Journal of Chromatographic Science, Vol.

40, November/December 2002

Planar Chromatographic Method Development Using

the PRISMA Optimization System and Flow Charts

Sz. Nyiredy

Downloaded from https://academic.oup.com/chromsci/article/40/10/553/534466 by guest on 01 February 2022

Research Institute for Medicinal Plants, H-2011 Budakalász, P.O. Box 11, Hungary

Abstract (14), role of chamber saturation (15) in the optimization of PC,

and transfer of the optimized analytical mobile phase to the var-
This study presents a modern planar chromatographic method- ious analytical and preparative forced-flow liquid chromato-
development procedure, based on the “PRISMA” optimization graphic techniques provided the basis for an automatic
system, in which the optimum separation is achieved systematically optimization system for nonpolar compounds (16). In PC method
and the structures and properties of the substances to be separated development, the selection of the stationary and vapor phases, a
are not known. The procedure consists of three stages. In the first of suitable development mode, and the forced-flow technique estab-
these the basic conditions the stationary phase, vapor phase, and lished the PRISMA optimization system (17).
individual solvents are selected with a TLC procedure (generally in In the last 15 years or more the applicability of the PRISMA opti-
nonsaturated chromatographic chambers). In the second stage, the
mization system has been discussed in hundreds of papers. This
optimum combination of the selected solvents is determined with
the PRISMA model. The third part of the procedure includes the
study describes a modern PC method development process based
selection of the development mode (circular, linear, or anticircular); on this optimization system. The optimum PC separation can be
the selection of an appropriate forced-flow chromatographic achieved systematically and the structures and properties of the
technique (over-pressured layer chromatography or rotation planar substances to be separated do not have to be known. For practical
chromatography) with high-performance thin-layer reasons the strategy is also given in the form of flow charts.
chromatographic plates; the transfer of the optimized mobile phase
to the various analytical, planar, or column preparative liquid
chromatographic techniques; and the selection of the operating
conditions. For practical reasons, the optimization process is Experimental
presented with the help of flow charts.
The PRISMA system for PC
The PRISMA system in PC (Figure 1) consists of three parts.
The first is the selection of the basic conditions: the stationary
Introduction phase, vapor phase, and individual solvents for the mobile phase
optimization process.
With the introduction of highly sophisticated and automated In the second part, the optimum combination of the individual
instruments for high-performance liquid chromatography
(HPLC), many mobile phase optimization procedures and criteria
(e.g., selectivity) have been described and summarized extensively
(1–4). On the basis of a three-dimensional representation, the sol-
vent classification of Snyder (5), and the seven-point optimization
method of Glajch et al. (6), the “PRISMA” mobile phase optimiza-
tion model was developed by Nyiredy et al. in 1985. The model for
manual selection of solvents and optimization of the mobile
phase was developed first for thin-layer chromatography (TLC)
(7) and HPLC (8) and later applied to forced-flow planar chro-
matographic (FFPC) techniques such as overpressured layer
chromatography (OPLC) (9) and rotation planar chromatography
(RPC) (10). A computer-assisted HPLC mobile phase optimiza-
tion procedure was published later (11,12). The applicability of
the PRISMA system for planar chromatographic (PC) separations Figure 1. The PRISMA system for planar chromatography (17).
has been presented elsewhere (13). The study of the vapor phase

Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission. 1
 Journal of Chromatographic Science, Vol. 40, November/December 2002

solvents selected is determined by means of the PRISMA model. of the chromatographic process have been distinguished on the
By use of this three-dimensional geometric design all mobile basis of the types of interaction involved. Because silica has excel-
phase combinations of between two and five solvents can be lent separating power and is the most used (90–95%) stationary
characterized. The third part of the procedure entails the phase in normal phase (NP)-PC, the PRISMA optimization pro-
selection of the development mode (circular, linear, or anti- cess usually starts with this adsorbent.
circular), chromatographic technique [high-performance In the event that the structures and properties of the sub-
thin-layer chromatography (HPTLC), OPLC, or RPC], and stances to be separated are known, the decision flow chart for sys-
the transfer of the optimized TLC mobile phase to the various tematic selection of the appropriate separation technique and
analytical, preparative planar, and column liquid chromato- stationary phase are given in Figure 2.
graphic techniques. If the sample contains compounds with a wide polarity
range, the automatic multiple development (AMD) mode is
Selection of the basic conditions (the first part of the considered. For aromatic hydrocarbons or isomeric compounds,
PRISMA system) either reversed-phase with aqueous buffers or ion-pair chro-

Downloaded from https://academic.oup.com/chromsci/article/40/10/553/534466 by guest on 01 February 2022

matography might be appropriate. If the sample to be separated
Selection of the stationary phase is extremely polar and contains homologous compounds,
PC separations have been performed on unmodified, modified, reversed-phase chromatography is the best option. If the sample
and impregnated stationary phases because of differences
between the chemical properties of the adsorbent material and Table I. Solvents for the Normal Phase TLC Optimization
the substances of the sample to be separated (18). Different types Procedure with Synder’s Suggested Solvent Classification
and Values (5)

Solvent
strength
Group Solvent (si) xe xd

– n-Hexane 0 – –
I. n-Butyl ether 2.1 0.44 0.18
Diisopropyl ether 2.4 0.48 0.14
Methyl-t-butyl ether 2.7 0.49 0.14
Diethyl ether* 2.8 0.53 0.13
II. i-Pentanol 3.7 0.56 0.19
n-Butanol 3.9 0.56 0.19
i-Propanol 3.9 0.55 0.19
n-Propanol 4.0 0.54 0.19
Ethanol* 4.3 0.52 0.19
Methanol 5.1 0.48 0.22
III. Tetrahydrofuran* 4.0 0.38 0.20
Pyridin 5.3 0.41 0.22
Methoxyethanol 5.5 0.38 0.24
Methylformamide 6.0 0.41 0.23
Dimethylformamide 6.4 0.39 0.21
Dimethylsulfoxide 7.2 0.39 0.23
IV. Acetic acid* 6.0 0.39 0.31
Formamide 9.6 0.36 0.23
Figure 2. Flow chart for selection of the stationary phase and separation tech-
V. Dichloromethane* 3.1 0.29 0.18
nique.
1,1-Dichloroethane 3.5 0.30 0.21
Benzylalcohol 5.7 0.40 0.30
VI. Ethyl acetate* 4.4 0.34 0.23
Methyl ethyl ketone 4.7 0.35 0.22
Dioxane* 4.8 0.36 0.24
Acetone 5.1 0.35 0.23
Acetonitrile 5.8 0.31 0.27
VII. Toluene* 2.4 0.25 0.28
Benzene 2.7 0.23 0.32
Nitrobenzene 4.4 0.26 0.30
Nitromethane 6.0 0.28 0.31
VIII. Chloroform* 4.1 0.25 0.41
Dodecafluoroheptanol 8.8 0.33 0.40
Water 10.2 0.37 0.37

Figure 3. Flow chart for selection of chamber type and saturation (17). * something here???????????

2
Journal of Chromatographic Science, Vol. 40, November/December 2002

contains mainly dissociated compounds, alumina is the most egories. The chambers used for OPLC are nonsaturated S-cham-
suitable stationary phase (17). bers that are devoid of any vapor space. This must be considered
when selecting the appropriate solvents and during optimization
Selection of the vapor phase of the mobile phase. In RPC, the size and therefore the extent of
Selection of the chamber type and vapor space is possible with saturation (10) of the vapor phase can be varied; the micro (M)-
PC only; in practice, little attention is devoted to this (14). and ultramicro (UM)-chambers are S-type chambers. In M-
In PC, two basic types of chromatographic chambers must be chamber RPC, the plate rotates with the small chromatographic
distinguished. In the generally used normal (N)-chamber, the dis- chamber and the vapor space is saturated rapidly. In UM-chamber
tance between the plate and the wall of the chromatographic tank RPC, there is almost no vapor space, similar to OPLC (9).
is more than 3 mm. If this distance is smaller, the chamber is said Among the forced-flow methods OPLC has the highest sepa-
to have the sandwich (S) configuration (15). Both types of rating power because of the possibility of using the optimum
chamber can be used for saturated or nonsaturated systems. As a mobile phase velocity on the HPTLC plate and the long separation
rule of thumb, if the sample contains more than seven substances distance. If the separation distance is an important factor in the

Downloaded from https://academic.oup.com/chromsci/article/40/10/553/534466 by guest on 01 February 2022

or the separation is very difficult, nonsaturated S-chambers should final separation, OPLC should be selected and the preassay for
be selected because this enables subsequent transfer of the opti- mobile phase selection should be performed in a nonsaturated S-
mized TLC mobile phase to a forced-flow separation technique. If chamber. If RPC equipment is available to improve the efficiency
the sample contains fewer than seven compounds to be deter- of the final separation, the chamber of choice for the preassay
mined quantitatively, saturated N-chambers can be selected (15). depends on the type of compound to be separated. If a vapor phase
Often the separation problem cannot be solved by use of capil- of an acidic or basic character is important for the separation, a
lary-driven TLC because of the relatively modest separating saturated S-chamber (if available) is the right choice for TLC pre-
power of this mode. If so, the use of different forced-flow tech- assay. If a saturated S-chamber is not available, a saturated N-
niques (OPLC or RPC) is necessary; this must be considered chamber is useful. Otherwise, a nonsaturated S-chamber should
during selection of the vapor phase. Selection of the vapor phase be chosen for transposition of the mobile phase to an RPC sepa-
therefore depends on whether or not forced-flow techniques are ration. The decision flow chart for selection of the vapor phase is
available for the final separation. The chambers used for forced- given in Figure 3.
flow planar separations can also be assigned to the noted two cat-
Selection of the individual solvents
The third step in the first part of the PRISMA strategy is solvent
selection. This is based on the solvent classification of Snyder (5),
who classified more than 80 solvents into 8 groups for NP chro-
matography according to their properties as proton acceptors (xa)
or proton donors (xd) and their dipole interactions (xn). From
these 8 groups, 27 solvents commonly used in PC have been listed
in Table I.
For the selection of suitable solvents, the first experiments are
conducted on silica TLC plates in nonsaturated chromatographic
chambers with the ten solvents indicated by stars in Table I. After
these first TLC experiments with the neat solvents, the solvent
strength must either be reduced or increased so that the sub-
stance zones are distributed between RF 0.2 and 0.8. The two the-
oretical situations are depicted as “A” and “P” in Figure 4. If the
compounds to be separated migrate in the upper third of the plate
(A-I in Figure 4), the solvent strength must be reduced by dilution
with hexane [solvent strength (si) = 0]. If the substances do not
migrate with the neat solvents, the solvent strength must be
increased (P-I in Figure 4) by the addition of a small amount of
water. In both circumstances the solvent strength should be
varied so that better distribution of the substance zones is
obtained. Consequently, the structures and properties of the com-
pounds to be separated need not be known. Their classification as
nonpolar (A) or polar (P) compounds (13) can be made in accor-
dance with their behavior in these TLC experiments.
If a good separation is obtained with particular solvents,
homologs or other solvents of the same group can also be tested,
as indicated by A-II and P-II in Figure 4. After these experiments,
Figure 4. Selection of the individual solvents with the first part of the system the solvents giving the best separations are chosen for further
(13). optimization for nonpolar compounds. For optimization of the
mobile phase for polar compounds, the solvents selected should

3
 Journal of Chromatographic Science, Vol. 40, November/December 2002

include one in which the compounds do not migrate—necessary tation of si and the proportions of the components that determine
for the subsequent transfer of the mobile phase to certain forced- the selectivity. Between two and five solvents can usually be
flow techniques—by use of the third part of the PRISMA system selected for construction of the PRISMA model.
(13). Occasionally suitable separation can be achieved with the If the si values of the neat solvents are plotted vertically above
first part of the system. The individual steps of solvent selection by an equilateral triangle (Figure 6A) and if the two-dimensional
use of the PRISMA system are summarized in a flow chart in representation of the solvent concentrations, which primarily
Figure 5. influence the selectivity, is plotted on the horizontal plane, a
prism is obtained with an equilateral triangle as its base. The
The PRISMA model (the second part of the system) lengths of the edges of the prism correspond to the solvent
The PRISMA model, the pivotal component of the PRISMA strengths of the neat solvents (SA, SB, and SC) in question
system (Figure 6), can be visualized as a graphic spatial represen- (Figure 6B). Because different solvents are of different strengths,
the lengths of the edges of the prism are generally unequal and
the top plane of the prism will not be parallel and congruent with

Downloaded from https://academic.oup.com/chromsci/article/40/10/553/534466 by guest on 01 February 2022

its base (Figure 6C).
If the prism is theoretically cut parallel to the base at the height
of the lowest edge (determined by the solvent with the lowest sol-
vent strength in the system) the lower part gives a regular prism
(Figure 6D) that has top and bottom planes that are parallel equi-
lateral triangles. The PRISMA model thus consists of three parts:
the regular part of the prism (white in Figure 6) with congruent
base and upper surfaces; the irregular truncated top prism (dark
gray in Figure 6); and the base, symbolizing the modifier (light
gray in Figure 6). The solvent strength values of the modifiers are
treated as an additive by the PRISMA model (13). For the sake of
simplicity, the solvent strength values of the modifiers (e.g., acids
and ion pairs) are neglected because they are usually present at
low, constant concentrations (generally 0.1% and 3%).
In NP chromatography, the upper irregular part of the model is
used for characterization of mobile phases for the separation of
polar compounds, whereas the regular part is used to charac-
terize mobile phases for the separation of nonpolar substances. In
typical reversed-phase chromatography the regular part is used
for the separation, irrespective of the polarity of the compounds
to be separated.

The irregular part of the model

The three corners of the top cover (irregular) triangle of the

FFigure 5. Flow chart for selection of the individual solvents.

Figure 7. Combination of three neat solvents (A, B, C) on the irregular top tri-
angle of the PRISMA model. (A) represents the volume fractions of the sol-
vents A, B, and C at point PS. and (B) is the selectivity points in the top triangle
Figure 6. Graphic representation of the PRISMA model. representing the combinations of the three solvents by three-digit numbers

4
Journal of Chromatographic Science, Vol. 40, November/December 2002

prism represent the three undiluted neat solvents (Figure 7). The = 333, because this composition is obtained if equal amounts are
corner corresponding to the longest edge of the prism is equiva- taken from all three (A, B, and C) solvents (Figure 7B). Because
lent to solvent A (the solvent with the highest solvent strength), the three solvents selected usually differ in solvent strength, all
whereas solvent C (the solvent with the lowest solvent strength) selectivity points on the surface of the cover plate of the irregular
corresponds to the corner of the shortest edge of the prism. In the part of the prism represent different solvent strengths and selec-
triangle shown on the left of Figure 7, a particular solvent com- tivity.
position (PS) can be characterized by the volume fractions of the The points along the edges of the cover plate represent mix-
corners. Here the volume fraction of solvent A is 0.1, solvent B is tures of two solvents (“A and B”, “B and C”, and “A and C”). Inside
0.3, and solvent C is 0.6 (this means that for the mobile phase the irregular triangle the selectivity points represent mixtures of
characterized by point PS, the concentration (%, v/v) of solvent A the three solvents (A, B, and C). Dilution of a mobile phase mix-
is 10%, solvent B is 30%, and solvent C is 60%). ture with a solvent of zero solvent strength (hexane in NP chro-
This point of the triangle where the tenfold values (PA, PB, and matography, water in reversed-phase chromatography) gives
PC) of all three characteristic volume fractions are integers can be mixtures of four solvents, which are characterized by the same

Downloaded from https://academic.oup.com/chromsci/article/40/10/553/534466 by guest on 01 February 2022

defined by a three-digit number. This number, in which the sum selectivity point but have lower solvent strengths. These solvents
of the digits is 10, can be obtained by multiplying the volume frac- are represented by the inside points of the upper, irregular part of
tions by 10 and arranging them in order of diminishing solvent the prism. Of course, mobile phase mixtures with a solvent
strength. The solvent composition shown in Figure 7A can there- strength lower than that of solvent C are excluded.
fore be defined by point 136 (which means a mixture of 10% sol- Because in the irregular top triangle the solvent strength differs
vent A, 30% solvent B, and 60% solvent C). The triangle on the at each selectivity point, which greatly influences the separation,
right of Figure 7 shows all compositions of the three solvents on the steps between two selectivity points are often very large (9).
the cover plate of the prism. They are characterized by integer The solvent mixtures between these selectivity points can then be
three-digit numbers and defined as selectivity points (PS). described by three two-digit numbers (e.g., PS = 57 – 18 – 25).
Because the center of the triangle cannot be described by use of Although even finer adjustments can be made (e.g., PS = 56.7 –
the system already defined, it is characterized by definition as PS 18.4 –24.9), 1% accuracy is usually sufficient.

The regular part of the model

The base and top plane of the regular part of the prism are con-
gruent equilateral triangles. The height of this part of the prism
corresponds to the solvent strength of the weakest solvent.
Because of the original selection of a decreasing order of solvent
strength for solvents A, B, and C, this is solvent ‘C’. This means
that corner ‘C’ of the regular prism represents undiluted solvent
C. The solvent mixtures represented by corners AD and BD
(Figure 8) can be obtained by diluting solvents A and B to the sol-
vent strength of C with the use of a solvent of zero strength.
Mobile phases characterized by other points on the top cover
plate can be obtained by mixing the solvents represented by the
corners of the top cover plate in the volume proportions that cor-
respond to the point in question. The selectivity points repre-
senting the two- and three-solvent compositions (the second and
Figure 8. Example of the calculation of mobile phase composition with the third solvent is the solvent strength regulator) are given on the
PRISMA model in NP chromatography
left of Figure 9. The selectivity points symbolizing four solvent
compositions frequently used in optimization are character-
ized—similarly to the points on the top irregular part—by three-
digit numbers (on the right in Figure 9). The 4 basic selectivity
points within the prism (PS = 333, 811, 181, and 118) for 4 sol-
vent mixtures are also shown in Figure 9.
The selectivity points on the vertical planes of the regular part
of the prism can be obtained by diluting the solvent mixtures with
a solvent of zero strength. The solvent strength values decrease
from top to bottom; at the base of the prism the solvent strength
value is zero. Hexane is used to reduce the mobile phase strength
in NP chromatography and water in RP chromatography. If sec-
Figure 9. The selectivity points in the regular part of the PRISMA model. (A) tions are prepared from the regular prism parallel to the base, tri-
selectivity points in the corners represent two solvent combinations; those
angles with different solvent strengths are obtained. Therefore, all
along the edges of the triangle represent three solvent combinations. (B) selec-
tivity points in the triangle represent four solvent combinations (the three neat
points on one of these triangles represent the same solvent
solvents and the solvent-strength regulator). strength, whereas all points on a vertical straight line correspond
to the same selectivity (13).

5
 Journal of Chromatographic Science, Vol. 40, November/December 2002

Calculation of mobile phase composition points on the top regular triangle can be described as PS* = PAD;
Construction of the irregular and regular parts of the prism PBD; PCD, where
and of the solvent composition corresponding to individual
points is demonstrated in Figure 9 for use of ethanol (SA = 4.3), PAD = (SA × PA)/ST; PBD = (SB × PB)/ST; and
chloroform (SB = 4.1), and diethyl ether (SC = 2.8) in NP chro- PCD = (SC × PC)/ST = PC Eq. 1
matography; hexane (SD = 0) is used to adjust solvent strength.
Point C on the top of the regular prism corresponds to undiluted The conversion of the selectivity points between the irregular
diethyl ether and the other two corners (AD and BD) are obtained and regular triangle can be described as follows:
by diluting ethanol and chloroform with the zero-solvent-
strength hexane to a solvent strength of 2.8. These correspond PS* = (SA × PA)/ST; (SB × PB)/ST; (SC × PC)/ST Eq. 2
to mixtures of 65.1% ethanol and 34.9% hexane (corner AD)
and 68.3% chloroform and 31.7% hexane (BD), as can be seen The projection of ethanol, chloroform, and diethyl ether on the
in Figure 8. top irregular triangle (ST = 4.09) in PS = 631, will be at:

Downloaded from https://academic.oup.com/chromsci/article/40/10/553/534466 by guest on 01 February 2022

The total solvent strength of a four-solvent (A, B, C, and hexane,
the solvent strength regulator) mobile phase is the sum of the sol- ST = 2.8, PS* = (4.3 × 60)/4.09; (4.1 × 30)/4.09;
vent strengths of the volume fractions (f) of the single solvents (2.8 × 10)/4.09 Eq. 3
(17): ST = fA x SA + fB x SB + fC x SC, which gives a solvent
strength of 4.09 for PS = 631 on the top irregular triangle (ST = which gives the selectivity point PS* = 63.08 – 30.07 – 6.85
0.6 x 4.3 + 0.3 x 4.1 + 0.1 x 2.8 = 4.09). The same relationship rounded off as PS* = 63.1 – 30.1 – 6.9. This means that the rela-
between the concentrations of A, B, and C is valid for dilution of tionship between the different solvents (cA: cB: cC) is constant
this solvent mixture to ST = 2.8. within the top irregular triangle for the different solvent
For the calculation of mobile phase composition it must be strengths (in sections parallel to the top irregular triangle) and is
considered that in the regular part of the prism a minimum of also constant within the regular part at horizontal sections (11).
two corners of the triangles represent solvents containing the sol- Therefore, if PS is defined as a selectivity point in the top irregular
vent strength regulator. For solvent strength 2.8 and PS* = 631 triangle and PS* and PS** in the regular part, PS PS* = PS**; the
(Figure 8), the f of the solvent mixture represented by corner AD mobile phase must therefore always be characterized by the sol-
(65.1% ethanol and 34.9% hexane) is 0.6, the volume fraction of
the solvent mixture represented by corner BD (68.3% chloroform
and 31.7% hexane) is 0.3, whereas that of corner C (100% diethyl
ether and 0% hexane) is 0.1. Accordingly, the concentration of
ethanol in the final mixture (in PS* = 631 at ST = 2.8) is 30.45%
(0.6 x 65.1), that of chloroform is 20.49% (0.3 x 68.3), that of
diethyl ether is 10% (0.1 x 100), and that of hexane is 30.95% (100
– 39.06 – 20.49 – 10).
Dilution of the solvent characterized by the point PS* (ST = 2.8;
PS = 631) with hexane in different proportions gives mixtures
with the same selectivity (PS* = PS**) but lower solvent
strengths [e.g., ST = 1.4; PS = 631 (15.23% [30.45:2] ethanol +
10.25% [20.49:2] chloroform + 5% [10:2] diethyl ether + 69.52%
[30.95:2 + 50] hexane)]. Obviously the same solvent composition
can also be obtained by first calculating the compositions at the
edges of the prism, corresponding to the solvent strength desired,
and subsequently mixing these in the proportions given by the
volume fractions (selectivity point).

Relationship between the irregular and regular

parts of the model
As can be deduced from the example calculation above, the rela-
tionship between the volume fractions of the top irregular and the
regular parts of the model are not identical, because, as is
apparent from Figure 8, the regular triangle with the highest sol-
vent strength is geometrically a projection of the top irregular tri-
angle. Only point C is identical in both triangles, all other points
characterize other solvent compositions.
The selectivity points in the top irregular triangle can be
described as PS = PA; PB; PC. In the highest regular triangle, Figure 10. Flow chart for mobile-phase optimization for non-polar substances
point C represents a neat solvent (PC = PCD), where the corners (13).
consist of PAD and PBD instead of PA and PB. So the selectivity

6
Journal of Chromatographic Science, Vol. 40, November/December 2002

vent strength and the selectivity point. (especially near PS = 811).

When the selectivity points on the top triangle are changed the
Optimization strategy solvent strength is also changed; a small change in the selectivity
The strategy for optimizing the solvent combination for non- point might therefore result in a large difference in resolution,
polar compounds is depicted in a flow chart in Figure 10. especially when the solvent strength of the selected solvents dif-
If three solvents were selected in the first part of the system, fers substantially. The subsequent procedure is similar to that for
optimization is achieved within the regular part of the model with the nonpolar substances but the solvent strength must be
the help of the four basic selectivity points. If two solvents were adjusted after suitable selectivity has been obtained (19).
selected, optimization is achieved along a side of the prism. In Optimization of the solvent strength and selectivity points
both instances the solvent strength is adjusted first, then different must be performed until the beginnings of a separation of the
selectivity points are tested. If 3–5 solvents are selected as best, compounds, at least, are obtained. This can usually be achieved
the number of solvents is reduced after the criteria number of with the first PRISMA combination, if the individual solvents
separated substances and RF values have been met. If the sol- were selected correctly.

Downloaded from https://academic.oup.com/chromsci/article/40/10/553/534466 by guest on 01 February 2022

vent combinations tested with the equivalent strategy did not
result in sufficient separation, or if the beginning of a separation Automatic mobile-phase optimization
of the essential substance pairs could not be observed, other sol- For the separation of nonpolar compounds by use of an NP sta-
vents must be selected and the process must be repeated, as indi- tionary phase and saturated TLC systems at a constant solvent
cated in the flow chart. strength the correlation between the selectivity points is
The flow chart for optimization of the solvent combination for described (16) by the function:
polar compounds is shown in Figure 11. For polar compounds
optimization is always started on the top irregular triangle of the hRF = a(PS)2 + b (PS) + c (horizontal function) Eq. 4
model, either within the triangle (if three solvents were selected)
or along one side (if two solvents were selected). Water is usually The hRF values were always measured along the edges of the
used as modifier to increase the solvent strength and reduce triangle through two basic selectivity points. The vertical correla-
tailing; if it is used as one of the selected solvents, several selec- tion at constant selectivity points between a variety of solvent
tivity points cannot be tested because of miscibility problems strengths can be described by:

ST = d ln hRF + e (vertical function) Eq. 5

Measurements on three solvent-strength levels are needed for

calculation of hRF values for all selectivity points in the spatial
design, because the vertical correlation can be linearized. These
correlations are also relevant when constant amounts of modi-
fiers are used. From these correlations the chromatographic
behavior of substances to be separated can be predicted at all
selectivity points within the PRISMA model in saturated chro-
matographic chambers. The separation quality of predicted chro-
matograms is assessed by use of the chromatographic response
function (CRF) (2). Twelve measurements are needed to find a
local optimum, and fifteen for the global optimum. Pelander et al.
(20) studied the application of these relationships to the irregular
part of the model. They found that for saturated chambers there
was a linear correlation between RF and the solvent strength at a
constant PS value. Again 12 measurements were needed to find a
local optimum and 15 for the global optimum.
For nonsaturated chromatographic systems using mobile
phases containing n solvents (multicomponent mobile phases), n
fronts can occur. Because the compounds to be separated can be
located between the different fronts, these correlations are not
valid.

The Third Part of the System

Selection of the mode of development

When the best possible mobile phase composition has been
determined on silica, a suitable development mode and separa-
Figure 11. Flow chart for mobile-phase optimization for polar substances (13). tion technique must be selected. If the beginnings of a separation
of the compounds are observed, use of linear development mode

7
 Journal of Chromatographic Science, Vol. 40, November/December 2002

and OPLC are indicated. If the separation problem is in the upper mobile phase can be used directly; no prerun is necessary
RF range the anticircular mode should be used; if the problem is because, compared with OPLC, U-RPC is not a completely closed
in the lower RF range the circular development mode should be chamber and some vapor phase is present. Both methods enable
chosen (21). For both, the development is first performed on a the possibility of “scaling-up” to the corresponding preparative
TLC plate; if the separation is insufficient HPTLC plates with a techniques and the use of column rotation planar chromatog-
suitable forced-flow technique can be employed (as summarized raphy (23).
in Figure 12). For fully off-line separation techniques dry filled preparative
If the separation problem cannot be solved by use of this columns (flash chromatography, low pressure liquid chromatog-
strategy and silica, other stationary phases must be tested and the raphy, and medium pressure liquid chromatography) can be equi-
optimization process must be started again with the first part of librated with the solvent used for the prerun in analytical OPLC,
the system. whereas for slurry packing the slurry must be prepared using the
same solvent as was used for the prerun of OPLC or U-RPC. In
Transfer of the optimized mobile phase both techniques air bubbles can be eliminated by pumping of the

Downloaded from https://academic.oup.com/chromsci/article/40/10/553/534466 by guest on 01 February 2022

The main point of the strategy for transfer of the optimized appropriate quantity of the solvent used for the prerun; the
mobile phase is that separation of samples is generally begun in a preparative separation can then be started with the optimized
nonsaturated TLC chamber, in which the compounds to be sepa- nonsaturated TLC mobile phase (24).
rated are placed between the hRF values of 25 and 80. In fully off- Figure 13A illustrates the possibilities of direct transfer after
line separations the mobile phase optimized in an nonsaturated optimization of the TLC procedure in nonsaturated chromato-
TLC chamber can be transferred without change—by use of an graphic chambers (24). Different lines show the transfers appli-
appropriate prerun—to analytical OPLC (22). For the separation cable for different methods. Dotted lines and thin lines denote
of nonpolar compounds this prerun can be performed with transfers between off-line and on-line methods, respectively.
hexane (si = 0); for separation of polar substances the prerun can Thick lines indicate that the optimized mobile phase can be
be performed with any component of the mobile phase in which transferred without change between different solid–liquid planar
the components do not migrate (selection of this solvent should and column chromatographic techniques, whether off-line
be considered during optimization of the mobile phase) (9). By or online.
use of ultramicro chamber RPC (U-RPC) the optimized TLC If the TLC mobile phase was optimized in saturated chambers
to improve the separations the M-RPC technique can be used,

Figure 13. Transfer of the optimized TLC mobile phase for analytical and
preparative forced-flow column and planar chromatographic techniques. (A)
Figure 12. Flow chart for selection of the mode of development and the Procedure starting with nonsaturated chromatographic chambers and (B)
forced-flow technique (17). starting with saturated chromatographic chambers

8
Journal of Chromatographic Science, Vol. 40, November/December 2002

without altering the TLC mobile phase (Figure 13B), irrespective Conclusion
of whether off-line or line techniques are used (24).
For both nonpolar and polar substances the optimization
strategy proposed is always started on silica. The individual sol-
Discussion vents can be selected by performing parallel experiments in ten
chambers. If a solvent was not characterized by Snyder (11), an
approximate value is chosen from one of its listed homologs. We
Application of the PRISMA method recommend the use of a nonsaturated chamber because this facil-
Several hundred papers have reported the use of the PRISMA itates mobile phase transfer to OPLC (5); this is also the reason for
system for method development in planar chromatography, for selection of a so-called suitable solvent in the first part of the
up-scaling procedures, and for physicochemical studies. The system for polar compounds.
system has also been used for biomedical, environmental, and For the separation of nonpolar compounds optimization with
toxicological analysis. Dozens of papers cover the separation of

Downloaded from https://academic.oup.com/chromsci/article/40/10/553/534466 by guest on 01 February 2022

the PRISMA model is generally a rapid process because a few
different classes of substance, some of them (25–58) are listed in experiments are sufficient to enable determination of the
Table II. optimum mobile phase composition. In contrast, optimization
with the second part of the system is a longer process for polar
Table II. Selection of Publications Using the PRISMA compounds because of the simultaneous change of solvent
Optimization System for Planar Chromatographic strength and selectivity. When water in particular is one of the
Separation solvents selected for construction of the triangle, a small change
in selectivity results in a large change in resolution. More chro-
Substance classes Authors References matographic experience is therefore necessary if the separation
problem is to be solved rapidly (17).
Alkaloids L. Botz, L.Gy. Szabó (25) Modern FFPC techniques can be employed systematically to
M. Waksmundzka-Hajnos, (26)
improve the separation efficiency on fine particle-size HPTLC
A. Petruczynik
Amino Acids Sz. Nyiredy et al. (27)
plates. The PRISMA system combines the appropriate develop-
M. Remelli et al. (28) ment mode with the appropriate forced-flow technique and use of
Amphetamine derivatives Zs. Fatér et al. (29) a mobile phase of optimized composition. This offers special pos-
Anilines M. Waksmundzka-Hajnos, (30) sibilities for solving difficult planar chromatographic separation
A. Petruczynik problems.
Anthraquinones K. Danielson, G.W. Francis (31)
Barbiturates Zs. Fatér et al. (32)
T. Aman et al. (33)
Benzodiazepines C. Cimpoiu et al. (34) References
Coumarins P. Bruno et al. (35)
P. Vuorela et al. (36) 1. J.C. Berridge. Techniques for the Automated Optimisation of HPLC
Dyes S. Gocan et al. (37) Separation. Wiley, Chichester, U.K.,1985.
Essential oils I. Wassmuth-Wagner, H. Jork (38) 2. P.J. Schoenmakers. Optimization of Chromatographic Selectivity.
P. Dugo et al. (39) Elsevier, Amsterdam, Oxford, New York, Tokyo, 1986.
Estrogens S.K. Poole et al. (40) 3. C. Nsengiyumva, J.O. De Beer, W. Van de Wauw, A.J. Vlietinck, S.
Fatty Acids A. Pyka, K. Bober (41) de Swaef, and F. Parmentier. Optimization of solvent selectivity for
Flavonoids K. Dallenbach-Tölke (42) the chromatographic separation of fat-soluble vitamins using a mix-
M.A. Hawryl, E. Soczewinski (43) ture-design statistical technique. Chromatographia 47: 401–12
(1998).
Hepatotoxins A. Pelander et al. (44)
4. L. Szepesy. Possibilities and pitfalls in defining selectivity in HPLC.
Herbicides J. Tekel (45) Chromatographia 51: S 98–107 (2000).
Indole derivatives T. Yrjönen et al. (46) 5. L.R. Snyder. Classification of the solvent properties of common liq-
Insecticides W. Funk et al. (47) uids. J. Chromatogr. Sci. 16: 223–34 (1978).
Iridoids K. Dallenbach-Toelke (48) 6. J.L. Glajch, J.J. Kirkland, and L.R. Snyder. Practical Optimization of
P. Junior (49) solvent selectivity in liquid–solid chromatography using a mixture-
Pharmaceuticals H.D. Ahmed, C.F. Poole (50) design statistical technique. J. Chromatogr. 238: 269–80 (1982).
H.E.M. Salomies, P.K. Salo (51) 7. Sz. Nyiredy, C.A.J. Erdelmeier, B. Meier, and O. Sticher. “PRISMA”:
Phenols J. Bladek et al. (52) Ein modell zur optimierung der mobilen phase für die
M. Waksmundzka-Hajnos, (30) dünnschichtchromatographie, vorgestellt anhand verschiedener
naturstofftrennungen. Planta med. 51: 241–46 (1985).
A. Petruczynik
8. Sz. Nyiredy, B. Meier, C.A.J. Erdelmeier, and O. Sticher. ‘PRISMA’: a
Pigments Q.M. Andersen, G.W. Francis (53) geometrical design for solvent optimization in HPLC. J. High Resol.
K. Morita et al. (54) Chromatogr. 8: 186–88 (1985).
Prostaglandins P. Bruno (55) 9. K. Dallenbach-Tölke, Sz. Nyiredy, B. Meier, and O. Sticher.
Saponins G. Bader et al. (56) Optimization of overpressured layer chromatography of polar natu-
K. Dallenbach-Toelke et al. (57) rally-occurring compounds by the ‘PRISMA’ model. J. Chromatogr.
Synthetic polymers L.S. Litvinova (58) 365: 63–72 (1986).
10. Sz. Nyiredy, C.A.J. Erdelmeier, B. Meier, and O. Sticher. “PRISMA”

9
 Journal of Chromatographic Science, Vol. 40, November/December 2002

optimie-rung der mobilen phase in der präparativen CLC. GIT Suppl. layer. Chromatographia 38: 520 (1994).
Chromatogr. 4: 24–30 (1986). 32. Zs. Fatér, B. Szabady, and Sz. Nyiredy. Two-dimensional overpres-
11. Sz. Nyiredy, K. Dallenbach-Tölke, and O. Sticher. Correlation and sured layer chromatographic separation of barbiturate derivatives. J.
prediction of the k’ values for mobile-phase optimization in HPLC. J. Planar Chromatogr. 8: 145–47 (1995).
Liq. Chromatogr. 12: 95–116 (1989). 33. T. Aman, I.U. Khan, and Z. Parveen. Spectrophotometric determina-
12. Sz. Nyiredy, W. Wosniok, H. Thiele, and O. Sticher. The ‘PRISMA’ tion of barbituric acid. Anal. Lett. 30: 2765–77 (1997).
model for HPLC mobile-phase optimization using an automatic peak 34. C. Cimpoiu, T. Hodusan, and H. Nascu. Comparative study of
identification approach. J. Liq. Chromatogr. 14: 3077–3110 (1991). mobile phase optimization for the separation of some 1,4-benzodi-
13. Sz. Nyiredy, K. Dallenbach-Tölke, and O. Sticher. The “PRISMA” azepines. J. Planar Chromatogr. 10: 195–99 (1997).
Optimization system in planar chromatography. J. Planar 35. P. Bruno, M. Caselli, and A. Traini. HPTLC and OPLC separation and
Chromatogr. 1: 336–42 (1988). detection of prostaglandin esters using 4-bromomethyl-7-methoxy-
14. F. Geiss. Fundamentals of thin-layer chromatography (Planar coumarin (BrMMC). J. Planar Chromatogr. 1: 299–303 (1988).
Chromatography), Huethig, Heidelberg, Basel, New York, 1987. 36. P. Vuorela, E.-L. Rahko, R. Hiltunen, and H. Vuorela. Overpressured
15. Sz. Nyiredy, Zs. Fatér, L. Botz, and O. Sticher. The role of chamber layer chromatography in comparison with thin-layer and high-per-
saturation in optimization of planar chromatography. J. Planar formance liquid chromatography for the determination of coumarins

Downloaded from https://academic.oup.com/chromsci/article/40/10/553/534466 by guest on 01 February 2022

Chromatogr. 5: 308–15 (1992). with reference to the composition of the mobile phase. J.
16. Sz. Nyiredy, Zs. Fatér, Automatic TLC Mobile-phase optimization Chromatogr. A 670: 191–98 (1994).
procedure using the “PRISMA” model. I. Separation of apolar com- 37. S. Gocan, L. Olenic, and I. Cristea. Optimization of the mobile phase
pounds. J. Planar Chromatogr. 8: 341–45 (1995). in thin layer chromatography for the separation of 4-benzylidene-1-
17. Sz. Nyiredy. “Essential Guides for Method Development”. In Thin- pyrimidyl-5-pyrazolone dyes. Rev. Roum. Chim. 35: 49–52 (1990).
layer (planar) Chromatography. In: Encyclopedia of Separation 38. I. Wassmuth-Wagner, H. Jork. Strategy for the chromatographic sep-
Science. I.D. Wilson, E.R. Adlard, M. Cooke, C.F. Poole, Eds. Vol. 10. aration of the essential oil of curry leaves. J. Planar Chromatogr. 5:
Academic Press, London, U.K., 2000, pp. 4652–4666. 383–87 (1992).
18. T. Kowalska. “Adsorbents in Thin-layer Chromatography”. In Planar 39. P. Dugo, L. Mondello, G. Lamonica, and G. Dugo. OPLC analysis of
Chromatography, A Retrospective View for the Third Millennium. Sz. heterocyclic oxygen compounds from citrus rruit essential oils. J.
Nyiredy, Ed. Springer, Budapest, Hungary, 2001, pp. 33–46. Planar Chromatogr. 9: 120–25 (1996).
19. A.-M. Siouffi and M. Abbou: “Optimization of the Mobile Phase”. in: 40. S.K. Poole, M.T. Belay, and C.F. Poole. Effective systems for the sep-
Planar Chromatography, A Retrospective View for the Third aration of pharmaceutically important estrogens by thin layer chro-
Millennium. Sz. Nyiredy, Ed. Springer, Budapest, Hungary, 2001, pp. matography. J. Planar Chromatogr. 5: 16–27 (1992).
47–67. 41. A. Pyka and K. Bober. Prediction of the RM values of selected
20. A. Pelander, J. Summanen, T. Yrjönen, H. Haarlo, I. Ojanperä, and H. methyl esters of higher fatty acids in RPTLC. J. Planar Chromatogr.
Vuorela. Optimization of Separation in TLC by use of Desirability 15: 59–66 (2002).
Functions and Mixture Designs According to the ‘PRISMA’ Method. 42. K. Dallenbach-Tölke. Anwendung des “PRISMA”-modells auf ver-
J. Planar Chromatogr. 12: 365–71 (1999). schiedene chromatographische methoden bei polaren inhaltsstoffen
21. B. Szabady. “The Different Modes of Development”. In Planar von pharmakopöe-drogen. Dissertation, ETH No. 8178, Zürich,
Chromatography, A Retrospective View for the Third Millennium. Sz. Germany, 1986.
Nyiredy, Ed. Springer, Budapest, Hungary, 2001, pp. 88–102. 43. M.A. Hawryl and E. Soczewinski. Normal phase 2D TLC separation
22. E. Tyihák and E. Mincsovics. “Overpressured Layer Chromatography of flavonoids and phenolic acids from betula sp. leaves. J. Planar
(Optimum Performance Layer Chromatography) (OPLC)”. In Planar Chromatogr. 14: 415–421 (2001).
Chromatography, A Retrospective View for the Third Millennium. Sz. 44. A. Pelander, K. Sivonen, I. Ojanperä, and H. Vuorela. Retardation
Nyiredy, Ed. Springer, Budapest, Hungary, 2001, pp. 137–176. behavior of cyanobacterial hepatotoxins in the irregular part of the
23. Sz. Nyiredy. “Rotation Planar Chromatography (RPC)”. In Planar “PRISMA” model for thin-layer chromatography. J. Planar
Chromato-graphy, A Retrospective View for the Third Millennium. Chromatogr. 10: 434–40 (1997).
Sz. Nyiredy, Ed. Springer, Budapest, Hungary, 2001, pp. 177–199. 45. J. Tekel. Separation of some phenylerea and s-triazine herbicides by
24. Sz. Nyiredy. The transfer possibilities of the mobile Phases between OPLC method with a binary mobile phase. J. Planar Chromatogr. 3:
different liquid chromatographic techniques for the analysis and iso- 326–30 (1990).
lation of compounds of biological matrices. Acta Pharm. Hung. 69: 46. T. Yrjönen, J.P. Haansuu, K. Haahtela, H. Vuorela, and P. Vuorela.
46–56 (1999). Rapid screening of indole-3-acetic acid and other indole derivatives
25. L. Botz and L.Gy. Szabó. Separation of tropane alkaloids by TLC, in bacterial culture broths by planar chromatography. J. Planar
HPTLC, and OPLC methods. J. Planar Chromatogr. 1: 85–87 (1988). Chromatogr. 14: 47–52 (2001).
26. M. Waksmundzka-Hajnos and A. Petruczynik. Retention behavior of 47. W. Funk, L. Cleres, H. Pitzer, and G. Donnevert. Organophosphorus
some alkaloids in thin layer chromatography with bonded stationary insecticides—quantitative HPTLC determination and characteriza-
phases and binary mobile phases. J. Planar Chromatogr. 14: 364–70 tion. J. Planar Chromatogr. 2: 285–89 (1989).
(2001). 48. K. Dallenbach-Toelke, Sz. Nyiredy, and O. Sticher. Application of
27. Sz. Nyiredy, K. Dallenbach-Tölke, and O. Sticher. Applicability of various planar chromatographic techniques for the separation of iri-
forced-flow planar chromatographic methods for the separation of doid glycosides from veronica officinalis. J. Chromatogr. 404:
enantiomers on chiralplate. J. Chromatogr. 450: 241–52 (1989). 365–71 (1987).
28. M. Remelli, R. Piazza, and F. Pulidori. HPTLC separation of aromatic 49. P. Junior. Recent developments in the isolation and structure elucida-
a-amino acid enantiomers on a new histidine-based stationary phase tion of naturally occurring iridoid compounds. Planta Med. 56: 1–13
using ligand exchange. Chromatographia 32: 278–84 (1991). (1990).
29. Zs. Fatér, G. Tasi, B. Szabady, and Sz. Nyiredy. Identification of 50. H.D. Ahmed and C.F. Poole. Determination of minor components in
amphetamine derivatives by uni-dimensional multiple development metoprolol tartrate tablets as dansyl derivatives by TLC and fluores-
and two-dimensional HPTLC combined with post-chromatographic cence scanning densitometry. J. Planar Chromatogr. 4: 218–22
derivatization. J. Planar Chromatogr. 11: 225–29 (1998). (1991).
30. M. Waksmundzka-Hajnos and A. Petruczynik. Retention behaviour 51. H.E.M. Salomies and P.K. Salo. Determination of oxycodone
of some phenols and anilines on CN-, diol-, and silica precoated hydrochloride in oral solutions by high-performance thin-layer chro-
HPTLC plates. J. Liq. Chromatogr. 22: 51–76 (1999). matography/densitometry. J. AOAC Int. 83: 149– 501 (2000).
31. K. Danielson and G.W. Francis. An alternative solvent system for the 52. J. Bladek, A. Rostkowski, and S. Neffe. The application of TLC to the
separation of anthraquinone aglycones from rhubarb on silica thin determination of phenol residues in water. J. Planar Chromatogr. 11:

10
Journal of Chromatographic Science, Vol. 40, November/December 2002

330–35 (1998). 56. G. Bader, V. Wray, and K. Hiller. The main saponins from aerial parts
53. Q.M. Andersen, G.W. Francis. “Natural Pigments”. In Handbook of and the roots of solidago virgaurea subsp. virgaurea. Planta Med. 61:
Thin-layer Chromatography, 2nd ed. Revised and extended. J. 158–61 (1995).
Sherma, B. Fried, Eds. Dekker, New York, NY, 1996, pp. 715–52. 57. K. Dallenbach-Toelke, Sz. Nyiredy, S. Y. Mészáros, and O. Sticher.
54. K. Morita, S. Koike, and T. Aishima. Optimization of the mobile TLC, HPTLC, and OPLC separation of ginsenosides. J. High
phase by the prisma and simplex methods for the HPTLC of synthetic Resolution Chromatogr. & Chromatogr. Communications 10:
red pigments. J. Planar Chromatogr. 11: 94–99 (1998). 362–64(1987).
55. P. Bruno, M. Caselli, A. Mangone, A. Traini, and C. Trisolini. OPLC 58. L.S. Litvinova. Practical aspects of the adsorption chromatography of
separation and quantitative determination of fluorescent derivatives synthetic polymers. J. Planar Chromatogr. 11: 114–18 (1998).
of prostaglandins at picogram level. J. Planar Chromatogr. 7: 362–67
(1994). Manuscript accepted August 13, 2002.

Downloaded from https://academic.oup.com/chromsci/article/40/10/553/534466 by guest on 01 February 2022

11