This section may seem trivial, but extracolumn volumes are critical to optimum HPLC performance. Dead volume in tubing, flow cells, and diverter valves from the injector to the column head can lead to band broadening. Dead volumes from column outlet to the detector can destroy a perfectly good column separation and remix separated elution bands. Inlet tubing is much more sensitive to room- temperature fluctuations than is the column itself. These effects can cause baseline drifting and cycling and are easily overcome by snapping rubber tubing split lengthwise over the inlet tubing to provide an insulating dead space.
Most of the stainless steel tubing used for these critical runs is 0.009-in.
internal ID, called ten-thousands tubing. Less critical areas such as flush valve outlets and splitter valves use 0.02- and 0.04-in. tubing, and it is important to separate these tubing types so that the wrong type of tubing is not picked up by mistake. Special 0.005-in. stainless steel tubing is available to use with microinjectors and columns. The larger sizes—0.010, 0.020, and 0.040 in.—can be cut with a tubing cutter and polished with a flat file. The finest tubing should be purchased precut and prewashed. Inert polymeric PEEK tubing usable to 4000 psi is available for use with inert HPLC systems with titanium or polymeric wetted surfaces used in applications that will not tolerate extractable metals, such as enzyme purifications.
Tubing is connected to system components using compression fittings made up of a ferrule-and-screw arrangement (Figure 2.8aandb). The screw and ferrule are placed on the tubing line, placed in the connection on the component in which the line will be used, and the ferrule is compressed onto the line by tightening the screw. Try not to overtighten the fitting! Usually, a quarter-turn with a wrench after the fitting is fingertight is sufficient to keep the fitting from leaking. Test it at pressure; if it leaks along the line, tighten it a bit more. Other fitting components are zero-dead-volume unions (Figure 2.8c) to connect two pieces of tubing with fittings and three-way diverter unions used in splitters and flush valves.
Fittings to be used with unions will need to be prepared in situ if the union is to remain a zero-dead-volume connector. An in-line splitter or diverter uses
HPLC TUBING AND FITTINGS 19
Ferrule
Ferrule
Compression screw
(a) (b)
(c)
Compression screw 1
16-inch tubing
FIGURE 2.8 HPLC compression fittings: (a) male fitting; (b) female fitting; (c) zero- dead-volume union.
0.010-in. tubing in the flow path and larger tubing, usually 0.02-in. tubing, in the diversion path. The fourfold increase in cross-sectional area diverts much of the flow away from the in-line path. For instance, a solvent splitter can be used to send part of the column effluent to a secondary detector while the main flow is sent to an ionizing interface for a mass spectrometer.
One of the most useful applications for tubing, unions, and fittings is to prepare a column blank (Figure 2.9). HPLC systems get dirty over time and must be
Compression fittings
f/f union
0.01-inch tubing
FIGURE 2.9 Column blank.
20 THE HPLC SYSTEM
cleaned. Buffers accumulate on check valves and plungers; sample precipitates in the injector line; and secondary detector flow cells become fogged by effluent materials. Most wetted surfaces of an HPLC system, except the column and mass spectrometer, can be cleaned with water, organic solvents, and 6N nitric acid.
Note that columns are an exception: Columns never should be washed with nitric acid! They can be washed with certain organic solvents if buffer is first washed out with water. This may seem obvious and trivial, but I have seen many columns plugged with precipitated buffer, on two occasions I have seen columns ruined with 6Nnitric acid, and on one memorable occasion I saw a $1000 silica protein purification column totally dissolved with Trisma base. To avoid these problems I advocate use of a column bridge. This 5-ft coil of 0.010-in. tubing is equipped with a compression fitting and unions and is used to replace the HPLC column. The column is washed out with water, removed, capped, and set aside. It is then replaced with a column bridge and flow is diverted away from sensitive detectors, such as conductivity, electrochemical, or mass spectrometers.
The remaining HPLC system can now be washed with organic solvents such as methanol, acetonitrile, or dimethyl sulfoxide, then with water, and finally with 6N nitric acid. I usually recommend that the system with the column bridge be washed next with water for 2 hours at 1 mL/min and then over a weekend at 0.1 mL/min before removing the bridge and replacing it with the column. This treatment once a month will prevent a multitude of check valve and injector problems and save a lab many visits from an instrument repairperson.
3