Hello!

There are many ways why the column clogged suddenly!

But one possibility could be their mobile phase and its gradient. If I have understood correctly its gradient?

Eluent A: 20 mM ammonium formate?

Eluent B: 100% acetonitrile?

Gradient of 0% B to 100% B in 5 minutes?

Here the ammonium format could fail at a very high concentration of acetonitrile. You can check this yourself, where you take two beakers. A beaker enter 20 mM ammonium formate and in the other beaker add 100% acetonitrile. Then pour the 20 mM ammonium formate solution in the beaker with 100% acetonitrile. The result is the ammonium formate falls out and the solution is cloudy!

If this is the gradient harm this gradient not only the column, their pumps can suffer damage.

This may be changing in the one the eluent B only as 90% acetonitrile (not 100%) was used.

Best regards

Burghard Cordes

Thank you for providing the chromatogram.

Many Comments in no particular order for you to think about:

• Could you provide us with the flow rate of your method? This is a critical piece of information and I do not see it in your post or on the report you attached. An HPLC method cannot be evaluated without it.
• What is your injection volume (also critical info)? Please provide. It must be very small as your column is only about 80ul in volume. Do you filter it?
• You stated that you dissolve the sample in pure ACN. Please, do not do this. You should dissolve your sample in the mobile phase or a weaker solution, not a stronger one such as pure ACN. Your method injects the sample in pure organic with a mobile phase of pure aqueous buffer. This may cause precipitation, fouling, poor loading of the sample onto the column and is generally not good technique to follow.
• PLEASE TURN OFF the Reference Wavelength feature (e.g. 550/100nm) in your method. It is adding noise and subtracting raw data from your actual output and will invalidate your method). Do not use this "feature" with any methods at all until after you have received proper chromatography training. Reference on ALL wavelengths should be set to 'OFF' as default. *Please read the attached article, "REFERENCE WAVELENGTHS (as used in HPLC UV/VIS)" which will provide you with more information on this topic.
• Why are you starting your gradient method with 100% buffer (vs ~ 70%)?
• Why are you ending the gradient with 100% ACN (vs perhaps 90 or 95%, depending on what is actually needed)? This may cause buffer precipitation which will clog or foul your tiny column (your column has a volume of only 80ul!). Switching to 100% organic is never a good idea after using buffer.
• Looks like initial peaks are starting to come out too early and too fast. Look at the initial peak shape (fronting) vs. gradient (too steep). Need flow rate to recalculate, but even without knowing your flow rate can see that your method needs to be optimized for the samples.
• It appears that you should initially address many of the obvious problems with your method before blaming the column batch as the problem. While large variations in RSD are commonly seen with such tiny columns (2.1 x 30mm), I would not concern myself with that issue until after my method was optimized on one column.
• If you have fouled the original (and probably 2nd column too) with sample or buffer, then both should be OK to try and clean. C18 columns are very rugged and can take a lot of abuse as long as ion pairing agents are not used. Run about 30 column volumes of a flush solution which contains a mix (NO BUFFERS!) of high resistance/purity water and MeOH (e.g. 50/50 to start with, then 10% Water/90% MeOH) followed by the same with Water/ACN, then retest the columns with a proper std (not your samples) to determine their condition. Your flush solution MUST include solvents which will dissolve ALL possible compounds that you have injected on the column. You may also want to use the same initial flush solutions with with 0.1% high purity formic acid added too. Make sure everything is fully dissolved before continuing with the flushing.

http://hplctips.blogspot.com/2011/03/reference-wavelengths-as-used-in-hplc.html

Dear Burghard,

I do apologize for late reply as I have been ill. The first thing I will do tomorrow morning is checking your suggestion.

Dear Bill,

First of all I believe I just started my learning in the HPLC approach, and I guess I need for long time practising to develop my skills with reading, and definitely with the guidance of such a generous experts (like you).

I will repeat your question in the same order to be followed by my comments

·         Could you provide us with the flow rate of your method? This is a critical piece of information and I do not see it in your post or on the report you attached. An HPLC method ACNnot be evaluated without it.

In all my experiments, I have used 1ml/min flow rate.

·         What is your injection volume (also critical info)? Please provide. It must be very small as your column is only about 8ul in volume. Do you filter it?

The injection volume is 2 µL as it is compatible with the size of used column.

·         You stated that you dissolve the sample in pure ACN. Please, do not do this. You should dissolve your sample in the mobile phase or a weaker solution, not a stronger one such as pure ACN. Your method injects the sample in pure organic with a mobile phase of pure aqueous buffer. This may cause precipitation, fouling, poor loading of the sample onto the column and is generally not good technique to follow.

The mean reason for using ACN as solvent is that this method was developed to follow the progress of a reaction where ACN used as solvent.

·         PLEASE TURN OFF the Reference Wavelength feature (e.g. 550/100nm) in your method. It is adding noise and subtracting raw data from your actual output and will invalidate your method). Do not use this "feature" with any methods at all until after you have received proper chromatography training. Reference on ALL wavelengths should be set to 'OFF' as default. *Please read the attached article, "REFERENCE WAVELENGTHS (as used in HPLC UV/VIS)" which will provide you with more information on this topic.

Actually I did not understand what you meant here.

·         Why are you starting your gradient method with 100% buffer (vs ~ 70%)?

Essentially in the first steps of developing my method, I started with 95% but it has not given a proper separation. After several attempts, this 100% worked well in separation these two compounds.

·         Why are you ending the gradient with 100% ACN (vs perhaps 90 or 95%, depending on what is actually needed)? This may cause buffer precipitation which will clog or foul your tiny column (your column has a volume of only 8ul!). Switching to 100% organic is never a good idea after using buffer.

I agree with you about that, I will double check this matter.

·         Looks like initial peaks are starting to come out to early and too fast. Look at the initial peak shape (fronting) vs. gradient (too steep). Need flow rate to recalculate, but even without knowing your flow rate ACN see that your method needs to be optimized for the samples.

I will try to decrease the flow rate in order to elute the first compound a bit later.

·         It appears that you should initially address many of the obvious problems with your method before blaming the column batch as the problem. While large variations in RSD are commonly seen with such tiny columns (2.1 x 30mm), I would not concern myself with that issue until after my method was optimized on one column.

I really appreciate your comment. However, I am blaming the column here only for one reason as I used similar column and I had no any problem. With using the new column, all these problems have contributed.

·         If you have fouled the original (and probably 2nd column too) with sample or buffer, then both should be OK to try and clean. C18 columns are very rugged and ACN take a lot of abuse as long as ion pairing agents are not used. Run about 30 column volumes of a flush solution which contains a mix (NO BUFFERS!) of high resistance/purity water and MeOH (e.g. 50/50 to start with, then 10% Water/90% MeOH) followed by the same with Water/ACN, then retest the columns with a proper std (not your samples) to determine their condition. Your flush solution MUST include solvents which will dissolve ALL possible compounds that you have injected on the column. You may also want to use the same initial flush solutions with with 0.1% high purity formic acid added too. Make sure everything is fully dissolved before continuing with the flushing.

Only the new column has fouled, and definitely I will try your procedure in order to get the column operates.

I have tested the solubility of all my compounds initially with only ACN and they completely dissolved.

Secondly, I tried to dissolve them in the buffer (20 mM ammonium formate), they did not, but they dissolved immediately with adding the same amount of ACN to the same undissolved mixture.

Finally, I highly appreciate your kind comments, and all of them will be considered.

1st. Apologies for an earlier typo. Your column void volume is ~ 80 ul (not 8ul as it initially appeared from the system truncating the character). At 1ml/min your Tzero would be ~ 0.08 min (4.8 seconds)!

Saddam: I feel that you would benefit the most if you could find someone at your University with proper chromatography training who could help you in this matter. You appear to be going about this method development process without any understanding of chromatography fundamentals and such an approach will result in poor quality data. You still seem focused on column cleaning and receipt of defective columns. I think your efforts would be better spent understanding the technique used first. Once you have gained an understanding of the technique and used the skills to develop a better method, then you can evaluate columns. As someone new to the field you can not expect to master this technique, the use of the instrumentation or even the software without first having the benefit of months and years of proper training first. - And then, years of practical experience with samples. Many scientists using this technique still do not have a good understanding of the technique and how the many parameters affect the data, even after many years of use. It is a complex analytical technique that takes time to master. Having an experienced professional on hand can greatly help guide you through this process as can reading many of the wonderful books on HPLC principles too. "We" can only provide so much guidance to you via this forum. You need more help with the basics and fundamentals first for our comments to be of value. I hope you have someone there who can provide training to you.

There are too many individual problem items in your method and description which need optimization to list here.  However, I wish to comment on one basic area that I hope you will think about and adjust when you are re-developing this method for the samples.

Solubility: You wrote that all of your compounds do dissolve in pure ACN, but none of them dissolve in your buffer solution. Yet, you start your gradient method with 100% buffer!  The same liquid that none of your samples dissolve in. You also mentioned that when you started the gradient at 95% buffer, you had very poor results so changed it to 100% buffer. This makes no sense at all.

I will leave you with a link to a free webpage which contains many short articles on a variety of chromatography topics. All are focused on teaching and stress the importance of learning to use good chromatography fundamentals. I hope you find it valuable as you begin to learn this wonderful analytical technique.

Good luck to you.

http://hplctips.blogspot.com/

It is very common for an method that was developed on an old column to work poorly when a new column is used. Although modern columns are robust they are not indestructible, and an old one with show different chromatographic behaviour than a new one.

Bill Letter's advice is sound.

Peter