ZIC HILIC Chromatography Issue

[mjordan]

I have been running an HPLC/MS/MS method for methylmalonic acid (MMA) using a published procedure for several months now. The method uses a ZIC HILIC column (2.1 x 100 mm), and a buffered mobile phase. The mobile phases we are using are (A) 100mM NH4Ac, pH = 4.5 (adjusted with formic acid) and (B) Acetonitrile. The method starts with an isocratic hold of about 3 minutes using 20% (A) and 80% (B). The methylmalonic acid normally elutes around 2.15 minutes, very consistantly. Then the method switches to a higher aqueous composition to wash the column and then equilibrates back to 20/80. This method has been running very consistantly for several months. The HPLC has recently been PM'd. Since the PM, the retention time of the methylmalonic acid has shifted to about 2.7-2.8 minutes. In addition, the peak is fronting quite significantly.

I have tried several things. I measured the flow rate from each line and verified that the pump and proportioning valve are delivering an accurate and consistant flow. I have tried premixing the mobile phases and running the method isocratically through one of the solvent lines. I have tried running the gradient though the other two solvent lines I typically don't use (lines C and D instead of A and B). I have measured the pH of the mobile phase to confirm that it is still 4.5. In fact, I have also tried running the method on two other preparations of the buffer. I have replaced the syringe purge and needle wash solvents. Nothing has helped. I even tried replacing the column and that hasn't helped either.

To add to the level of confusion, this instrument is used to run two other compound assays and the chromatography of those assays have not been impacted.

As I mentioned earlier, I am seeing fronting peaks. Well, because this is an LC/MS/MS method, the MS detector is in MRM mode. The internal standard is a deuterated form of methlymalonic acid (D3-MMA). We are looking at 3 transitions: 116.8>72.8 (Daugter of MMA), 116.8>116.8 (Parent of MMA), and 119.8>75.8 (Daughter of D3-MMA). The peak fronting seems to only be affecting the peaks from the two MMA transitions. The D3-MMA transition appears to have a symmetrical peak shape.

At this point, I can not tell if this is an application issue or an instrument issue, since the method has been working very consistantly up till the PM. Does anyone have experience with this column or have any ideas of what might be causing these problems??

[mjordan]

Come to find out, the reason for the issues that I was having (mentioned above) was due to the column sorbent. When I was initially seeing a shift in the retention time of my peak and I changed to a new column, the sorbent in the new column was made from the same lot as the sorbant in the original column. The "new" column never worked properly.

I happened to run across another ZIC HILIC column as I was cleaning my lab area and noticed that the sorbent lot # on this column was different from the other two columns. So, I tried this third column and it worked like a charm right on the first injection!

My advice to anyone using HILIC columns, particularly ZIC HILIC, pay careful attention to sorbent lot #'s. Just because you recieve a new column does not mean the stationary phase inside the column is also new.

Under typical circumstances I wouldn't see this as a problem. But if the manufacturer has a bad batch of sorbent then it will likely decompose in the new column just as it did in the old column. Especially since these columns are stored and shipped in similar mobile phases as what you may be using in your method (i.e. the chemical environment of these columns is the same, and over the same period of time).

Hopefully this information can help someone else!

[SIELC_Tech]

Lot to lot reproducibility for good columns suppose to be below 2%, in your case it is 30%. Also there is another issue, when you use older column it might be "modified" chemically (hydrolysis of stationary phase, modification of silica surface , etc.). My guess each time you order new column you are going to have this problem. T

[mjordan]

Well, like I said, I am using a published procedure and there are other users of this method over in Europe who all get 2.1 minutes for their MMA retention time so I really shouldn't expect to see this kind of variability from lot to lot. But what you are saying may be possible and I will certainly be keeping an eye on this.

[ym3142]

Mark

Thanks for sharing