Will solvent mismatch cause a (negative) peak?

[Anthony_Ng]

Hello all,

Today I made a blank run and noticed a negative peak somewhere. Here is the condition:

Eluent: 90:10 MeOH:H20 (isocratic)
Flow: 0.4ml/min
wavelength: 220nm, reference 360nm

When I inject 5ul pure MeOH, a small negative peak appeared at 1.9min, (t0 is around 1 min).

Hence, I changed the condition to 80:20, the peak came out later. And then 70:30 and even later.

But when I run it with 100:0 MeOH:H2O, the negative disappeared.

Just want to know.....if the injection solvent is stronger than eluent, will it cause a "peak"?

Thanks in advance.

[tom jupille]

Yes, it can happen. All that is required is that your mobile phase contain a low level of some contaminant (presumably from the water) which absorbs at 220 nm. When you inject 100% MeOH (which does not contain that contaminant) you get a "vacancy peak" at the retention time of the contaminant.

The other possibility is that your methanol is contaminated with something that has a higher absorbance at 360 nm than it does at 220nm. In that case, however, I would still expect to see the negative peak at t0 when you elute with 100% MeOH.

A quick way to check is to disable the reference wavelength function and see if your mystery negative peak disappears.

[HW Mueller]

Note that the explanations given by Tom have nothing to do with solvent strength.

[Anthony_Ng]

Here I change the injection solvent composition a little and more info obtained:

Injection solvent: 80:20 MeOH:H20
Flow: 0.4ml/min


Here I used 70:30 MeOH:H2O and a negative peak at 5.5min


When 75:25 MeOH:H2O the negative peak comes earlier (about 4.2min)


When the mobile phase is match with injection solvent (80:20 MeOH:H2O), there is no -ve peak (or +peak) at all.


If further increase to 85:15 MeOH:H2O, no -ve peak seen but a peak appeared at 2.4min


The peak comes earlier if 90:10 MeOH:H2O is used.


Do you think there is a relationship between the -ve peak and +ve peak? And how to explain the peak disppeared at 80:20?

By the way, why the pattern of t0 change from "up then down" in weaker solvent strength (70:30 & 75:25) to "down then up" in stronger strength (85:15 & 90:10)?

[HW Mueller]

You confused things here by stating that the injection solvent is 80:20, assuming that this is supposed to be the mobile phase than the most obvious explanation would be that your MeOH has crud in it. The retention time moves, because you inject too much.
Don´t worry about details of the peaks at tm (to).

[Anthony_Ng]

Thanks very much.

What I have prepared is: I premix 80%MeOH and 20%H2O, put it in a sample vial. And inject 10ul of it by autosampler. (And this 80:20 mixture should be without analyte/compound). Hence run the chromatogram with the above 5 different conditions. (all isocratic)

By the way, is the air peak (around 1.6min) seemed too big? How can I minimize it?

[HW Mueller]

It´s the other way around from what I understood? The mobile phase composition changed, not the injection composition?
If that´s correct now, it is simpler to analyze: The MeOH is dirty, the peak moves because the elution strength increases.
How do you know that there is an air peak? I don´t know how your autosampler works, but overfilling the loop and/or degassing the sample can help.

[HW Mueller]

Whew once confusion sets in . . . . .
Just noticed that the negative peak is optained when the mobile phase water is proportionally more than the H2O injected (always 20%?), and it is positive if the mobile phase has relative less H2O than the injection. That can only mean that this dirt peak comes from the water, not the MeOH.

[Anthony_Ng]

Yes it is the mobile phase composition changed, and keeping the injection composition no change (i.e. 80:20 in a sample vial)

At the very beginning of this troubleshooting, I thought there was dirt in either MeOH or H2O too. But....if so, why the "dirt peak" disappear when the mobile phase match with injection composition (i.e. both mobile phase and injection composition are 80:20)?

I tried mixing 90:10 as injection composition in a sample vial. Now the peak disappeared when the mobile phase is 90:10. And negative peak is there when H2O content larger than 10% in this case. (e.g. 80:20, 70:30 etc.)

For the air peak at 1.6min, I prepare a vial without any solvent inside, just cap an empty vial. Hence make the injection volume of 5ul, 10ul and 20ul respectively. And the area increase accordingly.

[Anthony_Ng]

I just surprisingly notice that, apart from the major -ve peaks changed to +ve peaks, there are 2 small -ve peaks also flip to +ve when mobile phase composition changed!

[HW Mueller]

You should be able to figure that out now. If you inject material which is the same as the mobile phase you neither add or subtract dirt that the mobile phase continuously puts on the column.