I am having problems with IP391 analysis(aromatics in middle distillate with RI detection). The mobile phase is heptane and the columns are either silica or amino. I am using a Shimadzu RID-10A detector, which is about 7 years old. The standards used to calibrate the method contain cyclohexane, xylene, methylnaphthalene and phananthrene. Shortly after the aromatic constituents pass through the detector the baseline begins to dip and continues until some time after the final peak has eluted, but eventually returns to the same steady level as before injection. I initally tried changing columns, with no effect. It does not depend on whether or not the injection valve is returned to its 'fill' position. If a blank (heptane only) is injected there is no baseline disturbance. If I leave the columns out, the phenomenon still occurs, but more quickly. I have cleaned the detector using the method recommended by the manufacturer. I have also checked that the lamp output is within specification.

My suspicion is that there is a leak between the sample and reference sides of the RI cell but does anyone have any other suggestions? Shimadzu are being singularly unhelpful.

Hello Malclom

defective solenoid valve or flow line blockage on outlet tubing may cause sample entering the reference cell. You may check this by closeing reference cell with blind plugs.

Good luck

Thankyou. I have bypassed the solenoid valve and that solves the problem, so clearly I need a new valve.