Column Overloading - An issue of some sensitivity

Injecting too much onto a chromatographic system can distort the chromatography, diminishing efficiency and reducing assay sensitivity. Reducing column diameters will reduce sample consumption, solvent consumption and also improve sensitivity, however this  needs to be done with care as column overloading can result. For columns that are equivalent in the type of stationary phase that is being used, and in length, a simple squared relationship exists between the different diameter columns and the injection volume. Thus reducing the diameter of the column by a factor of two will require a reduction in the injection volume of a factor of 4. Where different stationary phases and column lengths are being used the relationship is more complicated. The issue with injecting too much onto any form of chromatographic system is that of overloading the column. This article will highlight the symptoms, and identify the sources of this situation.
There are two types of overloading that can occur: mass overloading of the analyte as well as volume overloading. In addition to these two types of overload, care must be excercised with the selection of injection solvent as one with excessive elutropic strength can have deleterious effects which mimic volume overload.

Mass loading considerations

The peak shape associated with mass overloading is often referred to as a ‘shark fin’. This abnormal peak shape results from the bulk of the analyte eluting earlier than expected as a results of the stationary phase not having the capacity to retain the analyte.

Figure 1. Schematic example of mass overloading on a HPLC column, showing a characteristic shark fin.

When considering the acceptable mass loading of a column, it is important to appreciate that within any chromatographic system there are only so many active sites where retention can occur. Once all of these sites are consumed, any subsequent analyte molecules travelling through the column will not be retained by the stationary phase and will elute earlier than the analytes which were retained prior to reaching the overload state. This effectively shifts the centre of mass for the eluting peak further down the column, resulting in a reduced retention time. It is important to note that the surface area per pore volume plays a role in this loading capacity and it is not dcitated by surface area alone. Packing possessing a low pore volume  results in more mass being packed into a column, and the amount of surface thereby increases with the mass of packing in a column. Thus, the loading capability of a column is dependent on the structure of the stationary phase media and the surface area. The surface area on typical packing materials can vary from 4 m2/g for non-porous beads up to 400 m2/g for highly porous media. It is evident from these surface area numbers, coupled with the array of pore structures, that the amount of material that can be retained will vary substantially and will result in a very large, typically a hundred-fold difference in the loading capacity between columns of the same dimension.
It is possible to estimate the amount that can be injected on different columns by using the following formula, which gives the phase ratio, a measure of the loading capacity of a column [1];
Equation 1
Ap – surface area of packing (m2/g)
Vp – specific pore volume (mL/g)
εi – interstitial fraction
ρ - density of substrate material (g/mm3)

Table 1 provides an idea of what can be loaded onto a column before overloading occurs, however it should be noted that these are very approximate and the loading capacity for each analyte / column configuration should be investigated experimentally.

Table 1. A guide to how much can be loaded onto a column before overloading occurs.
When evaluating column mass overloading, consideration should also be taken of the retention mechanism. For purely hydrophobic interactions, the above statements are valid, however where a molecule has an associated charge, the charge interactions also have to be considered. Charged molecules typically load at levels10-50 times less than comparable neutral molecules. The understanding here is that if the analyte is adsorbed to the surface through a hydrophobic interaction then the charge will be exposed, effectively repelling other analyte molecules from the surface. For porous stationary phases the repulsion effect could result in analyte molecules being effectively excluded from the pore structure. In a purely hydrophobic retention mechanism, this clearly does not happen. To alleviate this issue, altering the pH to ensure that the analyte is in a neutral form will result in a larger loading capacity.
Another consideration is the elution mechanism and whether an isocratic or gradient system is being employed. A gradient system allows for refocussing of a peak as it progresses down the column, whereas an isocratic elution does not. In terms of overloading, both elution techniques will ultimately suffer from too much mass loading, however the refocussing that is observed in a gradient elution will help to alleviate some of the observed peak effects.

Volume overloading

The second type of overloading that can occur is due to injecting too much solvent. This results in tailing peaks, Figure 2. The tailing is caused by the analyte molecules having different elution start times due to the injection solvent taking time to load onto the column. This phenmomena is more noticeable under isocratic conditions, where there are no focussing effects, but can to a certain extent be disguised when using gradient elution.
Figure 2. Schematic of column overloading caused by an increase in injection volume
The type of injection solvent employed can also limit the amount of analye that can be loaded onto a column. For an efficient separation to occur, the analyte molecules must immediately partition into the stationary phase upon injection, and then be controllably eluted by the B solvent. If the injection solvent has too high of an elutropic strengththe analyte molecules will start to migrate down the column before the loading of the sample is complete. The injection solvent at that point is controlling the elution and not the the bulk mobile phase. The most common reason for use of too strong of an injection solvent is samples with poor solubility in weaker solvetns. To reduce this effect the analyte molecules have to migrate away from the injection solvent, which can happen through diffusional processes or through any form of retention on the stationary phase. Reducing the injection volume will favour the retention, also improving the peak shape. An added benefit of improving the peak shape is  the increase in peak height and hence better sensitivity. An example of the effect of injection volume and solvent composition on the peak height is given in Figure 3. It should be noted that the injection solvent may be a very relevant consideration when there are solubility issues or if some form of sample preparation is being performed on the original sample. An example of this could be protein precipitation or SPE, which could result in the injection solvent having a highly elutropic solvent.

Figure 3. The effect of varying the injection volume on peak height for two types of injection solvent. Methanol / phosphate buffer (50/50 v/v), Indomethacin (100 ng/mL) injected onto C18 column (120 x 1.0 mm), data taken from [2]

Scaling injection volumes

Equation 1 allows the phase ratio of two different columns to be determined which will give guidance on the relative amount that can be loaded onto each column. This equation is quite complicated and a different equation can be used if the same type of column is being scaled down. Thus if the same packing material substrate and the same column length are being used then the following equation can be employed;

Equation 2

Since the concentrations of the samples are nominally the same, the Amount term can be replaced with an injection volume term. It can be seen therefore that there is an inverse square relationship between the column radius and the amount that can be loaded onto the column. This would suggest that there are substantial sensitivity disadvantages to using capillary columns, however since most of the detectors that are employed are concentration sensitive detectors, there is also an increase in the sensitivity of the instrumentation which will offset the loss in sensitivity. The HelpDesk [3] has discussed this in a previous article. In summary the following equation determines the relative response that is observed for different column i.d.’s
Equation 3
Where
M – mass loaded onto column
K – retention or capacity factor
N – chromatographic efficiency of the system (column)
V0 – column dead volume

Prep Chromatography

In some forms of chromatography, column overloading is acceptable and indeed the columns are specifically run in an overloaded state to ensure that an optimum yield is obtained. The obvious example here is the field of preparative and process chromatography where the separation technique is used to allow for isolation of one or more of the analytes. The aim of this form of chromatography is to isolate specific analytes and although resolution is important as this will affect the final purity of the isolated analyte, it is typically the amount of analyte that can be separated per unit time that is the critical component. This means that resolution and amount that is separated have to both be considered, whereas most analytical separations it is the speed of analysis and the resolution that are critical parameters.
Other areas of chromatography where overloading may routinely occur is in trace and also impurity analysis, where in order to achieve the sensitivity of the main analyte, a large injection volume is used. For the impurity analysis the main analyte being detected is the impurity not the main component within the sample.

Conclusion

Reducing column diameters to reduce sample consumption, reduce solvent consumption and also to improve sensitivity needs to be done with care as scaling has to be done on the amount of sample injected as well as the sample volume. For columns that are equivalent in the type of stationary phase that is being used, and in length, a simple squared relationship exists between the different diameter columns and the injection volume. Thus reducing the diameter of the column by a factor of two will require a reduction in the injection volume of a factor of 4. Where different stationary phases and column lengths are being used the relationship is more complicated.
Other important variables to consider when looking at reducing overloading effects are the injection solvent, the mass loading and also the nature of the molecule that is being separated. Care has to be taken with ionisable compounds to ensure that peaks are not excluded from the pore structure. It is also useful to know that highly elutropic solvents can be used but only in small measure.

References

1.     K.K. Unger. Porous Silica: Its Properties and Use as Support in Column Liquid Chromatography. Elsevier. Amsterdam. 1979. D. 170.
2.     Handbook of Analytical Separations Vol 4 Bioanalytical Separations ed. Ian D. Wilson p. 287
3.     Chromatography Today, 20, Nov 2014