Hi. Have you tried re-injecting a previous sample from before you started seeing split peaks? If you have and you are still getting split peaks then it is likely that you need a new column. The only other thing that I could think of is the mobile phase. Have you changed the mobile phase, made a new batch, or changed vendors or lots of any of your mobile phase components?

Thanks for your answer Kendall Arrington Miller . I have tried re-injecting previus samples and prepred new MP, but the splitting occured again. After proper cleaning my belief was that I do need a new column, but wanted to get another opinion if there could be other possible causes.

That is all I can think of. It won't help with your current column, but was there anything different about the first sample that you saw splitting peaks with? Was the extraction different, or was it a different color or consistency? If so, that may have been the reason why you saw splitting all of a sudden. It could have done something to your column....just a thought. Good luck in all you're doing :)

Get a NEW column! If this separation worked it should work now. Evaluate your method since the excipients may be retained and coat the column thus yeilding split peaks.

In my experience, the peak splitting may caused by the void in the column. Do you have a guard column to protect your analytical column? Before changing the column, you may try to reverse the column to see if it will fix (temporarily),

Valuable information already given, frit blockage is very unlikely cause. You may check this by purging your mobile phase channels, if you experience high pressures then you need to change the frits. Physical examination also works.

Best wishes.

Do you have the same problema when you inject standards? What is the device you use to inject? You can have an injection problem rather than a column problem. Did you try to inject a standard into another column? If you keep having split peaks with standards and even with another column, you could think of injection problems.

Thanks everyone for your suggestions.

Ademario Junior after the cleaning I was only injecting standards to check the peak shapes. The first injection was better after wash, but the peak began tailing and eventually splitting after several injections of the same std.

Martin Ongas as this would be my first time to physically examine a possible frit blockage, do I need special tools to open up the column endfitting?

Dear Maltina Shkodra , in my previous response i was referring to the pump frit.

As for the colum frit, this could be helpful. http://delloyd.50megs.com/moreinfo/hplcfrit.html

It's however a delicate exercise that could damage the column packaging if not done carefully.

Best wishes.

Hi Maltina,

Please look into the chromatogram and answer the question:

• All peaks affected: This means there is a void volume before the column (between injector and column to be exact; Was the injection loop replace?)

Solution: First please check the column fittings there are two different styles for the connection (Waters style and Parker style) available. Sometimes there had been used the instrument with different column and the connection was destroyed. (Please check the PDF: https://www.waters.com/webassets/cms/support/docs/720003234en.pdf).

If you are sure that there is no void volume then turn the column and make a reverse injection, if the peak splitting still appears, so replace the column.

• Only the early eluting peak or peaks are affected – late eluting peaks are not affected

This means you may have used a too strong sample solvent (perhaps pure ACN or MeOH)

Solution: The best is you replaced it by the starting mobile phase, but often it is not able to dissolve the entire sample, then try to dilute it with water as much as you can. If you cannot change it, inject less sample volume.

A very common mistake is that the mobile phase contains acid or buffer, but it has not been added to the sample solvent.

I hope it helps.

Regards,

Falk-Thilo

It does not sound like you have addressed the basic chromatography fundamentals which cause this to happen. Unfortunately, no chromatography conditions are provided so it is near impossible to troubleshoot the exact cause for the issue. Please provide detailed method conditions for more info or read the article linked below to solve it yourself (or with the aid of someone at your school).

BTW: Peak splitting is often due to the use of the wrong injection solvent, insolubility or sample overload.

You can read more about how to troubleshoot and solve the problem in this article taken from one of my professional courses;

• "HPLC Peak Splitting. Common Reasons For It"; https://hplctips.blogspot.com/2016/09/hplc-peak-splitting-common-reasons-for.html

Falk-Thilo Ferse thanks for your suggestion.

I injected the same sample in direct and reversed injection (hydrocortisone acetate std in 50:50 MeOH:Water, which was the MP as well).

I also injected Nipagin 0.1mg/ml in MeOH with the same MP.

@William Letter thank you for the article.

All injections were 20 microliters. I have tried lowering the injection vol, but the peak splitting appeared again. Also I've been using the same method before many times and the peak shape was fine. Splitting began appearing recently and it's showing up in every sample I inject with the same column.

Here's the chromatograms attached.

No chromatography conditions are provided in your updated post. Without them, we can not understand what you are doing. Second request, please provide them so we can offer advice.

Your 'Y' scale appears to be HUGE!! What detector are you using? We see "uV" written, but is that an abbreviation for micro-volts (most commonly written as uV) or is it a typo for "UV", as in Ultraviolet (UV Signal in mAU or AU units)? It sure looks like signal overloading, but without your actual chromatography conditions, there is no way to know.

What is "reversed injection"? What is "direct" injection? No such scientific terms in HPLC. Please explain what you mean in scientific terms.

Looking at the chromatogram provided, I do NOT see any "Peak Splitting" for the peak with Rt of ~ 19 minutes.

The peak at ~ 4 minutes shows characteristics of a split peak due to detector or sample overload.

Chromatographic conditions:

Column: Zorbax XDB C18 4.6 x 150mm

MP: Methanol : Water 50:50, isocratic

UV detection @ 254.

Sample1: (Rt ~19) Hydrocortisone Acetate diluted in MP. Inj. Volume 20uL.

Sample2: (Rt~4) Nipagin 0.1mg/ml in MeOH.

Yes units are microVolts (uV).

Apoligies for my poor scientific vocabulary, but can’t seem to find the correct term to describe this with one term, so I’ll explain as follows:

Direct inj: 1st injection of Hydrocortisone sample was done having the column fitted in the appropriate direction (as per correct flow arrows on the column pointing towards the detector).

Reverse inj: in the 2nd inj.of Hydrocortisone, I fitted the column in the opposite direction (now the flow arrows on the column pointing towards the injector).

What does this indicate, that with the 2nd inj of the same sample (Rt~19) there was no peak split?

With Nipagin Rt ~4 yes this is a high concentration sample, but as explained earlier, I‘ve tried other samples with low conentration and differenr injc volumes (40, 20, 10uL) and still got peak splits in each injection.

Do these problems indicate that the column material has deteriorated? Or could there be another (or additional) problem?

Thanks

Some comments in no particular order.

• At ~ 19 minutes, you do NOT have true "peak splitting". That is a chromatographic term which does not really apply in your example.
• Thank you for providing a few more HPLC method conditions, but you did not include the flow rate in your post. What is the flow rate (ml/min)? Flow rate is critical to know as it would allow us to estimate the Tzero time of your method using the 4.6 x 150 mm C18 column. *For example, if the flow rate was 0.5 ml/min, then we would know that the first peak at ~4 minutes was unretained (no chromatography taking place, invalid method for Nipagin). If the flow rate was 1.0 ml/min, then we have some sample retention and can move on.
• Regarding injection solvent choice. For HPLC analysis, we dissolve our samples into the mobile phase OR in a solution which is weaker than the mobile phase, not stronger (pure MeOH is Stronger than, not weaker). Injecting samples onto the column which are dissolved in stronger solution may in fact cause many peak shape (tailing and true splitting are examples) and method problems (sample precipitating out). Read up on the fundamentals of chromatography to learn more.
• Why did you turn the column around and run it backwards? The flow direction arrows are there for a reason and indicate the direction the column was packed. You should always try and use a column in the direction shown. Your column may have already been contaminated or damaged when you started.
• The only time you observed a second peak appear next to your primary peak at ~ 19 minutes was when you operated the column in the wrong direction. Is this correct? If so, then the reason you see the second peak appear (not peak splitting) appear is because most likely there is a void in your column (throw it out and get a fresh new column). Sample degradation could also be the cause too, but if this only appears when you run the column backwards, it is most likely a void. A dirty column inlet frit would be more likely to show poor peak shape, not a separate peak 2.5 minutes away from the main peak as yours does. A void can result in a change in time (path and/or interactions) that it takes the sample to go through the column, eluting more than once. The column has been damaged. Columns are consumable items just like vials and filters. They wear out and/or get damaged.
• Thank you for the clarification on the detector type and scale. Now we know it is a UV/VIS detector operating at 254 nm (bandwidth unreported). Since the scale shows micro-volts we can see that the tallest peak is around 500 mVs, but we do not know what your detector's scale is for mVs to mAU so it may or may not be off-scale (500 mV may be 500 mAU (1:1) or something else). The shape of the peak at ~ 4 minutes does appear as an overloaded or off-scale peak would appear, but we would need the detector's AU scale equiv ratio to know which one it is. This is needed for clarification and understanding only. **This does not have any impact on your peak at ~ 19 minutes.
• If you continue to run isocratic methods, please incorporate column wash/flush runs after each analysis which incorporate stronger solvents than you use in your analysis method. This will flush off any accumulated materials. Be sure and use solvents and/or solvent mixtures which will FULLY dissolve all sample types used (ALL). You might want to start with something simple, such as 90% MeOH and 10% Water for your C18 column.

Injection of Hydrocortisone 40, 20, 10uL.

That looks more like a different or changing sample, then "splitting". You can see four peaks present. Re-run it on a scanning DAD using a properly developed HPLC method to see what is really there for comparison. Again, we would need to know the EXACT HPLC method used to generate the chromatogram. Your above example has no method details and elutes ~ 2 minutes so is different (*if it is the same method as before, then peaks which elute at 2 minutes are at or near your Tzero, so invalid).

Thank you very very much for your clarifications @William Letter.

Flow rate was 1ml/min.

The peak that is with good shape was done in the wrong column direction.

Meanwhile in the chromatogram with the close eluted peak to the principal peak (rt 19), the injection was done in the right column direction.

Replace the column with a new column. Wash the column as explained after each analysis. Dissolve samples fully in the mobile phase. Filter samples. Maintain sample concentrations within scale. Insure proper K primes for all samples.

* Good luck, I am leaving for the weekend.

Maltina, I think William Letter and Falk-thilo Ferse already give you many explanations and solution, but I risk in tell you that the frit have a obstruction. The cleaning of this part is trick, requires a some experience. So, if you dont have this, I don't advise you try to cleaning. Then, if you haven't condictions to do it, change the column.

Hope helped

You could use a very dilute solution of sodium azide( say 0.1% solution in water) to flush the column, followed by wash with methanol and water. This will help reduce such issues.

Mohandoss: Sodium azide is for the prevention of bacterial growth, not curing column fouling or avoiding void formation. Washing with MeOH / Water is often too weak a solution to use for column cleaning, esp for a method that already uses MeOH/Water as the mobile phase.

William: You are right that the sodium azide is for removal of bacterial growth.

This in fact gives secondary benefit of resolving the peak splitting issues also.

One can try this and see.

The chromatograms you showed us is the most interesting peak splitting issue I have ever seen. It is more like four compounds. To prove that all four come from the same analyte, please go to the top of each peak and use your DAD to scan the absorption spectra and overlay to see it has the same profile.

Adding to Narong's answer you use the DAD's peak purity module to ensure you do not co-elution of 2 or more peaks. See the attached manual.

This was all previously discussed. Please review the earlier comments.

Change the method development parameters or either column.

Hi William,

Its been recommended by the manufacturer to flush the column with mobile phase containing 0.5% sodium azide (2-3 CV) prior to storage and flush with 5CV of mobile phase prior to use. Is this a right way to store the column or clean prior to use?

Regards,

Harleen

Harleen, using sodium azide in the flushing solution is specifically to inhibit bacterial growth so would be for storage, but NOT for routine flushing and cleaning the column after each analysis.

We have provided a lot of information to help you, but you really need basic training in HPLC to move forward. This is a complex technique and you should really read and review the classic texts on the topic plus obtain some local help (not the web). Perhaps there is someone at your school who is knowledgeable in the the basic of liquid chromatography? Please ask them to help you.

Regarding all of your questions so far: " This was all previously discussed. Please review the earlier comments. "