Dear all,

We are considering getting a HPLC system and hopefully a LC/MS. I am wondering if we should get a quaternary pump or a binary pump. Our applications are somewhat similar to a forensic type of application. We need to analyze a small sample and it is ideal if we can analyze a wide range of compounds in one run. I have used a binary pump and a reversed phase 4.6 mm I.D. column, but I want to develop the method further. For example, the use of narrow bore columns, on-line sample preparation and the use of three different polarities of mobile phases in one run. A quaternary pump seems to be more flexible for these purposes. However, I heard that it is less reliable in terms of delivery of solvents, in particular, those with low flow rate and may be more trouble prone. A binary pump seems to be generally reliable, even when it is used with a low flow rate.
I would appreciate any advice you could give me about this. Thank you.

I frequently work with flow rates around 0.2 mL/min (LC/MS)and prefer a quaternary pump because of the flexibility. I have only used Agilent's 1100, so I am not familiar with others. We qualify our pumps at several flow rates, and I think that this could be useful with any pump. Another possibility is get your hands on an old HP1090 with a tertiary pump. Those were very reliable.

One point you may wish to note when looking at binary vs quaternary pumps is the differences in dead volumes for these. This could become more an issue at low flow rates. Speaking from the point of the Agilent 1100 series the differences can be quite significant. One of the most significant differences arises from the ways the different systems mix the solvent before delivering it. In the cases of the binary system what appears to be a small empty column is used and in the case of the quaternary pump a chamber filled with ball bearings is used to cause turbulence, hence the mixing. A smaller dead volume exists in the binary system. In the Agilent 1100 capillary system only a binary system is available at the moment and I suspect one of the reasons for this is what I've just mentioned above. This is just one point to consider in the scheme of things. The LC/MS system we use utilizes an Agilent 1100 quaternary pump and is very reliable and support is excellent.

we have 5 HP/Agilent 1050s and 1100s, all quaternary, all excellent.

Thank you very much for all of your advises. It is really useful. I somehow gathered that a quaternary pump has no problem, in particular if it is used more than 200 microliter/min. of a flow rate. I still wonder if anyone have had other problems with a quaternary pump than those related with low flow rate? And also, has anyone had examples of analytical method, which shows that a quaternary pump is useful than a binary pump? One of discussions in the forum was talking about a problem of a multi channel gradient valve "sticking" by acetonitrile in a quaternary pump. I have used a HP1090 binary pump system with acetonitrile, but I did not have such a problem. Does it sound that a quaternary pump is more trouble prone? I would appreciate any comments on this. Thank you very much for your help.

Check out the Agilent 1100 OQ/PV documentation. The gradient accuracy spec for the quaternary pump is 1.5% and 0.7% for the binary pump.

As mentioned above the two pumps have a significantly different dwell volume because the pump itself is part of the dwell volume for a quaternary pump but not for a binary pump.

The only method that I have had any trouble transfering between quaternary and binary pump systems was a peptide map on a 3.0mm ID column. We had some transfer difficulity even after compensating for, then matching system dwell volumes.

The problem with the binary pump is you lose the flexibility of online mixing of trinary mobile phases. This greatly hinders our method development.

I recommend the quaternary pump unless your separations require the binary pump. Capilary columns or a very demanding separation might require you to switch to a binary pump.

We have 10 1100s and generally use channel B as a 100% orgainc channel. I have seen the "sticking" MCGV four times in five years. It has always been channel B leaking while filled with ACN. The problem is easily fixed/prevented by purging channel B with hot water.

Good Luck.

Thank you very much for your advice, Mr. Tom M. I am now inclining to a quaternary pump after wondering for a long time. I wonder if you could explain a little more about transferring problem of a peptide map. You must have used more than 0.2 ml/min of a flow rate with a 3mm ID column. Do you know why it was difficult? Would it be possible to explain for me? Thank you.

Developed a peptide map on a quaternary Alliance system (0.4 mL/min). Tried second analyst validation on a quaternary HP1100 - total failure, elution order of an early eluting pair of peaks switched resolution between many early and late eluting peaks were different. It looked like a dwell volume problem - then realized the 1100 system the second analyst used had a 900ul syringe. Tried a regular HP1100 and everything was fine - thought the problem was over.

Tried the separation on 4 different quaternary 1100s saw some troubling minor variations but the method worked. It may have been possible to optimize the quaternary systems by adjusting stroke volume, solvent compressibility, etc.

Ran the separation on 4 different binary systems with the dwell volume increased to match the quaternary systems. Minor separation varitions were gone, all chromatograms were superimposable. I attributed the results to better gradient accuracy with the binary systems but didn't really study the problem. Restricted the method to the binary systems.

Added a dwell volume test to the annual OQ/PV procudure and matched the system dwell volumes between the binary and quaternary systems - have seen no other problems in development/validation/transfer/testing. Not currently using any 1.0 or 2.0mm ID columns for routine testing. Remove added dwell volume when testing 1 and 2 mm ID colums.

Good luck.

I belive it to be fairly well known that the 1100 binary system will give more consistant RT reproducibility than the 1100 quart. It should also be pointed out that the Waters Alliance 2690 (a four solvent system) will, in turn, out perform (retention time wise) the 1100. Tom points out that he had trouble transfering methods from an Alliance to an 1100. This is not uncommon, the problem usually lies in the fact that the column heaters perform very differently on the 2 systems. The 1100 does not provide a constant temperature across the entire oven, as well, a temperature setting of 40 C (for example) will provide temperatures ranging for ~35.5-38.5 across various oven points. The Alliance column heater, on the other hand, ranges from 39.9 - 40.1 across varois sample points. To further complicate issues the system volumes, of course, are different. I dont know the volume of the 1100 quart, but a stock 1100 binary has a volume of ~730 uL and the Alliance ~650 uL. Of course the actual volume of the 1100 is variable, depending on back pressure, due to the pulse dampener. As well may people choose to remove the mixer from the 1100 to reduce the volume, this can prove to be a bad idea at lower flow rates using some solvent systems (TFA/ACN as one example).

Enough rambling for now.

Droid

The Rheos 2000 pump avaialble from LEAP Technologies (800-229-8814) has a dwell volume of less than 100ul and has good reproducibility using the low pressure mixing approach for binary to quaternary solvent mixing.

I don't know how the waters column heater works but the 1100 series column heater heats the mobile phase and not the air surrounding the column. This could in theory affect the tranfer of methods from or to 1100 series instruments. Most column ovens that I have come across heat the air rather than the mobile phase. Another temperature consideration for the 1100 series is which side of the heater block you run the solvent through. There is a 3uL side and a 6uL side. I typically run through the 6uL side for better heating of the mobile phase.

Thanks for the advices by Mr. Mizukami and all others about a quaternary pump vs. a binary pump, and also about the heaters. They are very useful and helpful. Thanks again for all your help.