Using Selectivity to Optimize UHPLC Separations

Published on: 
Pharmaceutical Technology, Pharmaceutical Technology-10-01-2007, Volume 2007 Supplement, Issue 5

Ultrahigh pressure liquid chromatography maximizes efficiency, but, as defined by the resolution equation, the stationary phase is still a crucial consideration when attempting to resolve mixtures of compounds.

Ultrahigh pressure liquid chromatography (UHPLC) has altered liquid separation procedures, but it has not changed the fundamental theory. Although UHPLC enables the use of sub-2-μm high-performance liquid chromatography packing materials, it is beneficial for a practicing chromatographer to consider the overall benefits and limitations of this technology. The author discusses the principles of chromatographic separations and explains how specific parameters such as stationary-phase selectivity and efficiency affect chromatographic resolution.



Since the late 1960s, when modern HPLC became a viable tool for practicing chemists, continual advancements have been made in the technology. Considering analytical-scale HPLC columns in particular, over time there has been a movement from irregular to spherical particles, as well as an overall decrease in the size of the particles used in the packing material. For spherical silica particles, this shift has been from 10 μm to 5 μm, then to 3 μm, and most recently to sub-2-μm particles. When using a smaller particle-size packing material, it is possible to increase the efficiency, or number of theoretical plates, and expand the range of usable flow rates.

However, this technique results in proportionately higher back pressure as liquid mobile phase is forced through a much more tightly packed bed. Conventional HPLC systems are capable of handling pressures as high as 5000 psi, which until recently limited the usable particle size to approximately 3 μm, depending on flow rate, mobile phase composition, and column temperature. The advent of UHPLC, which uses instrumentation capable of handling back pressures greater than 14,000 psi, makes using columns packed with sub-2-μm packings possible. With UHPLC, extremely fast and efficient liquid separations are possible.

Figure 1: The resolution equation indicates that selectivity has the greatest influence on resolution.


The overall goal in HPLC, and now UHPLC, is chromatographic resolution, whether between target analytes or between an analyte and the sample matrix. Considering, as a guideline, the fundamental relationships given in the resolution equation (see Figure 1), one can better understand the process of resolving mixtures. The resolution equation comprises three terms: selectivity (α – 1),

and efficiency . (√N) Each of these terms is affected by the specific conditions chosen when developing an analytical method. How well one resolves the analytes and how quickly it is done depend upon the analyst's ability to control these three factors. Selectivity, which is governed predominantly by analyte interactions with both the stationary and mobile phases, is arguably the driving force behind separations because it affects resolution to the greatest degree. The smaller particles used in UHPLC primarily affect the efficiency, N, of the resolution equation. Although small particles can improve a separation, they are only one contributor toward the goal of resolution.

Figure 2: Efficiency, N, is inversely proportional to the particle diameter dp; therefore, as particle size decreases, efficiency increases.

Particle size and efficiency


Columns packed with sub-2-μm particle sizes can be used to increase resolution and shorten analysis times. How is this done? To better understand the effect of particle size, consider two more concepts in separation science—the theoretical plate and the van Deemter equation.

Figure 3: Efficiency, N, is directly proportional to the column length L; therefore, efficiency increases as column length increases.

As previously mentioned, particle size affects the efficiency term of the resolution equation. Efficiency is ultimately derived from the theoretical plate model of chromatography. Conceptually, a plate refers to one complete equilibrated transfer, or partition, of a solute between the mobile and stationary phases. N, therefore, is a qualitative term used to measure the number of theoretical plates in a given column or the extent to which a solute partitions between the mobile and stationary phases. In relation to particle size, N is inversely proportional to the particle diameter (see Figure 2). Put simply, as particle size decreases, efficiency increases, and more resolution is possible with the same length of column.

Figure 4: The van Deemter equation, an empirical equation describing the relationship between efficiency, as measured by theoretical plate height H, and linear velocity is governed by three cumulative terms: A eddy diffusion, B longitudinal diffusion, and C mass transfer.

Practically, this means that shorter analysis times through the use of smaller particle-size packings can be achieved by two means. First, N is directly proportional to the column length L (see Figure 3). Therefore, an analyst can decrease the length of the column as the particle size decreases, still maintaining resolution and shortening the analysis time. Second, and probably more significant, higher flow rates can be used without a substantial loss in resolution. The van Deemter equation is an empirical formula describing the relationship between plate height H (the height of one theoretical plate) and linear velocity μ (see Figure 4). This equation explains why higher flow rates can be used with small-particle columns without compromising resolution. Smaller plate-height values correspond to greater peak efficiencies because more plates or analyte partitions can occur over a fixed length of column. The van Deemter equation is governed by three cumulative terms: eddy diffusion A, longitudinal diffusion B, and mass transfer C. Eddy diffusion is caused by disturbances in the solute flow path and is largely unaffected by flow rate. Smaller particles give rise to less interstitial space and therefore, through a decreased A term, have higher overall efficiencies. Mass transfer is the movement of the analyte in and out of the stationary phase. Through this term, higher flow rates typically result in poorer efficiencies.

Figure 5: An empirically determined van Deemter plot for a biphenyl test probe demonstrates that sub-2-μm packings are minimally affected by flow rate, thus allowing higher flow rates to be used and resulting in faster analysis times.

If we look at an empirically determined van Deemter plot of efficiency versus flow rate, when using a 1.9-μm particle size (see Figure 5), we can see that, in practice, smaller particle sizes are minimally affected by higher flow rates. Column efficiency does not diminish when flow rate increases, as indicated by the relatively flat slope of the curve. Peak efficiency is comparable even when the flow rate increases to 1 mL/min on a 2.1-mm internal diameter column. This illustrates the most significant influence that small particles have on chromatographic separations—a much wider range of usable flow rates, which translates into significantly faster analysis times. This benefit, and the shorter column lengths needed for a given resolution, result in higher sample throughput without loss of separation quality.

Figure 6: A biphenyl stationary phase is more selective for aromatic or fused-ring compounds than a conventional C18 phase, providing improved resolution.

The role of selectivity

UHPLC does maximize efficiency (e.g., theoretical plates), but, as defined by the resolution equation, stationary-phase selectivity is still a crucial consideration when attempting to resolve mixtures of compounds. By choosing a stationary phase that produces optimum selectivity for the specific compounds of interest, one can maximize the benefit of UHPLC. For example, the use of a biphenyl stationary phase can greatly enhance a separation. A biphenyl stationary phase differs from that of an alkyl (e.g., C18) phase in that the phenyl rings present can employ pi–pi interactions. This creates a separation mechanism with improved selectivity for aromatic or fused-ring compounds (see Figure 6). When using a biphenyl stationary phase with a highly efficient 1.9-μm particle-size column, one can produce fast, highly selective separations of steroids (see Figure 7). A 1.9-μm biphenyl column can separate a test mix of seven hormones in less than 2 min, a feat not possible through C18 selectivity (1, 2).

Figure 7: A biphenyl stationary phase, combined with a 1.9-μm packing, makes possible a fast and highly selective analysis of steroids.

Another example of the advantage of a carefully chosen stationary phase for UHPLC is a separation using a pentafluorophenyl (PFP) propyl stationary phase. This fluorinated phase is extremely retentive and selective for organohalogens or other compounds containing basic or electronegative functionalities. The analysis of fluoroquinolones such as ciprofloxacin, norfloxacin, and enrofloxacin demonstrates this characteristic. When compared with C18 and cyanopropyl stationary phases, the PFP propyl phase is more selective and more retentive (see Figure 8). The US Food and Drug Administration recently announced import controls on seafood from China, where the use of fluoroquinolones in food animals is permitted. All incoming shipments are now detained at the border until proven to be free of drug residues. In this specific example, the use of LC–MS–MS is preferred. Using a PFP propyl stationary phase to heighten the retention of fluoroquinolones is advantageous in LC–MS analysis because it requires higher organic percentages to elute the compounds. With electrospray ionization (ESI), higher amounts of organic in the mobile phases help in desolvation, thereby lowering the detection limits. In addition, unwanted adduct formation can be reduced, and charge competition from matrix interferences are lowered as the fluoroquinolones are better resolved from the matrix components (3).

Figure 8: A pentafluorphenyl propyl stationary phase offers greater retention and selectivity for fluoroquinolones than either C18 or cyano phases, making it well suited for LC–MS analyses.

In a final example, using the same premise of heightened selectivity of a fluorinated phase toward halogenated drug compounds, benzodiazepines and two metabolites, a mix commonly assayed on a C18 column, can be resolved in nearly 4 min on a PFP propyl column (see Figure 9). To get the same level of selectivity from a C18 column, a shallower gradient would be needed, thus prolonging the analysis time. Because the selectivity of a PFP propyl stationary phase elutes the benzodiazepines in quick succession, a simple gradient allows for the earlier elution of the more polar metabolites, while maintaining a fast overall run time.

Figure 9: Fast, selective analysis of benzodiazepines is made possible by combining the speed of UHPLC with the enhanced selectivity of a pentafluorophenyl propyl stationary phase.


Although the sub-2-μm particle sizes used in ultrahigh pressure liquid chromatography make fast, highly efficient assays possible, selectivity is still a critical parameter to consider. The examples reviewed here demonstrate the impact of selectivity on overall chromatographic resolution and speed of analysis. Proper stationary-phase selection is the most important factor in optimizing resolution and, with sub-2-μm particles, allows the benefits of UHPLC to be fully realized.

Rick Lake is a pharmaceutical innovations chemist at Restek Corporation, 110 Benner Pike, Bellefonte, PA 16823, tel. 814.353.1300, fax 814.353.1309,


1. R. Freeman, R. Lake, and L. Nolan, "Enhanced Resolution of Endocrine Disrupting Hormones," The Restek Advantage 2006 (3), 8–9, 2006.

2. R. Lake, "Optimize Selectivity and Efficiency in UHPLC Separations," The Restek Advantage 2007 (3), 6–7, 2007.

3. R. Lake, "Simplified LC/MS/MS Analysis of Fluoroquinolones," The Restek Advantage 2007 (1), 6–7, 2007.