Chromatography Forum

Hello All perhaps someone can help me w/ an issue I am having w/ my method. I am separating oligonucleotides and degradation products using a gradient. My two mobile phases are 25mM TEAA pH 7.4 and ACN I run the gradient from 5% B to 85% over ~50 minutes. The gradient is not linear it proceeds in several steps (see below). The issue i am having is at the high ∆B i am seeing "junk" eluting off of my column resulting in baseline rise as high as .1 ODs. My inclination is that we have a problem w/ the ACN (we went to a different supplier and different containier) I have not had a chance to run a different lot of ACN because my first shot was to use a different water source (we recently moved to a new lab). That unfortunately did not pan out. So, my question is: is there something inherent in my gradient that could be causing the baseline instability? This is a newly developed method and though the old method also showed issues I would like to ensure that I am not casuing stabilty issues w/ too steep a gradient.
Thanks
Nat
New Method (Step Gradient)

Time % B
0 5
1 5
4 12.5 ∆B= 4.166666667 %B/min
6 12.5
9 17 ∆B= 1.5 %B/min
12 17
18 32 ∆B= 2.5 %B/min
20 32
24 42 ∆B= 2.5 %B/min
26 42
40 45 ∆B= 0.214285714 %B/min
42 45
50 85 ∆B= 5 %B/min
54 85
60 5 ∆B= -13.33333333 %B/min
62 5

I think that this is inherent with this type of gradient.

I have had to resort to a similar gradient with a large spread between the starting and ending organic concentration for amine containing compounds.

What is your sample diluent?

When I would use mobile phase A as the diluent, I observed alot of peaks at the end of the gradient, and visa versa when mobile phase B was used as the diluent. I observed many peaks on both ends of the gradient when I mixed 50:50 MP A:MP B for diluent.

My gradient was linear, with no steps, but I observed a large increase in the baseline. I believe this is normal, because of the large spread between the mobile phase concentrations.

Also, be careful to fully equilibrate your system after each injection. It may take up to 10 minutes of equilibration to get the baseline back to near the original AU or MV reading.

sorry for the thread hijack, I was going to suggest as a first option to try pre-mixing the mobile phases but then wondered how one should correctly calculate the new B percentages (see below)..
let A= 95:5 buffer/ACN
let B= 15:85 buffer/ACN
If I now treated the old 5 and 85% ACN as 0 and 100%, should this scale linearly for B inbetween (giving now, for example 14.7 %B at 12.5 mins) or is it more complex.. help please, Tom!

Does this junk come off at high ∆B, or at high %B? Other thing was just to ask about how pure the TEAA is, you don't see junk in blanks? If not, I guess it's the sample.

Hi all,
Thanks for you responses. I have "solved" the problem... I say solved because i am still getting a bunch of junk eluting off of the column at high %B and high DeltaB but the method is passable. There is a huge flat topped peak at high %B (85%) and a smaller peak coming out around 20minutes at high DeltaB.
This was my fix: The first thing I did was to premix the mobile phase and adjust the gradient (i only changed B though not A&B to make the conversion easier). I also added TEAA to mobile phase B to give a stable TEAA concentration throughout the gradient (really should have been the 1st place I looked). This allowed me to see my peaks of interest down to the LOQ of the previously developed method with good linearity for the product and degradation products (.999 product .99 degradation products). To answer some questions about my sample; the diluent is generally PBS or water for standards and vitreous humor or plasma for samples. I suspect that the junk is not from the sample becasue the salts in PBS should elute very early (solvent front) and the junk was present in blank PBS injections and 0uL injections. Thanks to all for the responses it is probably a contaminant in the TEAA or the TEAA itself causing the ghost peaks I have ordered another lot of TEAA hopefully I will see a cleaner chromatogram but it is more likely that this is inherent in the gradient.

Tom here is how I did my conversion. I struggled for a little bit trying to work with both A&B adjusted then realized that adjusting B alone was much easier (as I said in the previous post). I rounded to the nearest .1% and was relatively close to my target A:B ratios <= 0.04%

Time %A %B %TEAA %ACN
0 94.11764706 5.882352941 95 5
1 94.11764706 5.882352941 95 5
4 85.29411765 14.70588235 87.5 12.5
6 85.29411765 14.70588235 87.5 12.5
9 80 20 83 17
12 80 20 83 17
18 62.35294118 37.64705882 68 32
20 62.35294118 37.64705882 68 32
24 50.58823529 49.41176471 58 42
26 50.58823529 49.41176471 58 42
40 47.05882353 52.94117647 55 45
42 47.05882353 52.94117647 55 45
50 0 100 15 85
54 0 100 15 85
60 94.11764706 5.882352941 95 5
62 94.11764706 5.882352941 95 5