Negative Dip in HPLC Chromatography

[skeetch1979]

OK, brief run down of the instrument (following company temperature specs/SOP):
System: Agilent 1260 Infinity II
Detector: RID
Flow Rate: 0.9ml/min
Mobile Phase: DI Water (Milli-Q)
Pump: Quaternary pump (fed by mobile phase after Degasser, MCGV not connected - i.e. by-passed)
Sample loop: 100uL
Column compartment temperature: 50C
RID Temp: 35C
Column: LiChrospher 100 RP-8 (5um), LiChroCART 250-4, 4mm x 250mm
Injection volume: 20uL
Draw speed: 200 uL/min
Eject speed: 400uL/min
Delay between draw and eject: 1sec

I have recently been noticing a negative dip in my Chromatogram starting around 8.5 min and lasting through about 11in. The dip is only seen during an injection. It is not seen during a normal flow through of the mobile phase with the signal being monitored.

My mobile phase (Milli-Q DI Water) is degassed, but the sample solvent (which is also Milli-Q DI Water) is not.

The blanks that I have been injecting are vials filled with the Milli-Q DI Water.

I have attached an example chromatogram here: https://imgur.com/a/YFEbL2d

I have replaced multiple items within the System. Brief rundown: Sample Loop, Needle, Needle Seat, Rotor Seal, Purge Valve Filter, Pump Seals (2), Analytical Head Seal, manu-Cart NT Cartridge Holders, Phenomenex SecurityGuard and filter

All of these changes and no improvement. Based on the peak shape and dip, I am leaning towards gas/air being introduced with the injection but I do not know where. Any insight would be helpful.

Thank you in advance!

[Multidimensional]

A few comments in no order:
(1) Your column temperature and RID flow cell temperature should be the same temperature, not different, as you have setup. Insulate the lines, keep them as short as possible and use the SAME temp.
(2) Use a smaller injection loop. A 100ul loop making 20ul injections will lead to poor reproducibility. Use a 20ul loop, with 3x overfill for accurate injections.
(3) You state your mobile phase is pure water. It will grow bacteria, algae ect very quickly, esp in the column at elevated temps as you have. This leads to two issues: (a) Column fouling over time and contamination of the flow path. Wash the column down with a "stronger" solution frequently to remove retained material and prevent fouling. Refer to the manual that came with your specific column for suggestions regarding which chemical it is compatible with. (b) Change the water you use each day. Flush the lines with fresh filtered water before use and use the same water for diluting your samples with too.
(4) The chromatogram you included shows your injection peak around 3 minutes (as expected) and your "sample" eluting off at around 9-minutes. The "dip" you see is because the sample's RI is different than water (normal).

Other notes: You noted that the MCGV was disconnected for this analysis. Is that because you have a broken MCGV (leaking channels) or because someone suggested you simplify the flow path as you are running isocratically and do not need it? Running with it by-passed would only be recommended IF you are drawing liquid from two or more channels, not from a pre-mixed solution (or just water in your case).
If the HPLC system has been verified to be in proper working order, none of the replacement parts you added were needed to explain the "dip" observed in your chromatogram. The "dip" is not due to a hardware issue, it is a product of the method. No need to "fix" something that is not broken. If possible, speak with someone local who is an experienced chromatographer for help in using the instrument and also reviewing the proposed analysis method.

[skeetch1979]

Hi @Multidimensional,
Thanks for your reply.

So a response to your points.
1. I agree that the temperatures should be the same between RID and column chamber. These parameters, however, are a company wide accepted SOP for this type of analysis. Other regions use these parameters without issue. So, it is not an option to change the current set temps.
2. I have a 100uL analytical head, Agilent suggested I use a 100uL sample loop. Our QC department has a similar configuration and they are not seeing any issues.
3. I agree with you regarding potential Algae growth and using fresh solvent everyday. I will try and be more vigilant in this practice. My solvent/mobile phase is solely Milli-Q DI Water and the solvent bottle is an amber 2 Liter bottle.
4. So, the both the sample solvent and the mobile phase solvent is the same in this case - both from the same source, Milli-Q DI Water. The only difference between the two is that the mobile phase passes through a degasser, while the sample's water does not.

The MCGV was disconnected because one of the channels was not fully closing and allowed cross channel contamination. At the time I was investigating the analysis of other compounds using our DAD - which required an acidic mobile phase blended with Acetonitrile. The acetonitrile was leaching into my other channels causing peaks issues. By-passing the MCGV has resolved the issues up to this point in time. I am only running one mobile phase without any sort of gradient so in effect the purchase and inclusion of a MCGV was not necessary from the start.

If air or gas was being introduced where would this occur in an LC? If I am using DI Water as a mobile phase, is there anyway to use IPA as a needle wash solvent without the risk of residual IPA being injected along with my sample after a rinse?

[Multidimensional]

A response to your points:
(1) Just because someone else does something, that does not make it right. They often do not know what they are doing either. The column temp AND the RID flow cell temp should be the same. The idea is not cause an additional temperature gradient through the system. Doing things correctly is important so be careful when copying what "others" do.
(2) Loop size: I thought you were manually injecting samples... your comments imply you may be using an autosampler. This makes more sense. If so, and it has a 100ul volume loop, then you want to leave it alone. 20 ul is fine in this case.Note: MAX Injection volume should be no more than 3% of the column volume (max, far less in normal use). Your column is 4 x 250mm so has an estimated column volume of ~ 2200 ul. 3% of this is 66 ul. Practically speaking that would be the largest loop and/or injection size to use. Not an issue.
(4) The sample does not need to be degassed (no one does this). No need to worry about this as any bubble present would appear at Tzero. Not an issue.

MCGV damage: When these valves get to be more than 10-years old, sometimes one or more channels will show seal leakage resulting in cross flow contamination. Using ALL of the channels on a regular basis, flushing with IPA periodically, and maintaining a clean flow pathe will extend their lifetimes. By-passing it as you are doing is fine for now, but you really should have the valve replaced or serviced [More info here: https://hplctips.blogspot.com/2021/08/h ... valve.html ].

BTW: Other errors noted.
Your A/I's draw and eject speed should be the same, not different. You are injecting water, no issues with viscosity so set them both to 100ul/min or 200ul/min max. No one should be messing with these values for your application. Also, no delay needed either. Sounds like someone was messing with your instrument and they forget to create a new method (overwrote the old one leaving all of the previous settings in there)..Careful, they may have programmed other settings that make no sense too (check solvent compressibility too). We see a lot of people make these errors using settings that the "other" person used.

You wrote: "If air or gas was being introduced where would this occur in an LC?" Please have someone with HPLC training help you on this project. Unless you have a leak, no "gas" is being externally introduced into the system. The gas is already there, in your mobile phase of water. The pump compresses it, squeezing out the trapped gas inside causing problems for the pump and other modules. An inline degasser or helium sparging system will reduce the amount of trapped gas to a level that is low enough to not cause pump problems.

[skeetch1979]

1. I will speak with the analytical team about making the column and RID temps the same. Not sure why these temps were chosen. The mobile phase is water so viscosity should not be an issue.
2. I am using an autosampler. Even though my system is an Agilent system, company wide we use Chromeleon (I am using Chromeleon 7.2, other colleagues are still on Chromeleon 6 due to the age of their instruments and PC's). Chromeleon came with these default draw and eject speeds and delay. I can modify them to be the same and eliminate the delay to see if this helps any issues.
4. Good to know samples are not degassed. I manage an R&D lab and only have the LC and 2 GC's to follow our reactions and solvent adjustments. I am far from a QC lab!

To be clear, my flow path is: Solvent Reservoir bottle with Milli-Q DI Water -> 4-channel degasser (1 channel is only used so this has the DI Water, the other 3 channels are simply filled with IPA and not used) -> quaternary pump -> Rotor Assembly (connected to the pump, column, sample loop and analytical head) -> analytical head -> sample loop -> needle -> seat.

My mobile phase is degassed via the degasser before entering the quaternary pump, column and sample cell.

You mention air would be at Tzero, so in my chromatograph what could be coming out late between 8min and 11min? IS that not air or other gasses?

[TylerSmith123]

Multi was referring to air introduced during injections, that is the air the would be present at tzero following an injection. However, if you have a leak somewhere that is a different issue. I think Multi has pointed you in the correct direction with all of their advice, but I want to return to earlier in the conversation. What does your method look like, what are your washing protocols like and what is the regularity of which you make/change your mobile phase. Have you tried Multi's suggestion regarding washing the system down with a strong solvent (just go with manufacturer recommendations) and observed anything? What have you implemented thus far and have you seen any improvements?

[Multidimensional]

skeetch1979, You Wrote: (1) "You mention air would be at Tzero, so in my chromatograph what could be coming out late between 8min and 11min? IS that not air or other gasses? ".

- Please review my previous reply. NOT GAS. You are observing the signal deflection for YOUR sample in the chromatogram (not chromatograph). The sample material has a RI which is different than water (normal).

skeetch1979, You Wrote: (2) "4-channel degasser (1 channel is only used so this has the DI Water, the other 3 channels are simply filled with IPA and not used)".

- Is your Agilent 1260 series pump (not sure which model you have) 3 or more years old? If it is, then your degasser is broken. If used on a regular basis (normal), the degasser in this model lasts a maximum of 3-years before disintegrating from wear (normal for this model). Also, the vacuum degasser is designed to be used with ALL channels used, ALL of the time, not one channel. Plugging three channels filled with IPA is not allowed unless you are storing the degasser. The IPA quickly evaporates inside the channels leaving air. Vacuum degassing channels can not be left with air inside them during use. Doing so accelerates the wear on the vacuum system. The degasser is also designed in such a way as to not alert you that it is damaged until long after the internal system has been destroyed (by design), contaminating the HPLC flow path. The error indicators are worthless on these modules as they alert you to problems long after they should have been serviced. We see many fail after just one year (right outside warranty) with the longest lifetime recorded so far as 3-years (and that is after servicing thousands). Flow path contamination results from ignoring them.

[Multidimensional]

More Vacuum Degasser Info: By design, the vacuum degasser in your pump is designed to be used with all channels, all the time. If you do not plan on using all of the channels, then we recommend that you routinely flush each line (100%, one-at-a-time) at least every week to maintain flow of fresh solution through each channel (chamber). Alternatively, you can interconnect unused channels to channels you actually use, in a serial fashion (with 'C' tubes). This will increase the degassing efficiency overall plus maintain the chambers in flowing liquid. Never plug unused channels. Never leave them open to air. Use of plain water as mobile phase leads to bacterial contamination as noted earlier (inside the degasser chambers and tubing). This is one of many reasons why the complete flow path must be routinely washed down with a stronger solvent to both flush off retained material from the column AND flush out debris from the flow path. Contamination of the flow path must be avoided at all cost or all data collected will be invalid. We teach these basic skills in our HPLC training classes and find that 95% of the clients do not know this.Please get some local help from an experienced chromatographer to insure you are making the best use of the system..

[skeetch1979]

Hi Tyler,
I appreciate your reply.

My method looks like this:
Flow: 0.9ml/min
Mobile Phase: DI Water
Column Oven: 50C
RID Temp: 35C
Detector: RID
Run Time: 12min
Post Time: 2min
Needle Rinse: 5 seconds (DI Water from Mobile Phase reservoir)
Calibration Curve: 9 points (1% EtOH, 2% EtOH, 3% EtOH, 4% EtOH, 5% EtOH, 20% EtOH, 70% EtOH and 99.8% EtOH)
Injector: AutoInjector
Injection: 20uL
Auto Sampler: G7129A
Quat Pump: G7111B

Blanks: Filled with Water from Mobile Reservoir which is DI Water

Multi was referring to air introduced during injections, that is the air the would be present at tzero following an injection. However, if you have a leak somewhere that is a different issue. I think Multi has pointed you in the correct direction with all of their advice, but I want to return to earlier in the conversation. What does your method look like, what are your washing protocols like and what is the regularity of which you make/change your mobile phase. Have you tried Multi's suggestion regarding washing the system down with a strong solvent (just go with manufacturer recommendations) and observed anything? What have you implemented thus far and have you seen any improvements?

[skeetch1979]

That is good advice to not store or plug unused channels.

Where would I find a 'C' tube? Do you have a part number?

I do not flush them weekly but that is also a great suggestion.

I can have a peek at my system using lab advisor. There are no warnings at the moment and the degasser is set to and holding at 100mbar. It is about 3+ years old.

Thanks again for sharing your wisdom, knowledge and experience.

More Vacuum Degasser Info: By design, the vacuum degasser in your pump is designed to be used with all channels, all the time. If you do not plan on using all of the channels, then we recommend that you routinely flush each line (100%, one-at-a-time) at least every week to maintain flow of fresh solution through each channel (chamber). Alternatively, you can interconnect unused channels to channels you actually use, in a serial fashion (with 'C' tubes). This will increase the degassing efficiency overall plus maintain the chambers in flowing liquid. Never plug unused channels. Never leave them open to air. Use of plain water as mobile phase leads to bacterial contamination as noted earlier (inside the degasser chambers and tubing). This is one of many reasons why the complete flow path must be routinely washed down with a stronger solvent to both flush off retained material from the column AND flush out debris from the flow path. Contamination of the flow path must be avoided at all cost or all data collected will be invalid. We teach these basic skills in our HPLC training classes and find that 95% of the clients do not know this. Please get some local help from an experienced chromatographer to insure you are making the best use of the system..

[Multidimensional]

"C-Tubes" are simply short low pressure Teflon lines with fittings. In your case, lines ~ 6 or 10 cm long with 1/4-28 fittings to connect the degasser ports together. We make these 'C-tubes' up custom for clients that purchase our products and/or use our services, but I think Agilent sells them in their catalog too.

"I can have a peak at my system using lab advisor. There are no warnings at the moment and the degasser is set to and holding at 100mbar. It is about 3+ years old."

- The pressure (vacuum) reading is not evidence of the system working (the system will burn itself out before anything on the screen shows up). If it is 3 or more years old, then it is broken. As noted, error messages do not generally appear until long after it has failed (sometimes YEARS !!!). This is one of many disadvantages of using inline vacuum degassers. **DISCLAIMER**: In addition to provide professional advanced level training and consulting in the use of HPLC and LC-MS systems, our business professionally provides LC vacuum degasser testing and repair services for ALL brands and models of HPLC (and LC-MS) degasser modules (and this is just a small part of our overall business). We have serviced thousands of degassers over the past 30-years for the instrument manufacturers and end-users. If interested in having your degasser tested or serviced, then please review this link: https://www.hplctools.com/agilent_degas ... ooting.htm

[Multidimensional]

Note: I see you wrote that you include a "Needle rinse" as part of your method? WHY? This is not needed at all with the A/S you have. In fact, it will increase the chances of contaminating your HPLC flow path AND also may result in ghost peaks or peaks with unusual shapes etc. This is understood, once you understand the flow path of your specific HPLC instrument. Please use no needle wash as standard.

Please consider hiring a professional chromatograher to come on site and review your instrument setup and method(s). Troubleshooting can not be done via the web or phone for such things. My boss often says that after just an hour on-site, dozens of serious problems are identified and explained to the clients. This is a huge help to them as w/o an expert on-site, they do not know what they are doing right or wrong.

Also, should we assume that your whole method is for determining ethanol concentrations by HPLC? IF so, it looks like many more problems exist. Please have your school or business hire a chromatographer. Doing so will save so much time and money. HPLC is a complex technique that can not be learned by taking a class, watching videos, looking at web forums like this one. It takes the average scientist FIVE years, under someone who can help and train them, to acquire a basic level of experience. Emphasis on 'basic' level. NOt yet ready to develop methods. Anyone can "run' an HPLC, but in order to do so an generate valid data requires many years of training. This is not something that you can just figure out for yourself. We regularly visit labs that have done just that and hire us,after they have been audited (and failed) or shut down, to explain what they are doing wrong. The reason is usually that they have no professionally trained chromatographers on staff.

[skeetch1979]

An Agilent Service tech came today and worked to troubleshoot the instrument.

Unfortunately he left without resolving the issue, but will return with some replacement parts that are found in the autosampler module.

He did replace the Degasser in an effort to rule that out or confirm that it was contributing to the problem. He stated the degasser was still in functioning order (even at 3+ years) but wanted to be sure it was not the source of the problem. Replacing the degasser did not yield different results - the issue persisted at the same intensity.

He seems very knowledgeable and familiar around the working and operations of an LC and its components.

He hopes to return tomorrow and troubleshoot the system more - focusing on the autosampler.

------

Multidimensional: I am sure you have many many years of experience and knowledge but I take issue with the fact that you automatically wrote off the degasser as broken since it was 3+ years old. You had not seen my degasser or had hands on it but still came to the conclusion that it was broken.

In my many many years as a scientist I know that it is not good practice to form a conclusion or jump to conclusions without putting in the due diligence and testing. As many times as you may see something as an outcome in similar situations, it is not good practice to assume that in every case things will have the same outcome or result.

So while I appreciate your knowledge, experience and input I would be careful reaching a conclusion unless you have had 'boots on the ground' with that particular instance and were able to do your hands on testing and troubleshooting. It helps you and the people you are trying to assist.

Thank everyone though for their thorough input and I will keep everyone posted, hopefully with a solution, that will help the broader cause and others who may come across this problem.

"C-Tubes" are simply short low pressure Teflon lines with fittings. In your case, lines ~ 6 or 10 cm long with 1/4-28 fittings to connect the degasser ports together. We make these 'C-tubes' up custom for clients that purchase our products and/or use our services, but I think Agilent sells them in their catalog too.

"I can have a peak at my system using lab advisor. There are no warnings at the moment and the degasser is set to and holding at 100mbar. It is about 3+ years old."

- The pressure (vacuum) reading is not evidence of the system working (the system will burn itself out before anything on the screen shows up). If it is 3 or more years old, then it is broken. As noted, error messages do not generally appear until long after it has failed (sometimes YEARS !!!). This is one of many disadvantages of using inline vacuum degassers. **DISCLAIMER**: In addition to provide professional advanced level training and consulting in the use of HPLC and LC-MS systems, our business professionally provides LC vacuum degasser testing and repair services for ALL brands and models of HPLC (and LC-MS) degasser modules (and this is just a small part of our overall business). We have serviced thousands of degassers over the past 30-years for the instrument manufacturers and end-users. If interested in having your degasser tested or serviced, then please review this link: https://www.hplctools.com/agilent_degas ... ooting.htm

[Multidimensional]

Skeetch1979: The Agilent tech that examined your HPLC system has no formal training or ability to troubleshoot the degasser in your HPLC pump. Specialized equipment and training are required to do so and Agilent does not provide their techs with these resources. They, like other instrument manufacturers, have found it is better to just replace the degasser when a problem appears. This provides an easy solution for the tech (no lost diagnostic or part swapping time) and insures that the customer gets a clean functioning degasser in the process. Your degasser was most likely damaged, again based on age, so Agilent's only option is to replace it with a known new one per the troubleshooting process. Yes, it would be great to have it on the bench to confirm, but time is money and Agilent does not have the option to "test it", only replace it.

• Note: I do not believe any of my comments suggested that your main problem was caused by using a broken degasser (which you had). Instead, the degasser was mentioned because when they are 3 or more years old, they are always broken and it is something that you should address, just like the broken MCGV too. These two things should be repaired in you wish to use the instrument. Running a broken system complicates troubleshooting and often leads to problems with data validity and method reproducibility.

Based on the basic info you provided so far (and all of the information you have not provided), there are many reasons for your issues, but the number one issue is training. There are many problems with your proposed method and it is likely that someone knowledgeable on-site (most Agilent techs are not experienced chromatographers) could review your method setup and goals to provide constructive advice in how to proceed. It is very hard to help new users with troubleshooting advice over the web. Such problems can often be solved by an intermediate level chromatographer on-site in very little time.

[skeetch1979]

"The Agilent tech that examined your HPLC system has no formal training or ability to troubleshoot the degasser in your HPLC pump." <- How do you know? Do you personally know the tech that serviced my HPLC? How do you know that he did not have other training before joining Agilent? You assume that Agilent was his first and current job and does not have any other experience.

I agree that their solution is to replace versus diagnose - it is a cheaper quicker option.

"Note: I do not believe any of my comments suggested that your main problem was caused by using a broken degasser (which you had)." <- Again, you are making the assumption without 100% knowing that the degasser was broken. As I stated, I saw no differences in my baseline, quality of chromatogram or overall performance between the old/replaced degasser and the new/replacement. But if you think the old one was broken, based on the fact that nothing changed in the quality of the chromatogram or baseline or overall system performance, then the new degasser is broken as well.

Fact of the matter is that I still have a negative dip that has not changed in intensity with all of this work so far.

How often does the Stator, Rotor Seal or metering valve cause issues?

A complete system pressure test was performed today and in all cases it passed. No leaks present. The dip most likely is caused by air being introduced - but not from a leak.

You suggest an intermediate level chromatographer to help and not just a run of the mill Agilent tech (which obviously lack chromatography experience) - any recommendations for a servicer in NJ?

Skeetch1979: The Agilent tech that examined your HPLC system has no formal training or ability to troubleshoot the degasser in your HPLC pump. Specialized equipment and training are required to do so and Agilent does not provide their techs with these resources. They, like other instrument manufacturers, have found it is better to just replace the degasser when a problem appears. This provides an easy solution for the tech (no lost diagnostic or part swapping time) and insures that the customer gets a clean functioning degasser in the process. Your degasser was most likely damaged, again based on age, so Agilent's only option is to replace it with a known new one per the troubleshooting process. Yes, it would be great to have it on the bench to confirm, but time is money and Agilent does not have the option to "test it", only replace it.

• Note: I do not believe any of my comments suggested that your main problem was caused by using a broken degasser (which you had). Instead, the degasser was mentioned because when they are 3 or more years old, they are always broken and it is something that you should address, just like the broken MCGV too. These two things should be repaired in you wish to use the instrument. Running a broken system complicates troubleshooting and often leads to problems with data validity and method reproducibility.

Based on the basic info you provided so far (and all of the information you have not provided), there are many reasons for your issues, but the number one issue is training. There are many problems with your proposed method and it is likely that someone knowledgeable on-site (most Agilent techs are not experienced chromatographers) could review your method setup and goals to provide constructive advice in how to proceed. It is very hard to help new users with troubleshooting advice over the web. Such problems can often be solved by an intermediate level chromatographer on-site in very little time.