Anyone who analyses samples by capillary gas chromatography has most likely used a split inlet. Although simple in concept, there are a few things to keep in mind to get the best analytical results possible.
Split inlets were invented to address the issues originally faced with the invention and adoption of capillary columns. Wall coated open tubular (WCOT) capillary columns offer much higher plate count, and therefore much higher separation power than the previously used packed column counterparts. However, they suffer from two major deficiencies: lower capacity and the need to produce narrow initial band widths even though the column flow rate is low.
The available surface area per meter of column is much smaller on the smooth walls of a glass or fused-silica capillary column than on the packing materials typically used in packed GC columns. Due the related loss in efficiency and difficulty in coating stable, efficient thick films on the smooth fused-silica surface, capillary column stationary phase thicknesses rarely exceed 5 µm; they are typically ≤ 1 µm even in large-bore capillary columns. The high surface-area supports (typically ranging from 0.5 - 5 m2/g) on which stationary phases are coated for use in packed columns will yield higher masses of stationary phase per linear length of the column as well as a higher total amount in the column. Since the solute capacity is related to the absolute amount of stationary phase present, capillary columns are at a severe disadvantage compared to packed columns. I imagine (i.e., have not seen any related studies) that having higher mass stationary phase per length of the initial part of a column is even more important than total amount of phase because solute concentrations will be highest (peak widths are narrowest) there and overload will manifest more dramatically there than later in the column when the peak width spreads.
Figure 1: Stationary phase overloading in GC. In gas-liquid partitioning (as opposed to gas-solid adsorption), overloaded peaks are fronted and their the peak maxima shift to later retention times. Peaks of different polarities will have different overload thresholds (where peak width broadens by 10%). For example, peak B broadens and shifts more than peak A even though the amount on-column increases <1/4 as much as peak A.
Figure 1 compares peak shapes and retention shifts for overloaded peaks and their non-overloaded counterparts. Notice how the second peak (different functionality than the first) overloads much easier and more dramatically than the earlier one. Overloading is very solute specific. Capacity for a “friendly” solute that is compatible with the stationary phase might be several orders of magnitude higher than one that is not. Capacity is also related to temperature, although one would not usually be changing temperatures in order to address capacity problems.
Table 1 compares typical packed and capillary column characteristics including capacities. The example packed column has 400 times more stationary phase per unit length than the capillary column and 27 times more in total. Solute capacity is often stated as mass solute/total mass stationary phase, but it seems more realistic approach is to state capacity in terms of mass solute/mass stationary phase per meter (or cm) since the solute-stationary phase interactions are localized and not happening through the full length of column all at once.
Table 1: Comparison of packed and capillary column characteristics including solute capacities. Capacity for a given solute is related to the mass of solute per mass of stationary phase, however capacities for solutes of different polarities can be quite different by orders of magnitude.
The capacity limit for a given solute in a given stationary phase (at a specified temperature) is commonly defined as the loading of a solute “on column” that causes a 10% loss in efficiency (N) or resolution (Rs) relative to its non-overloaded condition. According to my calculations, this corresponds to a 5.4% increase in peak width (half height, wh or base, wb). In the example of Table 1, the capacity of the packed column for a given solute would be approximately 400 times more than for the capillary column. The only convenient way to accomplish this feat is through split injection.
Capillary column peak widths are narrow (typically 0.5 – 5 sec) and typical column outlet flow rates are low (0.5 – 5 mL/min at 298 oK and 1 atm). Depending on the liner chosen, split inlet internal volumes are typically several hundred µL to 1 mL. Packed-column inlet volumes are similar. With typical packed column flow rates, one inlet volume is cleared in a couple seconds, which is fine for the packed column. However, at 1 mL/min capillary column flow rate, clearing the same volume would take 60 sec, which is far too wide an initial peak width for capillary columns unless some form of focusing is used (thermal or solvent).
Split inlets provide both a convenient way to reduce the amount of sample reaching the column and a higher total gas flow rate through the inlet to minimize initial peak widths. The amount of injected sample reaching the column can be adjusted by changing the split ratio.
The split ratio in gas chromatography is typically defined as
Split ratio is usually stated as an integer. So with 30 mL/min vent and 1.1 column flows, one would state the split ratio as 27:1 or simply 27, instead of 27.3. As such, split ratio is a nominal descriptor; used as a simple “setpoint”.
Split ratio is not a proportion of the injected sample being vented versus reaching the column - and sometimes you may want to know that. Assuming that sample is quantitatively transferred with no discrimination (not always a good assumption), one can estimate the portion of sample reaching the column as
With a low split ratio of 5:1 then, 16.7 % of the sample reaches the column (1/6), not 20% that some might have assumed based on the inverse of the split ratio (1/5). At a ratio of 100:1, 0.99% reaches the column. The differences between the “split ratio” and the actual proportion of original sample being vented versus that reaching the column usually has no practical impact on an analysis, since calibration standards would experience the same splitting. The resulting response factors would reflect whatever splitting is actually happening. However, if one is attempting to determine a real solute capacity, for example, one must know the amount of solute “on column” and using the proper relationship is important (especially at low split ratios).
In the next part we discuss split liners and flow measurement.
This blog article series is produced in collaboration with Dr Matthew S. Klee, internationally recognized for contributions to the theory and practice of gas chromatography. His experience in chemical, pharmaceutical and instrument companies spans over 30 years. During this time, Dr Klee’s work has focused on elucidation and practical demonstration of the many processes involved with GC analysis, with the ultimate goal of improving the ease of use of GC systems, ruggedness of methods and overall quality of results. If you have any questions about this article send them to email@example.com