HPLC Products

Spikes often look like very sharp peaks. Zooming in will reveal them to be only a few data points wide, which is far too fast for an HPLC peak. So they cannot come from sample components.

They must therefore be a function of something happening in the detector, either as a high frequency light disturbance in the flowcell, or as a result of electronic spikes.

Spikes which are at regular time intervals and are of similar size usually are cause by a small air bubble in the flow cell. The bubble enters the cell at the inlet at the bottom and quickly finds it way to the outlet. However it requires the input of energy to convert a spherical bubble into a long thin bubble able to pass down the outlet pipe, so it sits over the outlet. If it blocks the outlet, the pressure in the flowcell builds, the bubble becomes smaller, and then with a rush the solvent moves past it and out of the waste pipe. It is this rush and the expansion of the bubble which causes the spike.

The solution is to block the waste pipe for around 1-1.5 seconds at a flow rate of 1ml/min, and then release suddenly. This causes the bubble to compress, and hopefully be carried out of the cell as the eluent rushes out. It may be necessary to repeat this several times. Use a rubber septum or similar to block the pipe (NEVER your finger!), and check that the end of the cell outlet pipe is not sharp, or it could pierce the septum and then your finger when blocking the pipe. Don't hold for longer than about 1.5 seconds or you could break the flow cell, and for the same reason, don't use a flow rate higher than 1ml/min.

Spikes which give a random appearance, in terms of their timing and amplitude, are usually the result of an electronic fault within the detector. It can occasionally arise from an old and very worn out lamp. A replacement lamp soon eliminates the latter. A circuit board fault usually requires a service engineer to fix it, but be prepared for a big bill! The high labour rate (sometimes £150 per hour incl travelling) of a service engineer may be nothing compared to what their employer thinks they can charge for a circuit board!