This edition of CHROMacademy Resolver focuses on the various approaches to achieving high efficiency separations in liquid chromatography.
High efficiency separations are characterised by narrow chromatographic peaks and increased peak capacity (the number of compounds separated with satisfactory resolution per unit time). More commonly known as Ultra High Performance (or Pressure) Liquid Chromatography (UHPLC), high efficiency separations are typically achieved using sub 2µm fully porous particles, which required specialist HPLC systems to cope with the increased back pressure generated, or latterly using superficially porous particles which generate UHPLC type efficiencies using conventional HPLC equipment. We will discuss high efficiency separation HPLC columns and equipment as well as comparing and contrasting the relative advantages and disadvantages of the various practical approaches to achieving reduced analysis times, increased resolution or a mixture of both using high efficiency HPLC techniques.
High Efficiency Separations – Benefits and Plate Theory
There are two main benefits associated with high efficiency separations;
Increased resolution of complex mixtures
Higher throughput via faster separations (i.e. the required resolution is achieved in a shorter timeframe)
The efficiency of a chromatographic peak is a measure of the dispersion of the analyte band as it travels through the HPLC system and column. In an ideal world, chromatographic peaks would be pencil thin lines – however, due to dispersion effects the peaks take on their familiar ‘Gaussian’ shape (or often slightly tailing shape in my case!).
The plate number (N) is a measure of the peak dispersion within the HPLC system and HPLC column, and reflects the column performance. N is derived from an analogy of Martyn and Synge who likened column efficiency to fractional distillation, where the column is divided into ‘Theoretical Plates’. Each plate is the distance over which the sample components achieve one equilibration between the stationary and mobile phase within the column. Therefore, the more (‘theoretical’) plates available within a column, the more equilibrations possible and the better quality the separation. The method of calculating column efficiency is shown below in Figure 1. A typical plate number for a 4.6 X 100 mm column with 5 µm particles is between 5000 and 8000. Of course from a practical perspective, for a given column length, higher plate counts dictate higher efficiency, narrower peaks and an increased chance of achieving the required minimum resolution.
Higher values for the Plate Number (N) are expected for each successive peak within a chromatogram. Later eluting peaks that look broad in comparison to early eluters may have a higher plate count – remember that efficiency is a measure of band broadening as a function of time
If each successive peak within your chromatogram does not have an increasing value for the plate count (N) then your system contains a large extra-column dead volume, which is dictating the overall system efficiency!
Figure 1 – Determination of efficiency
Factional Distillation Model of Efficiency Theory When ‘cracking hydrocarbon fractions, more ‘Plates’ there are – the narrower the distribution of carbon numbers from each trap (or plate) within a fractional distillation tower (see the animations in Figures 2 & 3). Therefore – the higher the number of plates (N) the narrower the ‘peak’ obtained from that trap – this can be directly related to the peak ‘efficiency’ in HPLC where a column with a high number of plates gives narrower (more efficient) peaks.
Figure 2 – High efficiency distillation process with many plates
Figure 3 – Low efficiency distillation process with only a few plates
Similarly – for a fractionating tower of a given length (L), the higher the number of plates, the lower the distance between each plate, shown as plate height in Figure 2. Therefore, for high efficiency separations, the plate number (N) should be high and the plate height (H) low. Note that plate height is often called – ‘Height Equivalent to a Theoretical Plate (HETP)’ These two terms are related through the expression: H = L / N (1)
High Efficiency Separations – Practical Uses HETP (H) is more convenient term for us to use because if we increase (N) (through the use of new high efficiency particle technology) but the column length (L) remains constant, a decrease in H will be observed. The benefit of this approach is that although analysis time, solvent usage etc will remain the same, the resolution will be increased, so increasing the number of components which can be separated using the column. Traditionally efficiency (N) has been increased by increasing the column length – this approach is limited however in that longitudinal molecular diffusion (peak dispersion) becomes the limiting factor as retention of the analyte within the column increases. The alternative use of high efficiency packing materials is to reduce the column dimensions (length and / or internal diameter) until comparable resolution is obtained to the original separation – the advantage being that a similar separation is achieved in a much shorter time, using much less solvent etc.
Practical interpretation of the van Deemter Equation
In order to appreciate how we can practically reduce plate height we need to understand the van Deemter equation and corresponding plots. The van Deemter curve shows height equivalent of a theoretical plate (HETP, H) (y-axis) against the eluent linear velocity (m) which is a function of eluent flow rate and column internal diameter (increasing flow rate will increase linear velocity for a given column internal diameter). This curve is a composite of curves made up from three individual effects which contribute to band broadening – Eddy Diffusion (A-Term), Longitudinal Molecular Diffusion (B-Term) and Mass Transfer Effects (C-Term). From the resulting composite curve a ‘theoretical optimum’ linear velocity (area shaded green in Figure 5) can be determined, and hence a flow rate chosen, in order to generate the narrowest, most efficient peaks. As HETP decreases, plate number (N) increases and chromatographic resolution should increase. Efficiency loss is considerable at very low (dictated by the B-Term) or very high (dictated by the C-Term) eluent flow rates, with an optimum practical minimum value lying somewhere in-between.
When the goal is to decrease analysis time (whilst maintaining the required resolution), the challenge is to find a set of HPLC conditions and hardware whose optimum HETP values occur at higher flow rates without the usual loss in efficiency. We should be mindful here that increasing eluent linear velocity (flow rate) will increase the pressure drop across the column ( and increased the observed system back pressure). For those of you who have generated van Deemter curves for you analysis you will most likely find you are operating slightly post-minima in the ‘practical optimum’ linear velocity. Here we sacrifice a small amount of plate height (efficiency), but markedly reduce analysis time by operating at higher flow rates. As we will see, the use of sub 2mm and superficially porous materials
In order to adjust the composite curve such that the minimum is at increasingly higher linear velocity, we must first identify the A, B and C-Terms and understand how their individual contributions to the composite curve can be reduced.
The A term is related to eddy diffusion and is near constant over a range of linear velocities
The B term is related to longitudinal molecular diffusion and decreases approximately exponentially with increasing eluent linear velocity – it is the predominant cause of loss in efficiency at lower flow rates. At high flow rates the B-Term contribution to efficiency loss can be negligible.
The C term is related to the mass transfer process and increases linearly with increasing eluent flow rate – it is the predominant cause of loss of efficiency at higher flow rates. At low flow rates the C-Term contribution to efficiency loss can be negligible.
Figure 4 – Van Deemter composite curve
Figure 5 – Van Deemter composite curve
with optimum velocities indicate
Eddy Diffusion (The A term) The A term in the van Deemter equation is often used to describe variations in mobile phase flow or analyte path within the chromatographic column.
Eddy diffusion itself relates to the fact that an analyte molecule can take one of many possible paths through the column. These multiple paths arise due to inhomogeneities in column packing and variations in the particle size / shape of the packing material. In fact, the Eddy diffusion in the van Deemter equation is often called the “packing” term as it reflects the quality (homogeneity) of the column packing.
Eddy diffusion can be minimized by:
Selecting well packed columns
Using reduced diameter packing materials
Choosing packing material with a narrow particle size distribution
Important considerations for high efficiency HPLC:
Analyte molecules move through different paths within the column packing material – leading to reduction in efficiency
Larger differences in path-length occur, and hence loss in efficiency, with larger particles and with larger particle size distributions
Figure 6 – Animations illustrating the relative contributions of eddy diffusion to analyte band broadening when using packing materials of varying diameter
Longitudinal Diffusion (The B Term) A band of analytes contained in the injection solvent plug will tend to disperse in every direction (both axially and longitudinally) due to the concentration gradient at the outer edges of the analyte band.
The B term in the van Deemter equation is related to the dispersion experienced by analyte molecules due to these concentration gradients. This phenomenon is known as ‘Longitudinal Molecular Diffusion’ because inside the column, the greatest scope for broadening is along the axis of flow. Longitudinal diffusion will occur within all system tubing but will be most pronounced in the column.
Longitudinal diffusion occurs whenever the HPLC system contains internal volumes that are larger than necessary:
Tubing length too long and /or too high internal diameter
Incorrectly connected or non-zero dead volume fittings
Incorrect column nuts and ferrules
Detector flow cell which is too large for the required analytical sensitivity
Longitudinal diffusion has a much larger effect at low mobile phase velocities (flow rates); therefore it is reduced when using high linear velocities. However these effects still need to be bourn in mind when designing and using high efficiency HPLC equipment as gains in efficiency can be easily mitigated by small, but additive, extra column system volumes. This is especially true when using reduced internal diameter (3mm i.d. and below) columns
The plot below depicts the loss of efficiency / increase in plate height associated with the B-Term by running at very low eluent velocities (green peak) as opposed to the much more efficient peak shapes associated by running at higher eluent linear velocities (blue peak).
Figure 7 – Peak shape differences at high and low eluent velocities considering the contribution from Longitudinal Molecular diffusion
Mass Transfer (The C Term) The C term in the Van Deemter equation is related to analyte ‘mass transfer’ and accounts not only for the dispersive convection in the mobile phase between and within the packing material pores but also for sorption and desorption of the analyte from the stationary phase.
Rather than having a unique residence time in the stationary phase, analyte molecules show a spread of retention times. As analyte molecules move through the stagnant mobile phase within the pores of the stationary phase support material they do so by diffusion only (i.e. the mobile phase is not moving with the eluent flow but is ‘trapped’ within the pore). Analyte molecules will be sorbed onto the stationary phase surface at different depths within the pore. As the diffusion process into and out of the pore is a fixed rate process, this will cause analytes to elute from the pore at different times depending upon the ‘depth’ at which they sorbed onto the analyte surface – broadening the band of analytes as they travel through the column and resulting in loss of efficiency / increased plate height.
Mass transfer effects can be minimized by:
using smaller diameter stationary phase particles
using low mobile phase flow rates (low linear velocities)
increasing column temperature (at high temperatures the diffusion processes speed up and the differences in elution time from the particle are reduced)
When smaller particle are employed the much shallower pore depth / length of through pores and the distance between pores ensure the analyte band stays much tighter and yields high efficiencies with small plate heights
Figure 8 – Loss of efficiency associated with
differing penetration depths of deep pores (C-Term)
High efficiency at high flow rates – practical approaches
The main contributors to band broadening at increased eluent velocities are the A and the C-Terms of the van Deemter equation (Eddy Diffusion and Mass Transfer respectively) Therefore, by reducing the individual contributions of the these terms, high efficiency separations will be possible at increased eluent flow.
Increasing Temperature to achieve improved efficiency One mechanism by which this can be achieved is increasing eluent temperature, which alters mass transfer kinetics (C-Term) by lowering the mobile phase viscosity and increasing the analyte kinetic energy so that differences in analyte residence times within the stationary phase pores is reduced. The resultant composite curves when α-napthol was chromatographed on a 50 x 4.6mm, 1.8µm column in a mobile phase of 60% acetonitrile / 40% water at various temperatures is shown below in Figure 9. Note how at higher temperatures, the optimum linear velocities are generated at higher mobile phase velocities and yield flatter composite curves with much shallower post-minima increases (i.e. the system retains good efficiency even at vastly increased liner velocity).
Figure 9 – Composite van Deemter curves
generated using α-napthol at various temperatures
Care should be taken when increasing separation temperatures as analyte degradation may be promoted and co-elution or peak inversion (swapping) may occur. Not all analytes will be affected to the same extent by changes to the separation temperature, although in nearly all cases a reduction in retention time will be observed, the rate of decrease will vary from analytes to analyte. One further benefit associated with increased temperature is a reduction in system back pressure
Reducing Particle Size to achieve improved efficiency A second, increasingly popular mechanism used to increase efficiency involves the use of reduced particle size packing materials, typically in the sub 2 micron range. This approach not only reduces analyte dispersion by reducing pore depth and distance between pores,(C-Term) but the smaller particles (and their associated smaller particle size distribution)s also reduce analyte dispersion by minimising Eddy diffusion ( A-Term). Shown in Figure 10 are the composite curves achieved by analysing the same analyte under the same conditions using various packing material particle sizes.
The collection of composite curves nicely demonstrates two benefits of High Efficiency HPLC when employing small particles.
Figure 10 – Composite curves generated for various particle sizes »»»
Firstly, as the particle size decreases, the H value minimum also decreases. If all separation conditions and column dimensions are kept the same with only the size of the particles reduced, increased efficiency is observed whilst the selectivity remains constant. This is best utilised where previous separations of complex mixtures have been optimised but full resolution has not quite been achieved. Figure 11 shows how the selectivity, α, for two different classes of compounds remains consistent but the efficiency and resulting resolution both increase as particle size reduces.
Figure 11 – Selectivity remains constant as Efficiency and Resolution increase in line with decreasing particle size»»»
Secondly, as plate height reduces, shorter columns can be employed to achieve the same plate count (N). This reduction in column length leads to a reduction in analysis time and an increase in sample throughput and a corresponding decrease in solvent usage. This combined with the fact the sub 2µm particles do no exhibit the characteristic post-minimum increase of the composite curve associated with larger particles, analysis can be further sped up by increasing eluent velocity, often without compromising efficiency. Figure 12 demonstrates how analysis time can be reduced by employing smaller particles with shorter columns, the reduced plate height means that no reduction in overall plate number is observed. This approach may often also be combined with an increase in eluent linear velocity, without compromising resolution.
It is also important not to make any changes that will be detrimental to the B-Term, Longitudinal Molecular Diffusion. Whilst only at very low eluent velocities can the B-Term be reduced in order to increase efficiency, an increase in the contribution from the B-Term at high flow rates can serve to negate the enhancements of optimising the A and C-Terms. In fact, as we see shortly, only the very modern UHPLC systems, the lowest of extra column dead volumes, can be used for very high efficiency separations.
Figure 12 – Analysis time reduced using shorter columns at higher eluent
flow rate without affecting efficiency and hence maintaining resolution
Increased Back Pressure with Decreasing Particle Size Unfortunately this increase in efficiency when using smaller particles comes at the cost of increased system back pressure. Equation 3 demonstrates how the operating back pressure can be expected to increase as any one of the variables is changed.
When any of the numerators (viscosity, flow or column length) are increased the pressure increases, when they are reduced the pressure falls. When any of the denominators (specific permeability, column radius and particle size) are increased, the pressure falls, when they decrease the pressure rises. By reducing the viscosity (by way of increasing the separation temperature for example) and column length (the standard approach n high throughput separations), we would expect to see the system back pressure fall, but this is counteracted by a reduction in the particle size and this factor (along with column radius), is raised to the second power. If the particle size is reduced by a factor of 3, the pressure will increase by a factor of 9x and column length would need to be reduced by a factor of 9 in order to maintain pressure parity. In order to cope with these pressures a UHPLC (ultra high PRESSURE liquid chromatograph) system will be required in most cases which are designed to routinely operate at increased system back pressure (see later).
We are currently experiencing an unheralded time when so many manufacturers are launching brand new high pressure instrumentation platforms. Shimadzu have the Nexera range, Agilent the Infinity, Perkin Elmer the Flexar, Waters the Acquity, Dionex the Ultimate and Thermo have the Accela range, to name some of the main vendors. What changes have been made to conventional HPLC instrumentation that facilitates high efficiency separations? There are four main areas which differentiate UHPLC from conventional HPLC instrumentation. Some of the conventional HPLC instruments on the market may be able to lay claim to one or even a couple of the UHPLC pre-requisites, but in order to carry out high efficiency separations and fully reap the benefits of either higher throughput and / or increased resolution, a UHPLC must be able to achieve all four.
Pump Adaptations The pressure limit commonly associated with conventional HPLC instruments is 400bar and the primary system limiting factor is the pump. At pressures of above 400bar the traditional sapphire pistons are liable to snap, especially if their mechanical stability is compromised by chips or scratches caused by mechanical abrasion of particulate materials in the eluent system. UHPLC instruments employ pistons manufactured from stronger materials such as silica carbide which can withstand the 1000+ bar (14,500+ psi) pressures required and some permit the use of high flow rates at these incredibly high pressures. In many cases piston seals are also re-designed and manufactured from mechanically less deformable materials to cope with increased pressures without leaking. The final main mechanical change to the pump is motors that drive the pistons, which are routinely updated to cope with the excessive pressures. Conventional motors which drive a cam to move the piston had the unfortunate tendency to catch fire when working against increased hydraulic resistance and new piston design principles with direct drive motors replace the conventional units.
Gradient Dwell Reduction In order to generate the fastest separations, ballistic gradients are often employed (rate of increase of one of the constituents is greater than 50% per minute) It is not uncommon for some gradients to run from 5% solvent B to 100% solvent B in less than 40 seconds in very high efficiency separations. Gradient mobile phase composition is adjusted using a proportioning valve when using quaternary pumps or by adjusting the relative individual pump flow rates for a binary pump. To operate a ballistic gradient, changes in mobile phase composition produced by the pump, must reach the column in the minimum possible time. The time between the pump adjusting the mobile phase composition and it reaching the column is referred to as either delay or dwell volume (VD). The dwell volume, usually exhibited as an isocratic hold at the start of a gradient run where no hold is programmed in the gradient table, is equal to the total volume of all mixing chambers, unions and capillaries between the point at which the mobile phase composition is changed and the column. It is designated as a volume rather than a time as a time delay is flow rate dependent. It is not uncommon for modern UHPLC instruments to quote dwell volumes of around 20µL, in contrast to conventional instruments whose dwell may be many milliters. Most modern UHPLC systems have configurable mixing chamber volumes, and hence dwell volumes, where their volume can be increased when the lower volume chambers have been shown to be inaccurate. Where methods are to be published, transferred to another laboratory or employed on other systems it is of upmost importance that the system dwell volume is quoted (and either matched or accounted for using isocratic holds or delayed sample injection). In gradient HPLC, the change per unit time in eluent composition can have a dramatic effect on the selectivity of a separation and this needs to be matched carefully between instruments by taking into account changes in the instrument gradient dwell volume. A step-by-step guide to calculating your systems dwell volume and adjusting gradient parameters between instruments can be found on CHROMacademy LC / HPLC Channel, Instrumentation of HPLC, Solvent Pumping Systems, Experimental Determination of System Dwell Volume.
Reduction in extra column volume The analyte band in the mobile phase stream will tend to diffuse in the axial directions diluting the analyte within the eluent stream (van Deemter) B-Term. This results in an increase in peak width and decreased analytical sensitivity. The amount of analyte broadening is directly related to the interstitial volume within the column and the volume of the system outside of the column that the sample encounters. This latter volume is generally defined as the extra column volume (VEC) and is the total volume contributed by all system components and capillary tubing from the point of sample injection to detection, which are not directly involved in the separation process – from injector to detector flow cell.
Assuming isocratic conditions, analyte peak width is proportional to retention time, with later eluting peaks being considerably wider than their early eluting counterparts. When conventional HPLC columns are used (4.6 x 150mm, 5.0µm) the band broadening exhibited is predominately due to analyte dispersion within the large interstitial volumes within the column as it elutes slowly under lower flow rate conditions. With reduced UHPLC column dimensions (2.1 x 50mm, 1.8µm) and increased eluent linear velocity, the extra column volume becomes the predominant factor in band broadening and considerable efforts must be made to reduce these volumes such that high efficiency separations can still be achieved.
The main components whose dimensions have been reduced to achieve low dead volume UHPLC systems include;
reducing the volume of the injection hydraulic pathway,
typically 1 - 5µL
using reduced id and shorter capillary tubing, total typically 5 - 15µL
minimising the volume of the heat exchanger(s), typically 1 – 4µL
condensing the flow cell volume, typically 0.5 – 3µL
using a minimum amount of high quality zero-dead volume unions and couplings
Figure 13 demonstrates the effects of tubing length on peak efficiency »»»
Extra column volume can be effectively reduced by ensuring that all connections are correctly made and are zero dead volume. Due to traditional end-fitting materials such as PEEK (polyetheretherketone) materials not being able to seal under the increased pressures generated in UHPLC, stainless steel fittings are commonly used. As the stainless steel fitting cannot deform to seal into the fittings in which they are installed it is essential that the correct type of fittings are used. Nuts are manufactured to fit one specific column end fitting type and the nut and the user should check the compatibility between the column and end fittings used. Figure 14 demonstrates what happens when three commonly utilised styles of end fittings are not correctly matched.
The use of optimised lengths and internal diameter of capillary tubing is also important to retain the highest efficiency in UPLC systems. Tubing length is important, but the internal diameter is of much greater importance and this is demonstrated by the Aris – Taylor equation shown below;
Where; F is the flow through capillary tubing, L is the length of the capillary tubing, DM is the analyte diffusion coefficient and id is the capillary tubing internal diameter
The overall dispersion of the analyte band is proportional to the length of the capillary tubing but is proportional to the capillary tubing internal diameter raised to the 4th power, thus producing a profound effect upon analyte peak dispersion.
This same principle may be applied to flow cell volumes, which are minimised in UHPLC systems to avoid similar dispersion effects.
Figure 14 – Schematic showing the effect of incorrectly matching fittings »»»
Improvements in Detector Speed The final area where improvements have been made to UHPLC systems is in the data acquisition rate of the detectors used. This is sometime referred to as the scanning rate and relates to the speed at which the detector can record data. An artefact of UHPLC separations is the narrow peak widths associated with efficient peaks. In order to generate a truly representative peak to be used for a qualitative separation a minimum of 15 readings across the peak width are required. When a chromatographic peak is going to be used for quantitative measurements a minimum of 25 measurements across the peak width should be made. The peak width is usually defined as the distance between the points where lines drawn tangentially to the sides of the chromatographic peak intersect the baseline. In conventional HPLC separations, where peak widths of 30 seconds are common, a data acquisition rate of 1Hz (1 measurement per second) often suffice. In UHPLC separations when peak widths of 2secs are more common, a data acquisition rate of a 12Hz or 0.08secs or 0.001min are required. If running really high efficiency separations and are generating peak widths of 0.5 seconds (or even less in some cases) a data acquisition rate of 50Hz or 0.02secs or 0.0003min are necessary.
Figure 15 demonstrates the effects of using different detector acquisition rates on analyte peak width and resolution. The effects of ‘peak broadening’ due to low data acquisition rates are obvious and the phenomenon is often mistaken for physical peak dispersion!
Figure 15 – Effects of data acquisition rate on chromatographic peak width and resolution
A recent addition to the ‘high efficiency’ range of columns available is the so-called Superficially Porous Silica (SPS) particle. These particles consist of a solid bead or core, surrounded by a shallow porous layer and are shown in Figure 16 compared with a conventional fully porous material.
Figure 16 – Morphology of Superficially Porous and Fully Porous Silica Particles »»»
Superficially Porous Particle Development SPS particles were first developed at the end the 1960’s for the analysis of large molecules, typically monomers and small polymers. Large molecules tended to exhibit very broad peaks due to the hindered mass transfer kinetics (C-Term) they experienced diffusing in and out of the pores. By reducing the pore depth to 0.5µm from 2.5µm or greater of a fully porous particle, the mass transfer kinetics were significantly improved and more efficient peaks resulted. SPS particles then re-emerged in the late 1990’s as a solution to the broad peaks and loss of efficiency experienced when analysing the increasing popular bio-molecules. They consisted of an overall particle size of 5µm but only a small portion was porous and available for penetration with a pore diameter if 300Å or 30nm.
Recent Advances in SPS Technology It is only recently that this technology was adapted slightly such that it could be used for smaller molecules by reducing the overall particle diameter to 2.6 – 3.0µm with a porous layer of 0.35 – 0.5µm and a pore diameter of 80 - 120Å or 8 – 12nm. These particles have been found to generate van Deemter curves comparable to sub 2µm particles but as they are closer to 3µm in size they generate much lower operating back pressure – 40-50% less in most cases. Figure 17 compares the van Deemter curves for a fully porous 3.5µm particle, a fully porous 1.8µm UHPLC particle, a SPS 2.7µm particle and two SPS 2.6µm particles.
Figure 17 – Typical van Deemter curves for SPS and fully porous particles
The fully porous 3.5µm particle exhibits the usual loss of efficiency (increasing plate height) at increased eluent flow rate, whereas all SPS particles tested demonstrated remarkably similar curves to the 1.8µm fully porous UHPLC particle.
Figure 18 compares the separation of several reversed phase test probes using sub 2mm and superficially porous silica particles. Overall run time, efficiency and resolution are comparable, however the SPS column achieves this at sub 400 bar - making it useful for those with conventional HPLC equipment
Figure 18 – Superficially porous particle demonstrating equivalent efficiency
at significantly lower back pressure than an equivalent Sub 2mm particle
65% A:35% B, A = 0.2% formic acid, B = methanol Isocratic, flow rate 0.5 mL/min
Injection:
1 μL
Column temperature:
26 °C
Detector:
sig = 220, 4 nm, ref = off
SPS Particle Efficiency – Mass Transfer or Diffusion Led?
Initial conclusions that the reduction in plate height was due to an improvement in the mass transfer kinetics were speculative. Smaller molecules, (< circa. 600Da ) do not suffer the same hindered mass transfer kinetics as larger molecules so alternative or supporting reasons for the reduction in plate height and increase in efficiency have been offered. The supporting increase in efficiency is found to lie in the very narrow particle size distribution that can be achieved when manufacturing particles with a solid core. These highly uniform particles can ultimately form a densely packed bed within the column. The A-Term, eddy diffusion, in the Van Deemter equation shows that efficiency is dependent on not only the particle diameter but the magnitude of the particle size distribution.
SPS Particles – Particle Size Distribution When HPLC packing materials are manufactured they are ‘sized’ using sub-sieve sizing technology and a range of sizes are collected and packed into a column with a specified particle diameter. Not all particles will be the nominal specified size diameter but will be a distribution with the mean, median and mode of the distribution matching the specified diameter. Figure 19 illustrates example normalised particle size distributions for a range of fully porous particles and an SPS particle.
Figure 19 – Fully porous and SPS normalised particle size distributions The absolute deviation from the mean decreases as the fully porous particle diameter decreases but even the 1.8µm particles show a much wider distribution compared to the larger 2.7µm SPS particle.
SPS Particles – Advantages of Larger Particles for Facile Analysis Another disadvantage of sub 2µm particles is that due to their size and wider particle size distribution, the standard frit used to retain the material within the column has had to be reduced. A frit with a porosity of 0.5µm is necessary to retain the packing material, which is a substantial departure from the standard 2µm frit employed on >3.5µm conventional HPLC columns. This reduction in frit porosity has lead to extra steps being implemented in order to filter all eluents and samples in order to stop them plugging the column frits.
On the surface it appears that the SPS particles, with only a shallow porous shell, will have much lower sample capacities and lead to column overloading (usually manifested by peak fronting) but this does not seem to be the case. This may be the result of the SPS narrow particle size distribution leading to columns which are highly uniformly (and densely) packed leading to an (anecdotally) 70% surface area availability compared to a column of the same dimensions packed with fully porous 3.5mm particles.
Conclusions So returning to a question posed previously, do we really need to invest the very large quantities in a UHPLC system? Moving forward manufacturers will improve their processes and be able to reduce the particle size distribution associated with sub 2µm particles thus reducing their plate height even further. Methods developed on UHPLC systems can usually be transferred with ease to conventional systems as long as sufficient safety measures are built into the method such as;
an initial isocratic hold to account for dwell volume differences
extra resolution to account for loss in efficiency
Transferring methods from traditional HPLC to UHPLC instrumentation can be more problematic and require greater re-validation. The cost of UHPLC instruments will fall as 3D-UHPLC, UUHPLC or whatever the next generation of UHPLC instruments is called are introduced. While companies, especially in the current economic climate, cannot afford to replace all conventional HPLC instruments with UHPLC equivalents, SPS particles have served to bring UHPLC like efficiencies to the mass market. Whilst tubing capillaries, detector flow cells and mixing chambers can all be replaced to alleviate some extra column affects, the pressure ceiling and data acquisition rates will still preclude even the most highly modified conventional HPLC systems from achieving the efficiencies and separation speed of ultra high pressure chromatography.
The following subjects are covered in full multi-media at CHROMacademy.com
The Theory Of HPLC
Introduction (1.5hrs)
Chromatographic Parameters (3hrs)
Band Broadening (3hrs)
Column chemistry (4hrs)
Reverse phase (partition) chromatography (6hrs)
Ion-Pair Chromatography (3hrs)
Normal phase (absorption) chromatography (3hrs)
Gradient HPLC (3hrs)
Quantitative and Qualitative HPLC (3hrs)
FAST HPLC (4.5hrs)
Theory and Instrumentation of GC
Introduction (1.5hrs)
Chromatographic Parameters (3hrs)
Band Broadening (3hrs)
Gas Supply and Pressure Control (2hrs)
Sampling Techniques (4.5hrs)
Sample Introduction (5hrs)
GC Columns (5.5hrs)
GC Temperature Programming (3hrs)
GC Detectors (2.5hrs)
Instrumentation of HPLC
Mobile Phase Considerations (3.5hrs)
Solvent Pumping Systems (4hrs)
Autosamplers (4.5hrs)
Detectors (4.5hrs)
