This lesson introduces the primary retention modes in liquid chromatography: normal-phase (NPLC), hydrophilic-interaction (HILIC), and reversed-phase liquid chromatography (RPLC). We will learn how analyte polarity and stationary/mobile phase composition dictate retention behavior across these modes. Through theoretical models and practical considerations, the lesson clarifies the mechanistic basis of retention and highlights the strengths and limitations of each approach. Particular emphasis is placed on understanding the log–log and log-linear retention models, as well as the versatility of RPLC for a wide variety of analyte classes.
Learning Goals #
After this lesson you should be able to
Distinguish between NPLC, HILIC, and RPLC in terms of retention mechanism, mobile-phase composition, and analyte compatibility.
Apply the concept of competitive adsorption in describing retention in NPLC.
Interpret and apply log–log and log-linear retention models for NPLC and RPLC, respectively.
Assess the suitability of LC modes based on analyte polarity and mobile-phase miscibility.
Describe the influence of organic modifiers on retention time, selectivity, and peak shape in RPLC.
READ SECTION 3.6.1
Normal-Phase Liquid Cromatography
1. Normal-phase liquid chromatography (NPLC) #
NPLC, refers to classical liquid-solid chromatography (LSC), using an adsorptive material with a polar surface as the stationary phase and a less polar mobile phase. It is the oldest form of LC, which entitled it to carry the classification “normal”, even though it is increasingly abnormal today. Silica is again the most-common packing material. Its polarity is derived from the presence of silanol groups on the surface.
1.1. Competitive adsorption #
Retention in liquid-solid NPLC can be described by competitive adsorption. The mental picture is for active sites on the stationary surface to be occupied by either analyte molecules or molecules from the mobile phase. The weakest eluents are non-polar, such as alkanes (e.g. n-heptane). A more-polar co-solvent “modifier” (e.g. ethyl acetate, dichloromethane) is required to displace analyte molecules from the active sites on the surface.
Figure 1. Schematic illustration of competitive adsorption in normal-phase LC. In this figure, the silanol groups of the stationary phase are depicted as pink with the yellow hydrogen atom. The polar mobile-phase molecules are green, and the analyte molecules are purple. The non-polar solvent molecules of the mobile phase are not shown. In this simple schematic, analyte molecules and modifier molecules both occupy a single adsorption site.
If we denote the modifier with \text{M} and the analyte molecules with \text{A}, we can describe the adsorption equilibrium as follows.
Equation 3.30: \text{A}_{\text{mob}}+\frac{n_{\text{A}}}{n_{\text{M}}} \text{M}_{\text{ads}} \rightleftharpoons \text{A}_{\text{ads}}+\frac{n_{\text{A}}}{n_{\text{M}}} \text{M}_{\text{mob}}
where the subscripts “ads” and “mob” refer to adsorbed and mobile phase, respectively. is the number of adsorption sites occupied by one analyte molecule and is the number of adsorption sites occupied by one modifier molecule. For example, if an analyte molecule occupies three sites and a modifier molecule one site, then the coefficient in the equilibrium equation is n_{\text{A}}/n_{\text{M}}=3/1 = 1. Neither \text{A} nor \text{M} needs to be an integer number.
The corresponding equilibrium coefficient is
Equation 3.31: K_{\text{AM,ads}}=\frac{[\text{A}_{\text{ads}}]}{[\text{A}_{\text{mob}}]} \cdot \frac{[\text{M}_{\text{mob}}]^{n_{\text{A}}/n_{\text{M}}}}{[\text{M}_{\text{ads}}]^{n_{\text{A}}/n_{\text{M}}}}
Here the ratio \text{A}_{\text{ads}}/\text{A}_{\text{mob}} is the adsorption coefficient of analyte A (K_{\text{A,ads}}), which is proportional to its retention factor k_{\text{A}} with a (phase-ratio) proportionality factor \Phi^{-1}. If we work at the low analyte concentrations required for good LC with reasonably symmetrical peaks, [\text{A}_{\text{ads}}] \ll [\text{M}_{\text{ads}}], and the latter concentration can be considered constant, we obtain
Equation 3.32:
\log{k_{\text{A}}}=\log{K_{\text{A,ads}} \Phi}
= \log{(K_{\text{AM,ads}} \Phi [\text{M}_{\text{ads}}]^{n_{\text{A}}/n_{\text{M}}})}-\frac{n_{\text{A}}}{n_{\text{M}}} \log{[\text{M}_{\text{mob}}]}
The first term on the right-hand side of this equation is numerically equally to the logarithm of the retention factor at unit concentration of modifier (\log{k_{\text{A},1}}).
Equation 3.33:
\log{k_{\text{A}}}=\log{k_{\text{A},1}}-\frac{n_{\text{A}}}{n_{\text{M}}} \log{[\text{M}_{\text{mob}}]}
Thus, the competitive-adsorption model for LSC leads to a log-log model for retention as a function of modifier concentration.
1.2. Modifier and applicability #
This means that the effect of small percentages of a strong modifier (such that [\text{A}_{\text{ads}}] \ll [\text{M}_{\text{ads}}] holds) can have a very strong effect on retention. Increasing the modifier from 1 mM to 2 mM may have the same effect as increasing it from 10 mM to 20 mM from 100 mM to 200 mM. In the extreme case of the strongest modifier, water, very small amounts dissolved in the non-polar solvents can have dramatic effect on the retention, as well as on peak shape and repeatability. This is one of the nuisances of classical NPLC. The technique is still used for separation of oil samples, but it is not attractive for (biological, environmental, etc.) samples that may contain water.
In most of the contemporary applications of NPLC the stationary phase is no longer “bare” silica. Instead, it may be a chemically bonded phase (CBP), in which polar functional groups are attached to the surface. While this may increase the robustness of the system and shorten the equilibration time, it also leads to lower selectivity. High selectivity in NPLC may be obtained by creating chiral CBPs for the separation of enantiomers (see Module 3.13 in the book).
The most-practical contemporary mode of NPLC is hydrophilic-interaction liquid chromatography (HILIC).
READ SECTION 3.6.2
Hydrophilic Interaction Chromatography
1.3. Hydrophilic interaction chromatography (HILIC) #
Water is an enemy in classical NPLC on silica stationary phases, but it is turned into a friend in HILIC. The mobile phase is a moderately polar organic solvent, such as acetonitrile, with a low concentration (typically a few percent) of water. In combination with a polar adsorbent this results in a dynamic water-rich stationary-phase, a water layer, covering the surface. HILIC, coined by Andrew Alpert [1], is more efficient than classical NPLC and it is compatible with gradient elution, but it is only applicable to highly polar analytes that are retained on the water-rich stationary phase.
The behaviour of a polar CBP in combination with water/ACN mixtures as mobile phase is illustrated in the figure below. When the water concentration in the mobile phase is high (left-hand side of the figure), the CBP is predominantly solvated with acetonitrile. The mobile phase is more polar than the stationary phase (i.e. RPLC, see below) and moderately polar analytes are strongly retained.
Figure 2. Graphical simplified comparison of three different stationary-phase chemistries used in HILIC.
Retention in HILIC is complex. Different stationary phase chemistries are used to obtain specific selectivities. Diol phases have been described to be chemically polar, with strong hydrogen-bonding capabilities and suitable for moderate retention of small polar analytes like carbohydrates, peptides and small biomolecules [2]. Amide phases have suggested to be less prone to irreversible adsorption, which is favorable for neutral analytes [2,3]. Due to their neutral chemistry, these phases are also said to require less salts in the eluent which enhances MS compatibility. Zwitterionic phases are reported to offer the thickest aqueous layer [4, 5, 6].
EXERCISE 1
Which of the following statements correctly compares HILIC and normal-phase liquid chromatography (NPLC)?
This is correct! HILIC specifically leverages the formation of a water layer to act as a dynamic stationary phase. This is certainly not true for NPLC.
This is unfortunately incorrect. NPLC certainly is not the standard LC mode. Of the remaining statements only one is correct!
2. Retention modelling #
Perhaps one of the most useful conclusions from Lesson 1 was Equation 1.39, which showed us that separation can be influenced through the efficiency (N), selectivity (\alpha) and retention (k. The efficiency was extensively discussed in the first lessons and is not highly dependent on employed LC mode. The different selectivities were just discussed above and of course to a degree also the retention. However, it is the latter that is of interest here.
In NPLC, HILIC and RPLC, the retention of analytes is strongly influenced by the composition of the mobile phase, particularly the fraction of organic solvent, typically denoted as \phi. To describe this dependency, empirical models are frequently employed, providing practical tools for retention prediction and method development. An example of such models is shown for two analytes in Figure 3.
Figure 3. Schematic illustration of RPLC (left) and HILIC (right) retention on a moderately polar column.
In Figure 3, the retention factor of two analytes is plotted as a function of the organic mobile-phase fraction, \phi. The resulting curves can be obtained through measurement of retention facturs for the different analytes at different \phi-levels. A model can then be fitted through it. One such model is the quadratic model:
Equation 3.49:
\ln{k_i}=ln{k_0}+S_1 \phi + S_2 \phi^2
Here, \ln{k_0} is the extrapolated retention factor at 0% modifier (\phi = 0), and S_1 and S_2 are empirical coefficients of the model.
In the HILIC situation, at high concentrations of ACN and very low concentrations of water (right-hand side of Figure 3), the CBP is predominantly solvated with water. The mobile phase is more polar than the stationary phase (thus, HILIC is a form of NPLC) and only highly polar analytes are retained.
HILIC separations can be performed on bare silica stationary phases, polar CBPs and (zwitter-) ionic phases. Both hydrogen bonding and coulombic interactions may play a role, giving rise to a complex retention mechanism.
READ MODULE 3.5
Reversed-phase liquid chromatography
3. Reversed-phase liquid chromatography #
Reversed-phase liquid chromatography (RPLC) uses the octadecyl silica (ODS) phases discussed in Lesson 5. The result of a reaction with a monofunctional octadecyl dimethyl silane is illustrated again schematically below.
EXERCISE 2
Which of the following statements are correct?
This is correct! Well done!
Unfortunately, this is not correct. Three statements are correct.
The mobile phase in RPLCD starts with water (weakest eluent), to which one (or sometimes more) organic modifiers and possibly additives (e.g. salts, buffers) are added.
Figure 4. Schematic illustration of octadecyl silica (ODS) chains on a silica stationary-phase particle.
Equation 3.49:
\ln{k_i}=ln{k_0}+S_1 \phi + S_2 \phi^2
Here, \ln{k_0} is the extrapolated retention factor at 0% modifier (\phi = 0), and S_1 and S_2 are empirical coefficients of the model.
3.1. Properties and advantages #
RPLC is the Jack-of-(nearly-)all-trades in liquid chromatography. Mobile phases, as well as sample solvents, can vary from water to water-miscible modifiers (such as methanol, ACN, or tetrahydrofuran, THF), to even non-water-soluble polar organic solvents. This creates enormous flexibility (i.e. applicability) in terms of analytes that can be separated by RPLC and in terms of retention times, which can be varied over (many) orders of magnitude by just altering the mobile-phase composition.
Separation can be achieved for all but the most-polar analytes (such as carbohydrates), for which retention may be insufficient even with pure water as the mobile phase. Apolar analytes, such as triglycerides, may be separated using strictly organic mobile phases, although in such non-aqueous reversed-phase (NARP) LC selectivity limited.
(Small) ions usually show little retention in RPLC. Retention of weak acids or bases may be achieved by suppressing their ionization through adapting the pH (within the specified range for the column). Retention of any ion may be achieved by adding an ion-pair reagent to the mobile phase (treated in Lesson 7).
Because there are no strong interactions in RPLC, efficiency can be high and equilibration fast, provided that there is no detrimental effect of residual silanols.
3.2. Limitations #
There are, of course, limits to the power of RPLC. For very polar analytes it does not offer sufficient retention and HILIC (see above) has developed into the preferred method. For very non-polar analytes selectivity is limited and NPLC is preferred.
ODS-silica stationary phases can be used across a limited pH range. For conventional materials this is typically 2 < \text{pH} < 7, but some newer materials cover a broader range. This is especially relevant for suppression the ionization of basic analytes, so that they can be retained in RPLC.
3.3. Effect of modifier on retention #
The volume fraction of organic modifier (\phi) added to water has an exponential effect on retention. A log-linear model (also frequently known as the linear solvent strength, LSS model) is approximately valid across the most-meaningful range of retention factors (0.5 < k < 20). This can be written as
Equation 3.46a:
\log{k}=log{k_0}-S \phi
or
Equation 3.46b:
\ln{k}=ln{k_0}-S' \phi
Equation 3.46b with the natural logarithm is more practical when we start discussion gradient elution. See Module 3.4. In the log-linear model k_0 is the extrapolated retention factor in pure water and S or S' is a slope parameter.
3.4. Iso-elutropic mixtures #
The effect of \phi (and the slope parameter S) tends to be greater if the molecular weight of analyte molecules increases. Elution order reversals often occur when the amount of organic modifier is varied, resulting in changes in selectivity. Such changes are even greater when varying the nature of the modifier, for example when substituting methanol for acetonitrile (ACN). Occasionally ternary mobile phases containing two different modifiers (e.g. water with ACN and methanol) are used to ensure adequate selectivity.
Figure 5. Schematic illustration of octadecyl silica (ODS) chains on a silica stationary-phase particle. Based on [7].
ACN is a stronger modifier than methanol. To obtain iso-eluotropic mixtures, i.e. mobile-phase compositions that yield the same retention on average, lower concentrations of ACN suffice. Tetrahydrofuran (THF) is an even stronger modifier. Figure 5 illustrates approximate iso-eluotropic compositions in RPLC.
EXERCISE 3
Considering Figure 5, which solvent mixtures are expected to be iso-eluotropic with 70% methanol and 30% water in RPLC?
This is correct!
This is incorrect. If you draw a vertical line at 70% in the figure, you can see that about 56% of ACN and 46% of THF are iso-eluotropic.
If you mix these two solvents in a 1:1 ration you obtain a mixture of 28% ACN, 23% THF and the remaining 49% water, which is also expected to be iso-eluotropic.
Concluding remarks #
In this lesson, we introduced principal modes of liquid chromatography: normal-phase (NPLC), hydrophilic interaction (HILIC), and reversed-phase (RPLC). Each mode operates on distinct retention mechanisms governed by the polarity of the mobile and stationary phases, as well as specific molecular interactions such as adsorption..
We emphasized the importance of mobile-phase composition, particularly the role of organic (or aqueous) modifiers, in controlling retention and selectivity. A basic understanding of these mechanisms enables rational selection of chromatographic conditions for a broad range of analytes.
In the next lesson, we will focus in a different mode of liquid chromatography that targets ionic interactions.
EXTENSIVE EXERCISE
For two low-molecular-weight analytes a retention log-linear retention model has been established in RPLC. For Analyte 1 \ln{k_0} = 4.05 and S = 7.34 have been obtained, whereas for Analyte 2 the values are \ln{k_0} = 5.31 and S = 10.02. Your chromatographic system is characterized by N = 5000 and t_0 = 1 min. For the peak area of both peaks you may assume \text{Area} = 0.05.
Part A
Sketch the retention model for the two analytes.
Retention factors for the two analytes can be obtained from Equation 3.46b.
Fill in Equation 3.46b for two to four different values of \phi (e.g. 0.2 and 0.4) and draw a line through the resulting points on your graph. If you do the exercise on paper, just take a ruler and draw a line. If you do it using Excel, fit a trendline.
Your sketch should look something like this.
Part B
Sketch the expected isocratic chromatograms at \phi = 0.35, \phi = 0.47 and \phi = 0.60.
To do this, you will need to calculate the expected retention factor (k), retention time (t_{\text{R}}; Equation 1.12), peak standard deviation (\sigma_{\text{t}}; Equation 1.26) and peak height (\text{Height}; Equation 1.25) for each analyte.
The following equations are useful:
\text{Height}=\frac{\text{Area}}{\sigma \sqrt{2 \pi}}See Lesson 1 for the other equations.
For \phi = 0.35, you should find t_{\text{R},1} = 5.40, \sigma_{\text{t},1} = 0.076, \text{Height}_1 = 0.21, t_{\text{R},2} = 7.07, \sigma_{\text{t},2} = 0.100, \text{Height}_2 = 0.16.
To sketch the chromatogram for example on paper, draw a baseline with time (x-axis) and intensity (y-axis) axes. Locatie the peak centers using the retention times, and then point out the heights and widths of the peak (W_0.5= 2.35\sigma; at the base the width is roughly 4\sigma). You can of course also use Excel and plot Gaussians using NORM.DIST (and use the PDF).
You should get the following data, with the corresponding chromatogram sketches. Note that the peaks co-elute at \phi = 0.47.
References #
[1] A.J. Alpert, Hydrophilic-interaction chromatography for the separation of peptides, nucleic acids and other polar compounds, J. Chromatogr. A, 1990, 499, 177-196, DOI: 10.1016/S0021-9673(00)96972-3
[2] B. Buszewski, S. Noga, Hydrophilic interaction liquid chromatography (HILIC)-a powerful separation technique, Anal. Bio. Chem. 2012, 402, 231-247, DOI: 10.1007/s00216-011-5308-5
[3] M.A. Guarducci, A. Fochetti, A. Ciogli, G. Mazzoccanti, A Compendium of Principal Stationary Phases Used in Hydrophilic Interaction Chromatography: Where Have We Arrived?, Separations, 2023, 10(1), 22, DOI: 10.3390/separations10010022
[4] Y. Guo, N. Bhalodia, B. Fattal, I. Serris, Evaluating the Adsorbed Water Layer on Polar Stationary Phases for Hydrophilic Interaction Chromatography (HILIC), Separations, 2019, 6(2), 19, DOI: 10.3390/separations6020019
[5] P. Jandera, P. Janás, V. Skeriková, Jirí Urban, Effect of water on the retention on diol and amide columns in hydrophilic interaction chromatography, J. Sep. Sci. 2017, 40(7), 1434-1448, DOI: 10.1002/jssc.201601044
[6] D.V. McCalley, Hydrophilic‑Interaction Chromatography: An Update. LCGC North America, 2020, 38(3), 112–122, [LINK]
[7] P.J. Schoenmakers, H.A.H. Billiet and L. de Galan, J. Chromatogr. A, 1981, 205, 13-30, DOI: 10.1016/S0021-9673(00)81809-9
