The 1200 and 1260 are both quaternary pumps? You have tested the same samples with the same column with the same method on both LCs? Both LCs have approximately the same length of tubing with the same inner diameter? The backpressures are comparable? The retention times are comparable on both LCs, only the resolution is worse on the 1200?

If the answer to those questions is "yes", the probability is high, that the different size of the sample loop is the cause of your problems.

If the sample loop size is the cause, I would also expect that late eluting peaks are less affected that early eluting ones.

If you are using a sample loop volume that is 90 times greater than the volume of your sample, the 'bolus' of sample will inevitably be diluted, the degree to which it is diluted will also depend on the internal diameter of the sample loop (the wider it is, the greater the mixing with the mobile phase that makes up the rest of the loop volume). I would switch to a 10-20ul loop, if possible, so that the sample completely fills the loop, or can be 'sandwiched' between a few ul of mobile phase each end, to ensure that the full 10ul of sample is injected, with the minimum of band spreading as it hits the head of the column.

If you do not have access to a smaller sample loop, you can minimise the dilution/spreading of your injected sample by filling the injection syringe to 885-890ul with the starting mobile phase, and then adding your 10ul sample via the injection needle (as it is narrow) followed by 2-5ul more mobile phase. This would have the effect of injecting the 10ul sample in a volume of 12-15ul onto the column first, followed by the balance of the (900ul of mobile phase) injection volume, which should not broaden the injected band any further since it would already be applied to the column.

I used both strategies, with success, when HPLC was my main occupation for several years. I hope this helps.

Thanks to everyone for your answers. Then I think that I have been told something different from Agilent. I got the idea that the semiprep metering head takes my 10 ul of sample, but the sample is retained in a small portion of the long loop. Then the sample is sent to the column. I think that my 10 ul don´t go to the piston to mix with 900 ul of mobile phase and dilute. Am I right?

Dear Alfonso,

I have taken a look at the type of injection loop that you are using for loading sample for preparative applications. It definitely would not be set-up in a way that would send the sample through the pump head before arriving on the column. They actually mention that it is possible to partially fill the loop with a smaller sample volume (with the sample closest to the entrance of the injection port - eg sample application syringe needle) so that the sample dilution is minimised. It shows the id of the sample loop as being 0.76mm, so you would get more sample mixing in this than in a 20ul injection loop but, done carefully, you could get good preliminary results with the 1ml sample loop, if a 20ul loop is not currently an option. It is important, with small sample volumes, to ensure that the tubing between the injector outlet and column head is as short and narrow as reasonably possible, to prevent further pre-column band-spreading. You can calculate the volume of the tubing using:-

Volume(V) = length of tubing(l) x π  x (internal radius(r) of tubing (diameter/2)2

This Agilent attachment might help - it is for the model 1100, but the sample injection information should be similar, if you have a manual injection port (see pages 5-10). Oh, and I have just found some information about the Agilent model 1200 system (it shows a similar manual Rheodyne injector):- 

Thanks a lot Zoë,

I agree with you, I also changed already the tube from the injector to the column, which was really long (I don´t know why). I´ll check the performance with an eye in the 1 ml loop, maybe I can change it as well.

Best regards,

Alfonso

Dear Alfonso,

Injection loop volume will increase your system dvell volume (volume of tubing from point of mixing two mobile phases to column head), which can change the gradient/peak separation if you compare your two systems.

Agilent 1260 has capillaries of smaller diameter than 1200 which lowers peak band spreading which results in  tighter, higher peaks. Also, dvell volume of 1260 will be lover than on 1200.

Maybe hese things influence worse separation on your analysis on 1200.

The 1100/1200-series A/S is a wonderful unit with a full flow-through design sample path (unique) and the ability to easily tailor it to different injection volumes using kits.

If you have the optional 900 ul metering unit installed, then you need to replace it with the std analytical metering unit (100 ul volume). It should have come with it when new (std configuration). That will solve the problem. The 900 ul metering system was not designed for small volume injections (IOW: less than 100 ul) and will substantially dilute a tiny 10 ul volume into a rainbarrel sized 900 ul volume! The "Loop" selection contributes to this extra volume, but replacing the 900 ul loop in your system with a 100 ul loop alone, without changing the metering unit, will not solve the problem. The specific metering pump (think of it as a high pressure syringe) and loop are tailored for each application. The std system is set up with a 100 ul metering unit and 100 ul loop. The 900 ul system is set up with the 900 ul head and 900 ul loop. Each one is designed to be used together to change the system from a std 100 ul analytical injector to a semi-prep 900 ul injector. NO other changes are needed and you can purchase the two parts (plus the appropriate needle seat capillary) from Agilent or a distributor. *I have made these changes dozens of times on client systems (very easy to do). Once you change the metering head, loop, needle and needle seat back to the std 100 ul configuration, you should obtain the same results as found with your std 1260-series A/S.

BTW: The design of the 1100, 1200 & 1260-series std A/S are the same. They label the 1260 version as 600 bar rated, but they really are the same. They use the same parts, same volumes and operate exactly the same way (I service them). All of these A/S's are true 'flow through" autoinjectors where the mobile phase is flushed through the complete flow path of the needle, metering head (piston has space around it for liquid), needle seat and loop during the entire run. The flow-path is washed clean with mobile phase the entire time with one exception, loading a sample into the loop. This design results in almost never having to use a wash vial at all (except for really sticky samples) and is common to all A/S modules in the Agilent/HP 1100, 1200, 1260 and 1290-Series. Other manufactures do not have this type of flow through design (they just flush the needle) and require wash vials to keep the flow path clean. *I design and use analytical instruments so really appreciate this feature as it makes their A/I's some of the best on the market due to the use of the high pressure, flow-through metering pump as a "syringe".

Alfonso: If you have any other questions regarding how these units operate, need part numbers or advice in how to optimize them, please let me know. I have over 25 years experience designing, working on and teaching advanced courses in their use and operation (for these and other brands).