure, an electric field or by centrifugal force. One-dimensional PC systems, however, often have inadequate power for satisfactory resolution of compounds present in complex samples, a failure that becomes increasingly pronounced as the number of compounds increases. For such separations multidimensional (MD) techniques are very important.1 MD-PC exploits combinations of different separation mechanisms or systems; such methods can generally be developed by combining almost any of the different chromatographic mechanisms or phases (stationary and/or mobile), electrophoretic techniques and field-flow fractionation sub-techniques.2,3 The correct definition of MD chromatography — according to Giddings4,5 — includes two criteria. “First, it is one in which the components of a mixture are subjected to two or more separation steps in which their displacements depend on different factors. The second criterion is that when two components are substantially separated in any single step, they always remain separated until the completion of the separative operation.” 2
RECENT APPLICATIONS IN MULTIDIMENSIONAL CHROMATOGRAPHY December 2003
Nyiredy
This latter condition, therefore, precludes simple tandem arrangements in which compounds separated in the first separation system can re-merge in the second. Many papers exist that describe the different MD chromatographic techniques (e.g., references 6–10). Very recently, Schoenmakers et al.11 published a paper on the nomenclature and conventions in comprehensive MD chromatography. Being consequent to those conventions, the most frequently used MD-PC techniques are listed and discussed below: Comprehensive two-dimensional PC (PCPC): An MD-PC technique, using the same monolayer stationary phase and two developments with different solvent systems (mobile phases) characterized by different total solvent strength (ST) and selectivity values (SV), or using a bilayer stationary phase and two developments with the same or different solvent systems (mobile phases). Targeted or selective two-dimensional PC (PCPC): An MD-PC technique, in which following the first development from the stationary phase a heart-cut spot (scraped off a TLC/HPTLC plate or cut out using paper chromatography) is applied to a second (same or different) stationary phase for subsequent development of the transferred compounds with the same or different solvent system/mobile-phase composition, respectively. Modulated two-dimensional PC (nPC): An MD-PC technique, using the same stationary phase (in one, two or three geometrical dimensions) using solvent systems/mobile phases with decreasing solvent strength and different selectivity values. Coupled-layer PC (PC–PC): An MD-PC technique, using coupled layers with orthogonal stationary phases (in which completely independent retention mechanisms ensure the separation) developed with a solvent system/mobile phase of constant composition. Combined MD-PC methods (cMD–PC): A combination of at least two of the above mentioned modes, or coupling of two chromatographic techniques in which PC is used as the second dimension and another separation method (e.g., gas chromatography, high performance liquid chromatography, etc.), as the first. The goal of this paper is not to present an overview of all relevant applications, but rather to stress the potential of the various multidimensional possibilities in PC. In the following the terms solvent system and mobile phase are used
interchangeably, however, for PC using capillary action, ‘solvent system’ is the correct term; while for forced-flow PC techniques ‘mobile phase’ is correct.
Comprehensive 2D PC (PCPC) The first (comprehensive) 2D separations were achieved using paper chromatography by Martin’s group12 in 1944. Kirchner et al.13 introduced 2D TLC in the early 1950s before it was put on a firm footing by Stahl.14 The term comprehensive 2D chromatography (PCPC) covers chromatographic development in one direction followed by a second development in a perpendicular direction to the first. The method consists of spotting a sample at the corner of a chromatoplate [Figure 1(a)] and enabling migration of the solvent system (characterized by ST1; SV1) in the first direction [Figure 1(b)]. For all substances not completely separated in the first development or primary “column” [see Figure 1(b)], baseline separation can be achieved by means of a second separation process with an appropriate mobile (stationary) phase. Therefore, after drying, sequential development of the plate (in a direction at right angles to the first development) can be started with a second solvent system (characterized by ST2; SV2). Figure 1(c) shows that in the second dimension a theoretically unlimited number of secondary “columns” can be applied. The overall separation obtained is superior if the interactive forces which bring about retention are different for the two consecutive developments through the use of solvent systems with different composition. Theoretically, the best separation can be obtained when the spots are spread over the total area of the chromatoplate [Figure 1(d)]. PCPC can also be achieved using a bilayer stationary phase and two developments with the same or different solvent systems (mobile phases). In this instance the primary “column” is the minor part (4 20 cm) of the bilayer, and the secondary “columns” are the major part (16 20 cm) of the bilayer. Unfortunately the use of RP-18 and silica as the bilayer is rather complicated, because the solvent used in the first Figure 2: Schematic illustration of the peak capacity (n2) of a comprehensive two-dimensional PC (PCPC) separation, where the number of squares represents the number of compounds which can theoretically be separated. y
'n' Compounds
Figure 1: Schematic illustration of a comprehensive two-dimensional PC (PCPC) separation, where the number of ’secondary columns’ is unlimited, demonstrating the multidimensionality of PC. (a)
(b)
(c)
(d)
ST2;SV2
1-front
2-front
2.
Primary 'column'
www.lcgceurope.com
2. Development
ST1;SV1
n secondary 'column'
'n' Compounds
1. Development
Primary 'column' 1.
x
3
Nyiredy
development can modify the other (major) part of the bilayer (the second stationary phase), which can result in poor reproducibility of the separation system. Therefore, the use of two single coupled plates instead of bilayers is suggested (see coupled layer PC). A PCPC separation can be seen as one-dimensional displacement operating in two geometrical dimensions. A combination of two selective displacements in PCPC will lead to the application of different separating mechanisms in each direction. As an extreme, if the solvent combinations are the same or very similar (ST1 = ST2; SV1 = SV2 or ST1 ~ ST2; SV1 ~ SV2, respectively), the compounds to be separated will be poorly resolved or unresolved, and a diagonal pattern will be obtained. In such circumstances a slight increase might occur in resolution because of an increase (by a factor of √2) in the migration distance of the zone. In this instance, the term comprehensive 2D separation should not be used. The point at which the sample is spotted can be regarded as the origin of a coordinate system.9 The process of development is performed in two steps; the first in the direction of the x axis to a migration distance Lx. After evaporation of the solvents used, the second development will be performed in the direction of the y axis to a migration distance Ly. The positions of the compounds after development in the x direction depend on the ST and SV values of the first solvent system applied. Similarly, the migration distances of the individual compounds also depend on the total solvent strength and total selectivity of the second solvent combination. After development in the x direction, the migration distances in y direction for all compounds are zero. For development in the y direction, their migration distances in x direction are determined by the first development. The final positions of the spots are determined by the coordinates x(i) and y(i), which can be expressed as follows: RFxy(i) = (RFx(i), RFy(i))
[1]
The principle of PCPC separation is illustrated schematically in Figure 2. The multiplicative law for 2D peak capacity Figure 3: Schematic illustration of the steps of a targeted or selective two-dimensional PC (PCPC) separation. (a) Schematic chromatogram and densitogram of a separation after the first development with the first solvent system; (b) Heartcutting (scraped off) of the targeted spots from the first layer; (c) The heartcut spot is applied to the second stationary phase for subsequent analysis of the transferred components; (d) Schematic chromatogram and densitogram of a separation after the second development with the second solvent system. (a) (b) ST1;SV1 1-front
(c)
(d) ST2;SV2 2-front
emphasizes the tremendous increase in resolving power which can be achieved; in theory this method has a separation capacity of n2, where n is the one-dimensional peak capacity. To achieve this peak capacity the selectivity of the solvent systems used in the two different directions must be complementary. In practice, the peak capacity in PCPC is less than the product of the two one-dimensional developments. The sizes of the spots of the compounds being separated are always larger in the second development than was the initial sample spot. Consequently, in the second development a higher resolution is needed to obtain a sufficient resolution in the second direction as well. The most efficient PCPC system involves contrasting solvent systems. Although computer simulation has been used to optimize PCPC separations based on one-dimensional data,14 the search for appropriate solvent systems for normalphase PCPC separations is a simple and rapid method. Härmälä et al.15 suggested starting solvent selection according to the first part of the ‘PRISMA’ system for PC,16,17 with a selection of 10 solvents chosen from the eight Snyder18 selectivity groups, which can be evaluated in parallel in unsaturated chromatographic chambers. The compounds to be separated should be applied to the chromatoplates in groups of three or four, for ease of identification. The solvent strength of each neat solvent is adjusted individually by using n-hexane, so that the zones of the compounds to be separated are distributed in the RF range 0.2–0.8. The exact positions of the spots can be located densitometrically. If solvents afford good separation, their homologues or other solvents from the same group can also be tested. The RF values of the substances to be separated are obtained from the TLC runs using the most promising binary solvents and must be compared with each other in a correlation matrix. The next step is validation of the solvent systems by regression analysis; the solvents with the poorest correlation values must be selected. After choice of an appropriate pair of solvents the PC technique [e.g., HPTLC or OPLC (overpressured layer chromatography)] and other separation conditions must be selected, according to the third part of the ‘PRISMA’ system. In this way an excellent separation can be achieved for closely related compounds in the minimum time, generally within a few hours. As was discussed, in comprehensive 2D PC (PCPC) the fractions are not always transferred to another separation system. Rather, a secondary separation is developed orthogonally on the same (mono or bilayer) chromatoplate. Because of this the terminology ‘2D PC’ is not sufficiently expressive; instead of the term comprehensive 2D PC (PCPC) should be used as the most powerful type of MD-PC. PCPC is highly important in its own right, because this is the first MD-PC separation method in which all compounds can be passed to a next dimension; it therefore serves as the reference system with which all other multidimensional systems (GCGC, LCLC, etc.) can be compared.
Targeted or Selective 2D PC (PCPC) The goal of the selective or targeted mode of MD-PC is to separate small groups of compounds from a complex sample. Target analysis by 2D PC is normally performed by using the principle of “heartcutting”. In 2D PC “heartcutting” can be achieved in two different ways. In the first approach — after the first development with an ST1; SV1 solvent system [see Figure 3(a)] — from the first layer
4
RECENT APPLICATIONS IN MULTIDIMENSIONAL CHROMATOGRAPHY December 2003
Nyiredy
a scraped-off spot [heartcut, Figure 3(b)] is applied to a second stationary phase for subsequent analysis [Figure 3(c)]. Thus, the first stationary phase is used to extract and enrich analytes (e.g., from complex biological matrices, such as plant extracts), while the second layer is used to separate the compounds of interest. In this manner the target compound(s) is(are) preconcentrated and/or can be detected in the absence of interfering substances. As a result, enhanced selectivity, sensitivity and lower detection limits can be obtained and densitometric evaluation is highly improved [demonstrated in Figure 3(d)]. A possible symbol for this selective 2D PC is PCPC because these are two consecutive steps. The second “heartcutting” mode can be performed in situ on the same chromatoplate. Therefore, after the first development is finished and the chromatoplate dried [Figure 4(a)] two lines must be scraped into the stationary phase perpendicular to the first development, in such a way that the target compound(s) are between these lines. A certain portion of the stationary phase must be removed [hatched lines in Figure 4(b)], to ensure that the solvent system of the second development affects the target fraction of the sample only [Figure 4(c)]. If, in a complex sample, more than one target fraction is present, these selective 2D separations can be repeated for other parts of the sample with another, more adequate solvent system (third development) as Figure 4: Schematic illustration of the steps of in–situ targeted or selective two-dimensional PC (PCPC) separation. (a) The dried chromatoplate after the first separation was finished using the first solvent system; (b) The prepared chromatoplate for the separation of first targeted fraction between the scraped lines (scraped lines and hatched lines indicate the removed part of the stationary phase); (c) The dried chromatoplate after the second separation was finished using the second solvent system; (d) The prepared chromatoplate for the separation of second targeted fraction between the longer scraped lines (scraped lines and hatched lines indicate the removed part of the stationary phase); (e) The dried chromatoplate after the third separation was finished using the third solvent system. (a)
(b)
1. Development
ST1;SV1
it is demonstrated in Figure 4(d) and (e). The suggested general term for this technique is: PCPCPC. If capillary action is used for the separation, TLCTLCTLC; if forced-flow is used (e.g., OPLC), then the term OPLCOPLCOPLC could be used, as the separation is not comprehensive, not on-line, but selective for certain parts of a complex matrix. Based on our experiences (working with capillary action) on TLC plates, three target fractions can be separated, while on HPTLC plates, two target fractions can be separated with in situ “heartcutting” MD-PC operating mode, because of the decreasing solvent front velocity. Using OPLC, several target fractions can be analysed, however, the handling is complicated because the layer must be prepared for all selective separations.
Modulated 2D PC (nPC) A special feature of PC is multiple development, in which development is repeated two or more times with the same or different solvent systems. Between consecutive developments the stationary phase must be dried. From the points of view of development distance and solvent-system composition, multiple development techniques can be classified into different categories.1 The most powerful multiple development technique is bivariate multiple development (BMD) in which the development distance and the solvent-system composition vary simultaneously during successive chromatographic runs. Using this technique, developments are performed in the same direction with a stepwise negative gradient that becomes progressively weaker (decreasing ST value) over distances that increase by 1–5 mm for each stage. In contrast to the situation in reversed-phase HPLC, the gradient starts with the most polar solvent system, for which the shortest developing distance is employed, which is progressively decreased. The longest migration distance is used with the most non-polar solvent system. BMD is typically fully off-line, because the Figure 5: Schematic illustration of modulated two-dimensional PC (nPC), based on negative solvent strength gradient and zone refocusing mechanism. (a) The applied sample before starting the separation; (b) The solvent front of the first solvent system reaches the first stage (n 1), where the zone starts to modulate (refocus); (c) The solvent front of the second solvent system reaches the second stage (n 2), where all compounds together start to migrate (ST2 < ST1); (d) The solvent front of the third solvent system reaches the third stage (n 3), where a part of the compound is separated as narrow bands (ST3 < ST2). (a)
(c)
(d) 2. Development
ST2;SV2
www.lcgceurope.com
(b) ST1;SV1
(e) 3. Development
(c) ST2;SV2
(d) ST3;SV3
n
n
n
n
3
3
3
3
2
2
2
2
1
1
1
1
ST3;SV3
5
Nyiredy
chromatoplate must be dried between consecutive development steps. Figure 5 shows a schematic of modulated 2D PC (nPC) on a monolayer stationary phase, in which the composition of the solvent sytem, characterized by the total solvent strength (ST) and total selectivity value (SV), is changed for each of three development steps and in which the development distance increases linearly. As it is shown, after the sample application [Figure 5(a)], between the first [Figure 5(b)], second [Figure 5(c)] and third developments [Figure 5(d)], migration distance changes, as does the total ST and SV of the solvent system. Figure 5(d) illustrates the principle of AMD (automated multiple development) — developed by Burger19 — employing a negative solvent strength gradient (decreasing ST values). The number of re-chromatography steps, the development distance, and the total solvent strength and/or selectivity value of the solvent system can be freely varied, depending on the separation problems. The efficiency of nPC is partly a consequence of the zone refocusing (modulating, see later) mechanism, as depicted in Figure 5(b). Each time the solvent front traverses the stationary sample in multiple development, it compresses the zone in the direction of development. The compression occurs because the solvent system first contacts the bottom edge of the zone, where the sample molecules start to move forward before those molecules still ahead of the solvent front. When the solvent front moves beyond the front edge of the zone the refocused zone starts to migrate and is broadened by diffusion in the normal way. By use of optimum conditions a balance between zone refocusing and zone broadening can be achieved. Figure 6 illustrates the principle of AMD employing a Figure 6: Universal used gradients during the stages of AMD separations. (a) Change in the solvent system combination; (b) Change of total solvent strength (ST) of the solvent system combinations; (c) Change of total selectivity value (SV) of the solvent system combination.
negative solvent strength gradient (decreasing ST values). In the first step the compounds to be separated were moved and modulated (re-concentrated). In subsequent steps both values characterizing the solvent system were changed and, as a consequence, the more non-polar compounds migrate near the -front. By correct selection of the total solvent strength and total selectivity value all the compounds could be separated. AMD under controlled conditions results in high reproducibility, and the convenient facility for vacuum drying of the chromatoplate and the use of a nitrogen atmosphere reduces the chance of degradation during multiple development. Unfortunately, separations are slow because of the large number of development and intermediate drying steps. However, the increase in zone capacity is significant. The procedure generally used in nPC starts on normal-phase (NP) stationary phase with a strong solvent (100% methanol; si = 5.1, sv = 2.18). The methanol concentration is reduced in 15 steps to 0%, while the amount of solvent B (diethyl ether or dichloromethane) increases from 0 to 100%. From the 15th to the 25th steps the concentration of solvent B is reduced and the concentration of n-hexane is increased to 100% [Figure 6(a)]. Figure 6(b) shows that if dichloromethane (si 3.1, sv 1.61) or diethyl ether (si 2.8, sv 4.08) is used, the change in solvent strength is not very large. However, the selectivity value [Figure 6(c)] decreases when dichloromethane is used and increases when diethyl ether is applied. It can be concluded that selectivity values must also be considered in the search for suitable solvents in nPC. Based on the arguments above, use of the term modulated 2D PC (nPC) is suggesed, because both the solvent system composition varies and sample refocusing will occur.
Coupled-Layer PC (PC–PC) Coupled layer systems can be divided into two different approaches; namely when a stationary phase is mainly used as Figure 7: Schematic diagram of cross section of coupled layer PC (PCPC) to ensure MD separation on stationary phases of decreasing polarity. (Dark shading, stationary phase (a); light shading, stationary phase (b); black zone, sample).
(a) Me
tha
nol
ane
eth
m loro
(b) ST
5
10
xan
thy
h
Dic
e
or D ie
He
let
15
he
r
20
Steps SV
Methanol-Dichloromethane-Hexane
5.10
Spring
ST
3.10
2.18 1.61
SV (c) ST
5
10
15
20
25
Steps SV
Methanol-Diethyl ether-Hexane
5.10
SV
ST
4.08
2.80
Supporting glass plates Stationary phase (b) Stationary phase (a)
2.18 5
10
15 Steps
6
Distance holder
25
20
25
Sample
RECENT APPLICATIONS IN MULTIDIMENSIONAL CHROMATOGRAPHY December 2003
Nyiredy
the first layer for fast sample enrichment and/or a clean-up procedure, and a PC–PC coupling employing two separation layers operating in orthogonal mode, which provides a highly enhanced spot or peak capacity relative to two one-dimensional techniques. From the point of view of the sample this MD-PC type is an on-line operating mode, therefore the suggested term is PC–PC. Chromatoplates can be coupled irrespective of whether capillary action or forced-flow is the driving force for the separation. The first technique is denoted “grafted” planar chromatography,20 and the second, long distance (LD) OPLC.21,22 The idea of coupled (grafted) TLC plates was first published in 1979. In this technique two plates with different stationary phases are grafted together and clamped in the fashion of a lap joint with the edges of their adsorbent layers in intimate contact. In this way a compound from a chromatogram developed on the first chromatoplate can be transferred to the second plate without the usual scraping of bands, extraction and re-spotting. Nowadays almost all commercially available HPLC stationary phases are also applicable to PC,1 which enables the use of coupled layers for MD-PC separations. A simple method for coupled (serially connected) TLC with different stationary phases is illustrated in Figure 7, in which the two chromatoplates are turned face to face and pressed (grafted) together such that certain regions of the layers overlap. The configuration illustrated in Figure 7 is an ultra-micro chamber in which the vapour phase is practically unsaturated.35 The distance between the supporting glass plates and the stationary phases are twice the layer thickness, which depends on the type of analytical (10–25 µm) chromatoplate applied. The sample to be separated (black zone in Figure 7) is applied to stationary phase A. The two chromatoplates must be clamped in lap-joint fashion with the edges of their stationary phases in close contact (Figure 7) so that compounds from the first chromatoplate can be transferred to the second plate. This arrangement of the coupled layers can be regarded as MD-PC if all compounds — finishing the separation — are located on the second stationary phase (B), and if the second stationary phase is much less polar than the first (A). Otherwise the second criterion of multidimensionality is not fulfilled. Working with the same solvent-system composition, the serial connection of stationary phases under capillary-controlled flow conditions is much more complicated, for two reasons: first, for HPTLC plates the mobile phase velocity might no longer be adequate to maintain optimum separation conditions; and second, the handling of more than three chromatoplates within a separation distance of 7 cm is practically impossible. On this basis, orthogonal separations can be performed if, for example, RP and NP stationary phases are coupled, in such a way that 20 cm separation distances are ensured on both stationary phases. The sample to be separated can be applied to the first, 2.5 20 cm RP plate. After development with the appropriate solvent system, the first (RP) chromatoplate must be dried. The second, 20 20 cm silica plate must be clamped to the first (RP) plate in such a way, by use of a strong solvent system, that the separated compounds can be transferred to the second chromatoplate. The re-concentrated samples are ready for development in the second geometrical direction. By use of a second solvent system with appropriate selectivity an effective www.lcgceurope.com
orthogonal MD-PC separation can be observed, irrespective of whether capillary action or forced-flow techniques are applied. Because 10 20 cm glass-backed plates coated with chemically modified layers are available and can be cut to 2.5–5 20 cm, many stationary-phase combinations can be realized by the coupled technique. At the beginning of the 1990s Botz et al.21,22 proposed a novel OPLC technique with significantly increased separation efficiency, in which the separation distance can be increased as a result of a special arrangement of chromatoplates. This category of multilayer forced-flow PC involves serial connection of the chromatoplates with different (favourably with decreasing) polarity during a single development [Figure 8(a)]. Specially prepared plates are necessary for this linear separation technique.1 Therefore, all four edges of the chromatoplates must be impregnated with a special polymer suspension. Three plates can be placed on top of each other to create the LD for the separation. The end of the first (top) and the second (middle) chromatoplate has a slit-like perforation to enable transfer of the solvent system to the second and third layer. On this basis 60 cm separation distances can be achieved. As a consequence of the manner in which the layers are prepared, glass-backed plates can only be used at the bottom of the stack. Movement of the solvent system (for off-line separation) or mobile phase (for on-line separation) with a linear front can be ensured by placing a narrow plastic sheet with a solvent system/mobile phase inlet channel on the top layer, or by scraping out a channel from the top layer, as it is shown in Figure 7. The slit (width ~ 0.1 mm) can be produced by cutting the layers with a sharp blade; this enables easy passage of solvent system/mobile phase and separated compounds without any mixing. The cushion of the OPLC instrument is applied to the uppermost layer only, and each plate presses onto the sorbent layer below. Figure 8: Schematic illustration of coupled layer OPLC (OPLC–OPLC) to ensure MD separation on three stationary phases of decreasing polarity. (a) Principle of serially connected OPLC using three chromatoplates; (b) OPLC–OPLC using the fully off–line operating mode; (c) OPLC–OPLC using the fully on–line operating mode (thin lines, the mobile phase inlets and outlet; thick lines, slit-like perforations to enable transfer of the mobile phase between two chromatoplates). (a) Stationary phase (a)
(b)
Stationary phase (b)
Stationary phase (c)
(c)
Stationary phase (a)
Stationary phase (b) Stationary phase (c)
Stationary phase (a)
Stationary phase (b) Stationary phase (c)
7
Nyiredy
Several samples can be applied for the fully off-line mode [Figure 8(b)], in which the separation is complete when the -front of the solvent system reaches the end of the bottom chromatoplate, which is prepared with the polymer suspension on only three sides. After drying the plates, the separations can be evaulated by a densitometer. Figure 8(c) illustrates the fully on-line operating mode of coupled layers, in which all the compounds of the single sample to be separated must be moved at the same separation distance. Therefore, the mobile phase can be directed from the lower plate in a manner similar to that in which it is introduced. This gives the possibility of on-line detection, using a through-flow DAD detector. For this fully on-line mode of operation the length of all layers placed between the uppermost and lowest plates must be reduced by ~1 cm, to ensure room for the mobile phase outlet. It has to be noted, that satisfactory combination of more than three chromatoplates for LD-OPLC is always difficult because only the bottom plate can be glass-backed (the others must be aluminium-backed21). The potential of serial connected OPLC can be used for MD-PC separations on different stationary phases of decreasing polarity for all compounds that are located on the bottom chromatoplate (fully off-line mode), or are eluted from the third layer (fully on-line mode).
Combination of MD-PC Methods A theoretical model whereby maximum peak capacity could be achieved using a 3D PC separation was proposed by Beaver and Guiochon.23 However, this idea could not be realized in practice because of technical problems. To realize PCPCPC, several chromatographic papers or EmporeTM layers24 (silica entrapped in an inert matrix of polytetrafluoroethylene microfibrils) can be carefully stacked on top of each other, with the developed 2D sheet (paper or layer) resting on the bottom. After the sheets have been pressed together, the separation can be started into the third geometrical dimension by introducing the solvent system through an appropriate porous sintered glass plate. If the number of chromatographic papers or EmporeTM sheets is sufficient, a separation cube can theoretically be constructed, enabling a spot capacity of n3. After separation in the third geometrical dimension, the separated compounds would be found on different sheets. However, it might also happen that the same spot could be found on more than one sheet. The separation cube represents the PCPCPC separation. It must be noted that the separating power of EmporeTM layers is only 60% of that of conventional TLC;24 this has been attributed to very slow solvent migration velocity resulting from capillary action. Unfortunately the separation efficiency of chromatographic papers has not been increased in recent decades, although hundreds of laboratories still use paper chromatography in their daily routines. In the introduction of this paper MD-PC was defined as a procedure in which substances to be separated were subjected to at least two separation steps with different retention mechanisms. Because of the versatility which resulted from the combination of solvent systems/mobile phases of different composition, more than two development steps can easily be realized using nPC techniques. A powerful possibility is the parallel combination of stationary and mobile phases (solvent systems) and multiple development techniques, in which total solvent strength and 8
mobile phase selectivity are changed simultaneously. This is an extended version of MD-PC and can be regarded as double MD-PC, because it changes not only the stationary phases (bilayer), but also the solvent-system composition to ensure the criteria of MD-PC. It is apparent that the number of compounds separated is more than for any other version discussed. Another way to increase separation power might be obtained using multiple development for both orthogonal developments. Unfortunately, this seems to have been rarely recognized. In this combination, after two or more developments during which the total solvent strength can be reduced stepwise at constant selectivity, the compounds can be separated according to differences in polarity. In the second geometrical dimension, perpendicular to the first, the chromatoplate can be re-developed several times in such a way that the selectivity would be changed at constant total solvent strength to achieve maximum spot capacity. This nPCnPC technique is a promising route to real improvements of planar chromatographic separation power. Further possibilities to significantly increase the separation efficiency of MD-PC include the combination of stationary phases, multiple development technique and forced-flow technique. Separation in the first dimension can be performed by application of the first stationary phase and mobile phase combinations using BMD (ST1; SV1 ➝ STn; SVn) technique. After drying of the chromatoplate the separation can be continued in a perpendicular direction by using three linear OPLC prepared chromatoplates, with different characteristics (decreasing polarity), in combination with another mobile phase. It must be pointed out that the separating power can easily be increased using BMD in the second dimension also. Increasing the dimensions of the separation procedure always increases the number of theoretical plates, which leads to enhanced resolution and, therefore, the efficiency of the separation. Several other possibilities exist to increase the number of dimensions. Between the first and second developments, the characteristics of the chromatoplate or the properties of the sample can also be modified. Although interfacing of on-line OPLC with one- or two-dimensional TLC is not particularly difficult, it is not yet widely practised. It must be concluded that full exploitation of the versatility of MD-PC is at an early state of development; as a consequence several significant changes in practice might be expected in the next few years.3
Coupling of PC with Other Chromatographic Techniques Another means of realizing MD-PC involves the combination of two complementary separation techniques that use different methods of separation. In such multimodal separations different techniques can be coupled such that PC is used as the second dimension and another separation method as the first. The first published multimodal separation was the coupling of electrophoresis and TLC.25 Some possible variations, in which PC is used as the second dimension are GCPC, SFCPC, HPLCPC, CEPC, CCCPC, OPLCPC, RPCPC and TLC. However, current interest mainly surrounds the coupling of HPLC and TLC, to which considerable attention has been devoted for the solution of difficult separation problems. As the direct coupling of HPLC and TLC is known26 several papers27–29 have been published describing the on-line
RECENT APPLICATIONS IN MULTIDIMENSIONAL CHROMATOGRAPHY December 2003
Nyiredy
coupling of HPLC and PC. If PC is used as the second method, all MD-PC methods discussed above can be applied to further increase the separating power.
Outlook MD-PC separations require not only a multiplicity of separation stages, but also the integrity of separation achieved in one stage to be transferred to the others. The process of separation on a 2D plane is the clearest example of MD separations. The greatest strength of correctly applied MD-PC is that compounds are distributed widely over a 2D surface with high zone (peak) capacity.3 Based on theory and experimental observations it can be predicted that a zone capacity of ~1500 could be achieved by 2D multiple development. Because the same result can be achieved by 2D OPLC development on HPTLC plates, the combination of 2D OPLC with multiple development encompasses all the advantages of two geometrical dimensions, forced-flow planar chromatography and the separating capacity of multiple development. Favourable conditions could start the OPLC separation and multiple development in the first dimension, reducing the total solvent strength stepwise at different mobile-phase selectivity to achieve a crude separation on the basis of the polarity of the compounds to be separated. In the second, perpendicular direction multiple development could be performed at constant ST but with variation of mobile phase selectivity. Another way for the second dimension is the use of coupled layers (LD-OPLC) with decreasing polarity of stationary phases for polishing the resolution of complex matrices. It can be stated that the combination of stationary and mobile phases, as well as forced-flow and multiple development techniques are the most powerful modes of instrumental MD-PC.
11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.
P. Schoenmakers, LC•GC Eur., 3, 2–8 (2003). R. Consden, A.H. Gordon and A.J.P. Martin, J. Biochem., 38, 224–232 (1944). J.G. Kirchner, J.M. Miller and G. Keller, Anal. Chem., 23, 420–429 (1951). D.S. Risley et al., J. Planar Chromatogr., 3, 216–221 (1990). P. Härmälä et al., J. Planar Chromatogr., 3, 515–520 (1990). Sz. Nyiredy, K. Dallenbach and O. Sticher, J. Planar Chromatogr., 1, 336–342 (1988). Sz. Nyiredy, J. Chromatogr. Sci., 40, 553–564 (2002). L.R. Snyder, J. Chromatogr. Sci., 16, 223–234 (1978). K. Burger, Fresenius Z. Anal. Chem., 318, 228–233 (1984). R.C. Pandey, R. Misra and K.L. Rinehart Jr, J. Chromatogr., 169, 129–139 (1979). L. Botz, Sz. Nyiredy and O. Sticher, J. Planar Chromatogr., 3, 352–354 (1990). L. Botz, Sz. Nyiredy and O. Sticher, J. Planar Chromatogr., 4, 115–122 (1991). L.A. Beaver and G. Guiochon, USA Patent No. 4.469.601 (1984). S.K. Poole and C.F. Poole, J. Planar Chromatogr., 2, 478–481 (1989). E.L. Durrum, J. Colloid Sci., 6, 274–290 (1951). J.W. Hofstraat et al., J. Planar Chromatogr., 1, 220–226 (1988). E. Mincsovics, M. Garami and E. Tyihák, J. Planar Chromatogr., 4, 299–303 (1991). K. Burger, Analysis, 18, 1113–1116 (1990). O.R. Queckenberg and A.W. Frahm, J. Planar Chromatogr., 6, 55–61 (1993).
Acknowledgement The author wishes to acknowledge the input and discussion of Professor Peter Schoenmakers to the nomenclature that has appeared in this paper. Thanks are rendered for the fruitful discussions to Dr J. Harangi and Dr G. Tarján.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
Sz. Nyiredy (Ed.), Planar Chromatography — A Retrospective View for the Third Millennium, Springer, Budapest, 2000. H.J. Cortes (Ed.), Multidimensional Chromatography, Marcel Dekker, New York, 1990. L. Mondello, A.C. Lewis and K.D. Bartle (Eds.), Multidimensional Chromatography, Wiley, New York, 2001. J.C. Giddings, Anal. Chem., 56, 1258A–1270A (1984). J.C. Giddings, HRC & CC, 10, 319–323 (1987). C.F. Poole and S.K. Poole, J. Chromatogr. A, 703, 573–612 (1995). J.B. Phillips and J. Beens, J. Chromatogr. A, 856, 331–347 (1999). J. Dallüge, J. Beens and U.A.Th. Brinkman, J. Chromatogr. A, 1000, 69–108 (2003). T. Wehr, LC•GC Eur., 3, 2–8 (2003). A.E. Herr et al., Anal. Chem., 75, 1180–1187 (2003).
www.lcgceurope.com
9
