I'm experiencing problems with my LC pump. When using low pressure mixing, I often observe proportioning errors, exhibited as retention times that are too short, but inconsistent. If I run premixed mobile phase in the A-reservoir, sometimes bubbles appear in the pump. The same solvent pumped from the B-reservoir works perfectly. The system uses low-pressure mixing.

I suspect that the inlet-line frit in the A-reservoir is partially blocked. Low pressure proportioning works by alternate opening of proportioning valves to allow solvent A or solvent B into the mixing chamber. For example, to make a 50/50 mixture of A and B, the A-valve would be open half the time and B would be open the other half. The cycle time for the valves is very short,usually less than a second. The pump pumps a steady flow rate. If the inlet frit in the A-reservoir were partially blocked, it would restrict flow into the mixer. Thus, when the A-valve opened, less than the desired amount of the A-solvent would enter. This also would create a slight vacuum. When the B-valve opened, solvent would flow freely, and a little extra solvent would enter the mixer to make up the pressure deficit created by low flow from A. Two simple tests should confirm this hypothesis. The easiest test is to remove the inlet frit on inlet line A. If the problem goes away, the frit is faulty. Alternatively, exchange the A and B solvent reservoirs and make the appropriate adjustments to the system controller to give the same solvent mixtures. If the retention times now are longer (resulting from a deficiency of the strong solvent, now in A), the A-reservoir is under-delivering. In either case, replace the frit. Don't try to clean the frit - this is only marginally beneficial and the cost of time spent would exceed the cost of a new frit. I like to use large-porosity frits, such as 10mm, so that there is little resistance to solvent flow.

The purpose of the frit is two-fold. First, it serves as a sinker to hold the inlet tubing in the bottom of the reservoir. For this reason, I prefer stainless steel over PEEK, but it is a personal choice. The second function of the frit is to provide minimal filtering of the mobile phase. This is not a substitute for normal mobile phase filtration, but rather to eliminate dust particles that might enter the reservoir. Finally, I prefer one of the frits that is designed to draw solvent from the bottom of the frit. Sooner or later, you'll be running a method where you need every last drop of solvent in the reservoir, and this type of frit will help you get out of a bind when you didn't prepare enough mobile phase.

This problem occurred a few times while I worked on intruments in the field. In many cases, as Dr. Dolan noted, the inlet filter was the culprit. However, I have seen some cases where the proportioning valve was actually malfunctioning. In the situation described by the reader, if the solenoid valve controlling position B was malfunctioning, that might also explain the situation as described. Allow me to explain.

Low pressure mixing operates as Dr. Dolan mentioned above. However, because this is an electronic and mechanical process, there are times when this mechanism fails. A possible scenario might be that the B position remains partially open at all times, either due to mechanical failure or even something partially occluding the valve, thus preventing it from closing all the way. If this was the case, then when solvent was pulled pre-mixed from position A AND if there was no solvent connected to the B position, then each time the pump's piston(s) pulled back, some air would leak into the mixing chamber built into the low pressure proportioning valve through position B. Likewise, the problem would most certainly go away if the solvent was pulled isocratically through position B, as in that case position B has to remain open anyway. Additionally, when the pump's mixing valve is utilized to provide the solvent mixing as opposed to using pre-mixed solvent, the same symptoms described might also occur.

I hope this helps!
--John Batts
Upchurch Scientific
Technical Support

In relation to the above advice. The proportioning valves can usually be dismantled (only the suspect valve is dismantled) and the sealing poppet inspected for contamination. If some contamination is apparent this can be gently wiped away and the valve re-assembled. If this procedure does not meet with any success then I suggest the complete valve be return to the manufacturer for repair or replacment.
Tim Colverd

I also recently experienced a proportioning valve malfunction. After tracing the problem to the valves, I disassembled each valve (Varian 9010 pump) and found two of the valves with broken springs. I assume that they break due to metal fatigue, one had one coil broken off and was partially opening and weak, the other had two or three broken and was hardly opening at all. They probably need to be checked regularly as a PM step.

Why would some one use a low pressure mixing technique. A 2 pump high pressure system as is available in the QuantaSep, Biocad or the Akta systems, give much better control and give better performance as far as chromatographic seperation is concerned

I'll argue with blanket statements about high-pressure vs. low-pressure mixing systems. I've seen good and bad examples of each.

I have also experienced proportioning problems with low pressure mixing valves (Kontron 425). In my case the problem was a small piece of dirt jamming the valve permanently open. By the way we use low pressure mixing because the reproducibility is much better at the >95% and <5% ends of the mixing process.