It has been just over 100 years since Russian botanist Mikhail Semionovich Tswett invented the separation technique that he named chromatography.1 This term combines two Greek words, χηρομα (chroma) and γραπηειν (graphein), meaning literally “color writing or recording.” Tswett defined his approach the following way: “A method which separates different components on an adsorptive column in a flow system is a chromatographic method.” Unfortunately, his approach did not attract much attention and remained unnoticed for a long time. Despite some experimental work during the 1930s, the major thrust came only after World War II. The seminal works of Archer J.P. Martin and Robert Synge that launched the era of modern gas and liquid chromatography were awarded the Nobel Prize in 1952.2 The following developments then led to rapid acceptance of this separation method.3 Today, chromatography is the third most widely used analytical technique, surpassed only by weighing and pH measurement.
The official definition of chromatography was coined by the International Union of Pure and Applied Chemistry (IUPAC) in 1993: “Chromatography is a physical method of separation in which the components to be separated are distributed between two phases, one of which is stationary (the stationary phase) while the other (the mobile phase) moves in a definite direction.”4 A number of developments have occurred in chromatography during the last 50 years. It is believed that the introduction of rigid stationary phases based on porous silica by Csaba Horváth, Georges Guiochon, and others at the beginning of the 1970s was the most important development that opened the wide field of current HPLC.5,6
Archenemy of chromatography: Diffusion
According to the definition presented above, chromatography is used to separate components of a mixture as best as possible. In simplified terms, the chromatographic process can be described as follows: The sample is injected into the packed column and the mobile phase drives it through the interstitial spaces between porous particles of the packing that represent the stationary phase. The mobile phase is always present both between the particles and within the pores. While the mobile phase in the interstitial voids moves, it does not move in the pores and remains stagnant there. Once the sample mixture passes by the packing particle, a concentration gradient or the difference between the concentrations in the stream and within the pores drives the components to enter the pores. Here, they interact with the surface functionalities. The strength of these interactions differs for each component. As the sample plug in the mobile phase passes, its concentration in the vicinity of that particle decreases, and the reversed concentration gradient forces the components to move back in the stream. The components that interact within the pores more strongly remain there longer than those that interact less or not at all. This process, which lies at the heart of adsorption chromatography, is repeated again and again until the component leaves the column. Thus, the difference in delay of components in pores enables the desired separation of the mixture.
In an ideal world, all the components of the sample would enter the pores quickly and at the same time and, similarly, they would become rapidly liberated from the pores. Therefore, the only difference would be the amount of time the components reside in the pores as a result of differential interactions. Under these conditions, the band of the separated component, called the peak, detected at the column outlet would be very narrow. The width of the peak defines column efficiency. Clearly, the Holy Grail of chromatography is to achieve separations featuring very narrow peaks that do not overlap. In other words, each component should leave the column completely separated from all the other components. Unfortunately, we do not live in an ideal world. The speed at which the components enter and leave the pores is controlled by their diffusion in the liquid phase and is strongly dependent on their size. The larger the component, the slower its diffusion. The slow diffusion then negatively affects the peak width, and the separation is less perfect.
The problem of diffusional mass transport in the stagnant mobile phase within the pores that leads to undesirable peak broadening has been pointed out by Martin and Synge. Thus, the history of modern chromatography can be characterized as a permanent war with slow mass transport.
Obviously, the simplest approach to avoiding diffusion with pores is to not have any pores or to use completely nonporous packings. Indeed, this approach has been demonstrated in academic research, and commercial columns packed with nonporous beads appeared on the market. These columns enabled very fast separations, even of large molecules. Despite this apparent success, the technology has not been completely embraced by the chromatographic community since diffusion is only one side of the coin. The other side is the need for interaction of individual components with the surface functionalities that define selectivity of the separation. The larger the surface available for these interactions, the larger the number of molecules that can interact simultaneously, and the more sample that can be injected and separated.
The nonporous packings bear functionalities exposed at their outer surface, which is only a fraction of the surface area exhibited by typical porous stationary phases. Therefore, only a small quantity of sample can be injected, and very sensitive detection is required to monitor the separation. However, this approach is useless for preparative separations such as those run in the biotechnology industry, where large sample capacity is essential. Clearly, waging war with diffusion by dropping off porosity is not a good solution. Thus, why not make the particles smaller and decrease the diffusional pass length?
Another problem: Backpressure
Launched in 1973, silica-based μBondapak (Waters Corp., Milford, MA) with a particle size of 10 μm was the first commercial column packing broadly used in HPLC. However, columns packed with 5-μm particles emerged only five years later and became the industry standard for several years. They were followed by 3.5-μm beads in the late 1990s. What is the driving force behind this development? The answer is simple: Improvements in technology have enabled the production of these particles and instrumentation facilitating their use. To better understand this, it is important to bear in mind that the column is only one part of the chromatographic system, which also includes pumps, injector, and detector. The smaller the particles, the higher the pressure that has to be applied to achieve the required flow rate. Hydraulics 101 teaches that pressure drop in a packed column, ΔP, is defined as
ΔP ~ FηL/A.dp2 (1)
where F is the volumetric flow rate, η is the viscosity of the mobile phase, L is the length of the column, A is its cross-section, and dpis the size of the packing. This equation shows that the backpressure in a column increases exponentially with the decreasing particle size, since the size of the interstitial voids through which the mobile phase must flow also decreases. However, the higher pressure requires use of adequate instrumentation. Until recently, all typical instruments worked at pressures of up to 40 MPa (ca. 6000 psi). In order to stay within the pressure limits, the column length L had to be shortened. However, the column efficiency is directly proportional to its length and indirectly proportional to the packing size. For example, a decrease to one-half in both length and particle size does not provide for any gain in efficiency, which remains unchanged. Yet, the overall column volume is also only one-half and, at the same flow rate, the separation is completed in a shorter period of time. The overall benefit is then faster analysis and higher throughput. However, this trend has certain physical limits since the decrease in column length cannot continue indefinitely. So, what’s next? Fortunately, chromatographers are creative people and came up with a few interesting solutions that are attracting a lot of attention. The following text will describe four approaches that emerged recently.