Gradient HPLC for Practitioners

RP, LC-MS, Ion Analytics, Biochromatography, SFC, HILIC

Stavros Kromidas

• Publisher: Wiley-VCH
• Year: 2019
• ISBN:
  9783527812769
• Edition: 1
• Topic: Physical Sciences
• Subtopic: Analytic Chemistry

This practical guide for analytical scientists explains the use of gradients in liquid chromatography. The fundamentals of gradient separations, as well as the most common application scenarios are addressed, from LC-MS coupling to biochromatography to the separation of ionic substances. Throughout, this handy volume provides detailed hands-on information for practitioners, enabling them to use gradient separation methods reliably and efficiently.

Part 1
Principles of Gradient Elution

Chapter 1
Aspects of Gradient Optimization

Stavros Kromidas (translated from German by Steve Ross)

1.1 Introduction

Gradients are versatile and therefore find wide application. For example, gradients are just as essential in method development of unknown samples as for quantification at trace levels. The theoretical background of gradient elution is quite complex, because what happens in the column during gradient elution, compared to isocratic separations, is affected by more factors; these sometimes act in opposite directions or are multiplicative.

Herein, we will focus on selected aspects of the optimization of gradient separations in RP chromatography in deliberately simple form. Other important aspects of the gradient such as theory, equipment, and troubleshooting are left to other sources [1–4]. First, we briefly describe the action of a gradient in the column, then using some basic formulas we discuss the characteristics/features of the gradient. On the basis of this, possibilities for optimization of the following objectives will be shown: low detection limit, high peak capacity, sufficient resolution, and the shortest possible retention times. Finally, there is a summary with some basic rules and recommendations.

1.2 Special Features of the Gradient

In HPLC, interactions of different strengths between the analytes on the one hand and eluent components and the stationary phase on the other usually occur during separation. In the case of isocratic separations there is a predetermined, constant eluent composition, consequently during chromatography an interaction of constant strength takes place between the eluent molecules and the phase material.

What happens now in a gradient run? During gradient separations the strength of the mobile phase increases, consequently its interaction with the stationary phase also increases during the gradient run. The rule in RP chromatography is: the more organic, nonpolar/hydrophobic the eluent becomes during the separation (more % B, ACN or MeOH), the stronger its interaction with the organic, nonpolar surface of an RP material becomes – it is indeed “like with like,” that means the nonpolar ACN or MeOH molecules naturally “like” for example nonpolar C18 alkyl groups.

In the course of a gradient, because of the ever increasing concentration of ACN/MeOH molecules, the substance molecules become subject to increasingly strong competition in their interactions with the C18 alkyl groups. Because of this, the substance molecules are increasingly forced to leave the stationary phase faster, go into the mobile phase earlier and thus elute earlier compared to isocratic separations. With 100% MeOH or ACN at the end of the gradient even the very hydrophobic components of the sample elute, maybe even persistent organic contaminants that may have accumulated on the surface of the stationary phase – as a side effect the column is flushed at the same time.

Focusing on the peak form, with gradients we have two opposing trends. On the one hand, the later the peaks elute, the more the substance zone is subject to dispersion processes in the column and thus band broadening initially increases – analogous to isocratic separations. On the other hand, the acceleration of the migrating substance zone increases to the same extent, since the elution strength of the eluent permanently increases from the beginning to the end. As a result, these effects compensate each other and with a gradient we usually have narrow peaks. Note that with a gradient the concentration of the elution band constantly increases leading to lower band broadening in comparison with isocratic separations, consequently resulting in low detection limits.

This is true both for the front and for the end part of the chromatogram, in the ideal case the peak width remains constant. For this reason, in conjunction with the gradient speaking of a “plate number” is not allowed. The plate number, a measure of band broadening, is defined only for isocratic conditions. The phenomenon described here means, among other things, that in practice a reduction in packing quality and a suboptimal hardware (system dead space), which with isocratic separations leads to broad peaks, is not as noticeable with gradient separations. Even with “poor” equipment and “poor” columns chromatograms from a gradient elution look good, especially if the gradient is steep and ACN is used as the organic content of the eluent – a welcome fact for sample chromatograms in manufacturer’s brochures …

Positive from the user perspective is, that simple gradient separations using 20–50 mm columns on conventional equipment generally prove to be no problem, at least as far as the peak shape is concerned. Also the advantage of smaller particle sizes, for example 2 or 3 μm particles compared to 5 μm particles, is less relevant in many applications. In the case of a difficult matrix, 3.5–5 μm material should therefore initially be considered. Unless one has to separate a large number of very similar analytes – then of course the separating efficiency of ≤ 2 μm particles also becomes relevant for gradients. In this context, it is also pointed out that as the eluent permanently becomes stronger (= nonpolar), the migrating substance molecules at the end of a peak, i.e., at the trailing edge, move faster than those at the beginning of the peak as the later eluting molecules of the substance band are always pushed “forward” faster. This fact, known as “peak compression,” has the effect that in gradient separations tailing is rarely observed. Peak symmetry is about 10% better compared to an isocratic run with equivalent eluent composition (H.-J. Kuss, personal communication).

1.3 Some Chromatographic Definitions and Formulas

Let us now consider some chromatographic definitions which are known from theory – which, by the way, was developed originally for GC and much later for isocratic LC separations. The derivation of the formulas used below is omitted, they are only used to elaborate the consequences for practical optimization. For a more detailed discussion, see references [2–4] and in particular [1].

The resolution R is, in simplified form, the distance between two peaks on the baseline. The retention factor k (formerly the capacity factor k′) is the ratio of the time a component spends in/on the stationary phase and in the mobile phase, that is the quotient of the net retention time

(time spent in the stationary phase) and the flow-through or dead or mobile time t₀ and t_m (time spent in the mobile phase). It thus represents a measure of the strength of the interactions of these components on this column under these conditions:

. However, the retention factor is not constant for a gradient. Very high at the beginning (with 100 or 95% water/buffer the substances literally “stick” to the beginning of the stationary phase), it becomes less during the separation and at the end of the gradient is very small. With 90 or 100% MeOH or ACN, the substance molecules hardly have a chance to stay on the stationary phase, because the competition for the “attraction” of the C18 group has now become huge. Put simply, with a gradient from 100% water/buffer to 100% MeOH/ACN, the k value at the beginning is virtually infinite – in some references numbers between 3500–4000 are given – and at the end almost zero. Since the k-value changes during gradient elution, a k^∗ value (or

) was introduced to take account of this particular feature [1]: this is the k-value of a component when it is just in the middle of the column.

Although the need for such a term to describe the gradient may be questioned, the k^∗ value is used here because it has advantages for our deliberations. And that the interactions, and therefore a measure for them, a retention size, is important for optimization considerations, is clear – however such a term may be defined.

The separation factor α is the quotient of the retention factors of two components that one wishes to separate, k₁ and k₂, and describes the ability of the chromatographic system to separate these two components. In the literature, different formulas are used for R and k^∗. However, they are quite similar and ultimately lead, when the focus is on the practice, to similar numerical values and thus to similar propositions.

Here is an example: in Eq. (1.1) for the second term (the selectivity term), in addition to (α − 1) the terms ln α or α − 1/α are also to be found in the literature. Assuming a α-value of 1.05, the following numerical values for the selectivity term are found: 0.048, 0.049, and 0.050. However, these different numbers affect the value of the resolution only in the second decimal place.

Five simple equations are given below. They are sufficient to draw conclusions for practical optimization.

(1.1)

(1.2)

(1.3)

(1.4)

(1.5)

With:

R	Resolution;
N	Number of theoretical plates, defined fundamentally for isocratic conditions;
α	Separation factor (formerly selectivity factor);
k	Actual (measured) retention factor of a component;
k^∗	Reten...

Table of contents

• Cover
• Table of Contents
• Preface
• The Structure of the Book
• Notes on Contributors
• List of Contributors
• Part 1: Principles of Gradient Elution
• Part 2: Specifics of the Gradient in Different Elution Modes
• Index
• End User License Agreement