Jakob wrote: "It was detected at 240 nm. But at 220 nm it looks even worse."

The "drift" problem you described has NOTHING TO DO WITH THE HPLC Column or LAMP used. It has everything to do with first having an understanding of basic HPLC fundamentals. Please work under the guidance of an experienced chromatographer. To have just a BASIC level of proficiency at HPLC analysis requires several years of professional training and industrial experience with a wide range of samples and instruments (something few students have time for). You are running an HPLC gradient analysis and increasing the TFA amount over time. TFA absorbs strongly, esp in the UV region, so the absorbance should and does increase over time (*you provided incomplete method information, but did say; "5 to 100 % ACN (+0.1 % TFA) gradient (in water) ", which is enough to know the cause). No column problem. No Lamp problem. No HPLC problem at all. The observation of the rising baseline with increasing TFA concentration over time is 100% normal and will be more exaggerated as you: view lower signal levels (Yes, it will look more extreme at 220 nm and 210 nm and 200 nm...) and/or increasing TFA concentrations (such as in a gradient analysis as the TFA increases in concentration over time).

• NOTE: Old or "dirty" TFA will greatly exaggerate this baseline effect. Always use FRESH TFA. Be sure and prepare freshly made mobile phase solutions (made up fresh every day. Do not store solutions. Make only what you need). Buy TFA in tiny bottles, not large ones. Yes, they cost more, but you will have fresh solution and save many days, weeks, months trying to troubleshoot problems caused by old solutions. *IF you are using OLD TFA solution, then this problem will appear to get worse over time. Are you using FRESH, new solutions?

To reduce the drift" seen from use of TFA, there are many solutions. Examples: (1) Run an identical analysis after the first one with only mobile phase injected, no sample. Then use the signal subtraction feature in your CDS to subtract out the "blank" sample from your original. (2) Change the concentration of sample used. Increase the sample concentration to reduce the baseline drift. (3) Decrease the acid concentration (only if the method used is reliable with the lower amount) to perhaps 0.05 %. (4) Change the acid used to a more UV transparent acid (again, make sure this is acceptable and results in a reliable method FIRST). (5) Collect a secondary signal in real-time for subtraction. At the same time the analysis is running, collect a second signal to later use for baseline subtraction. Note: The wavelength used must show NO SAMPLE peaks of any kind (outside of the detected range used), but at the same time be near the selected wavelength so it will still show some of the increasing baseline effect to be useful during subtraction (i.e. you use 240 nm. Maybe 240, 8 as you signal 'A' and 310, 10 as signal 'B'. Please note that this is an advanced feature and should only be used by a more advanced chromatographer, as their is a danger of collecting invalid data if not setup and monitored correctly).

the column might be a secondary. do you have 0.1% TFA both in buffer A and B? what wavelength are you monitoring? how are you conditioning your column every run? what's your source of acetonitrile? are you aware that different manufacturers have different qualities of ACN (see image)?

https://americanlaboratory.com/913-Technical-Articles/759-Quality-Evaluation-of-HPLC-Grade-Acetonitrile/#:~:text=When%20using%20air%20as%20a,from%20400%20to%20250%20nm.&text=Figure%205%20shows%20that%20the,significant%E2%80%94up%20to%200.03%20AU.

The question is incomplete- what wavelengths are you using for detection? Lower wavelengths tend to show baseline drift in this solvent system. Another possibility are contaminants in the TFA or acetonitrile which slowly elute from the column. Try using fresh solvents and new TFA. If your analysis allows it, try a longer wavelength.

See: https://www.sciencedirect.com/science/article/abs/pii/S0021967301869065

https://www.researchgate.net/post/How_can_i_remove_Baseline_drift_from_my_HPLC2

Jack Silver Sorry I forgot to mention the wavelength, it's only shown really small in the top left corner of the chromatogram. It was detected at 240 nm. But at 220 nm it looks even worse. Contamination is not the problem, I used fresh reagents/bottles for this run.

Our analyses in the past didn't show such a strong drift. I came to realize, that we have this problem a few weeks ago.

Anyways, thank you for your suggestions.

Savita Rani

Does the chromatogram trace show the full lenght of the gradient run (including both the washing step - at 100% ACN - and the following reconditioning step at 5% ACN)? Is it representative of a blank (or wash) sample?

Diego Pinetti yes it shows the washing step at 100% ACN and the equilibration. The gradient from 5 to 100% ACN is running over 18 min. It is a blank sample (5:95, ACN:water, v:v).

I see that at 24 min the baseline is still decreasing, is there a further equilibration step (not shown) or do you usually inject the following sample at this timepoint?

How is the system pressure behaving, can you show it?

Jakob wrote: "It was detected at 240 nm. But at 220 nm it looks even worse."

The "drift" problem you described has NOTHING TO DO WITH THE HPLC Column or LAMP used. It has everything to do with first having an understanding of basic HPLC fundamentals. Please work under the guidance of an experienced chromatographer. To have just a BASIC level of proficiency at HPLC analysis requires several years of professional training and industrial experience with a wide range of samples and instruments (something few students have time for). You are running an HPLC gradient analysis and increasing the TFA amount over time. TFA absorbs strongly, esp in the UV region, so the absorbance should and does increase over time (*you provided incomplete method information, but did say; "5 to 100 % ACN (+0.1 % TFA) gradient (in water) ", which is enough to know the cause). No column problem. No Lamp problem. No HPLC problem at all. The observation of the rising baseline with increasing TFA concentration over time is 100% normal and will be more exaggerated as you: view lower signal levels (Yes, it will look more extreme at 220 nm and 210 nm and 200 nm...) and/or increasing TFA concentrations (such as in a gradient analysis as the TFA increases in concentration over time).

• NOTE: Old or "dirty" TFA will greatly exaggerate this baseline effect. Always use FRESH TFA. Be sure and prepare freshly made mobile phase solutions (made up fresh every day. Do not store solutions. Make only what you need). Buy TFA in tiny bottles, not large ones. Yes, they cost more, but you will have fresh solution and save many days, weeks, months trying to troubleshoot problems caused by old solutions. *IF you are using OLD TFA solution, then this problem will appear to get worse over time. Are you using FRESH, new solutions?

To reduce the drift" seen from use of TFA, there are many solutions. Examples: (1) Run an identical analysis after the first one with only mobile phase injected, no sample. Then use the signal subtraction feature in your CDS to subtract out the "blank" sample from your original. (2) Change the concentration of sample used. Increase the sample concentration to reduce the baseline drift. (3) Decrease the acid concentration (only if the method used is reliable with the lower amount) to perhaps 0.05 %. (4) Change the acid used to a more UV transparent acid (again, make sure this is acceptable and results in a reliable method FIRST). (5) Collect a secondary signal in real-time for subtraction. At the same time the analysis is running, collect a second signal to later use for baseline subtraction. Note: The wavelength used must show NO SAMPLE peaks of any kind (outside of the detected range used), but at the same time be near the selected wavelength so it will still show some of the increasing baseline effect to be useful during subtraction (i.e. you use 240 nm. Maybe 240, 8 as you signal 'A' and 310, 10 as signal 'B'. Please note that this is an advanced feature and should only be used by a more advanced chromatographer, as their is a danger of collecting invalid data if not setup and monitored correctly).

It is normal, it increases because the concentration of UV absorbing material in the mobile phase increases over time during the graduent run. If you continue monitor the output after the gradient is back to the initial condition, it will drop. To prove this, run 100% A over time, you will see the low respons but constant (drift). If you do again with 100% B, you will see the same but at higher response (more UV absorption). No need to put the column in. Again, you should study the fundamental of HPLC method as William suggested or ask your professor at the school. Read a lot of books about it like I did 40 years ago.

The following discussion may come in handy: https://www.researchgate.net/post/How_do_you_get_rid_of_TFA_contaminations_in_your_LC_system