Waters e2695 Injection Peak Area Repeatability Problem

[Terry]

Dear all,

I have been stuck with standard injection peak area repeatability problem in the past few days, 50% of the runs can not meet the acceptance criteria defined in the test method. The typical data is as follows:

Area for 5 injections of standard: 2750277, 2624696, 2587515, 1925169, 2004414
RSD: 16.1%
Acceptance criteria: NMT 2.0%

Some troubleshooting work has already been performed: the repeatability result is very good (RSD = 0.1%) with another HPLC using the same standard, column performance is good, injector has been purged, compression check okay, adjust seal test okay.

Could anyone direct me on other possible cause?

Regards,

Terry

[mattmullaney]

On the second HPLC--is it also a Waters 2695?--what are the peak area values? Are they similar to the scattered values you receive with the 2695 having the repeatability issue?

What are the loop volumes on both 2695s--are you injecting no more than one-half the loop volume on each HPLC system?

[Terry]

The second HPLC is Agilent 1200, the peak areas are: 46687197, 46727225, 46682198, 46682859, 46624908, RSD = 0.1%.

The loop volume for both HPLC is 100 microliter, the injection volume is 20 microliter

[mattmullaney]

Good injection volumes being used for installed loops, check. Next, is the needle wash solution used for the 2695 stronger in elution strength than the sample diluent and the highest percent organic solvent used in your eluent composition?

Also, have you tried repeated injections of, say, water, using the 2695 and obtained the difference(s) in mass that result? This way we can see if the injector is working well mechanically.

[Terry]

Good injection volumes being used for installed loops, check. Next, is the needle wash solution used for the 2695 stronger in elution strength than the sample diluent and the highest percent organic solvent used in your eluent composition?

Not actually, the needle wash solution is water, the sample diluent is 70% pH 1.9 phosphorate buffer and 30% Acetonitrile (ACN), the mobile phase is 70% aqueous TFA (0.1% in concentration) and 30% ACN. However, the carryover is negligible in the post standard blank.

Also, have you tried repeated injections of, say, water, using the 2695 and obtained the difference(s) in mass that result? This way we can see if the injector is working well mechanically.

Yes, we have weighed the HPLC vial prior to and after injection of 20 μL water via a 0.01 mg balance, the weight difference (in mg) is as follows: 22.64, 20.88, 20.29, 20.75, 20.49, 20.59, 20.49, 20.36, 19.84. In addition we have performed Injector linearity and accuracy test by injecting 5, 20, 40, 80 and 100 μL cafffein standard, the R2 is 1.0000 (after rounding up to 4 decimal place), the calculated accuracy is 0.2 μL, as per Waters initial IQ/OQ package. The acceptance criteria is: R2≥0.9990 and Injection Accuracy≤1.0μL. The result from weighing and caffein injection are confusing for me.

[AA]

Check for an air bubble in the syringe (just open the door and look). This is often the cause of issues like this. If you have air bubble, you need to get rid of it, check the users guide for how to do that.

[DR]

If the plunger tip and/or barrel of the syringe is worn, they will need to be replaced. Worn syringes slide a lot more easily at the start of draw than they do later in the draw (farther down the barrel).

[mattmullaney]

I am agreed with AA and DR both...either of these could cause your repeatability issues with the injection.

As to the good caffeine results but poor results with masses...sounds like the caffeine results were an aberration. For this kind of wackiness, I'd guess DR's hypothesis may be the actual trouble and its an old syringe plunger...air in the syringe barrel can also be brought about by a leaky syringe.

As to the Needle Wash composition...isn't Waters general recommendation that the Needle Wash be as least as strong in elution strength if not stronger than the eluent composition and/or sample diluent? Even if you do not want to make this kind of change, unless you replenish the needle wash (100% water) with some high frequency, won't you eventually get mold growing in the Needle Wash if there is no organic solvent present at all?

[Terry]

We have changed the syringe, it works fine now. We will revise our SOP to add the requirement of using stronger solvent mixture as needle wash solution.

Thanks for your input.

[mattmullaney]

Hey Terry,

My thanks for letting everyone know of the resolution to the problem. I must also take the time to say--I Was Wrong!! An old trick I learned...if you're rushed, take off the syringe, remove the plunger and drop it from about a 10" height (I've dropped the plunger through drinking straws) onto a clean, flat, level surface (such as a mirror), and a beat-up Teflon plunger tip will work a while longer. Replacement, though, is the best way to go.

The Alliance 2695/e2695 injector systems are of the flow-through needle design (FTN), thus there is no limit to one-half the loop volume per injection...in essence the "entire loop volume" may be injected. The sample loop is not isolated from the eluent flow path during a chromatographic run...same thing with Agilent 1100/1200 systems, or the Waters Acquity-H system.

This time, I wish Mr. Jupille had caught my error!

Too many hardware manufacturers...but I should actually know what I say that I know, too.
All else about Needle/Seal Wash, though, is on-the-money.

[Terry]

Dear Mattmullaney,

Thanks very much for the update. Just curious, for what kind of HPLC should less than half loop volume be used?

Terry

[mattmullaney]

Hi Terry,

There isn't a "kind" of HPLC, per se. If the instrument has a "flow through needle," where the needle and the sample loop are in the eluent flow path during the chromatographic run, then the only limit to the injection volume is the loop volume itself.

If the instrument has a "fixed loop," then the loop is segregated from the eluent flow path during the chromatographic run. The sample "slug" is pushed, or pulled, into the eluent flow path...eluent is not sweeping the volume of the sample loop during the run.

In general, high-pressure mixing systems employ "fixed loop" design, though there are exceptions...and low-pressure mixing designs generally employ "flow through needle) design, though again, there are exceptions.

In the Waters world...the Acquity-H is a flow through needle, and both flow through needle AND fixed loop are employed in the Acquity-I...you can see both at Waters Web Site (fixed loop first, then flow through needle). Both Acquity-I instruments are binary high-pressure mixing, though one has a fixed loop and the second a flow through needle.

http://www.waters.com/waters/library.ht ... =134619338

http://www.waters.com/waters/library.ht ... =134619288

The difference...there will be less gradient delay volume in the fixed loop design...fixed loop has a bit of an advantage on that count. Flow through needle "ought" to have an advantage in sample carryover, as the loop is multiply flushed.

Does this help a bit? You can find an explanation of the fixed sample loop injector at the IDEX web site...Rheodyne has been making this design for many years in their manual injection valves...as well as their automated ones commonly employed in Dionex ICs and Thermo Separation Products LCs. I'll send a link to this thread later on today if I can--see below after an edit...

http://www.idex-hs.com/support/rheodyne ... note_5.pdf

[Terry]

Thanks for information, it is very helpful.