HPLC novice going crazy - please help!

[lube chemist]

A while back, I posted this thread looking for some help with method development. I've been continuing to work on this off and on and am still not confident in what I am seeing. Here's the story since then:

I purchased an Atlantis T3 column (with guard) and installed it. I started with a 50:50 ACN: water isocratic run and injected my analyte and the impurity I am trying to detect, as well as an ACN blank. I tried a bunch of injections with not much reproducibility and some extra peaks that were coming and going.

Today I went back to 100% ACN and actually think that this is working. Apparently my analyte is retained very well on the column. Running at 1.2 mL/min, it looks like my impurity elutes just after the solvent front and my ester analyte elutes at around 9 minutes. I switched to 10% water and after 15 minutes, no analyte. I went back to 100% ACN and both compounds showed up as before. HOWEVER, an additional large peak also showed up, almost dwarfing my expected peaks. I thought maybe it was something that came from the water, but the odd thing is that it gets bigger with every injection. What's up with that?

I'm flushing with 100% ACN right now and will rerun the analytes before the end of the day. I can post some chromatograms if that's helpful. I'm just totally baffled by this thing and I would really like to get it resolved and just use the method to analyze my reaction product. I hope you guys can give me some more helpful advice.

Thanks in advance!

[M. Farooq]

Hello,

Assuming are using RI detector alongwith PDA, few questions about your problem:

You see a peak that gets bigger and bigger with each injection when you change solvent composition with RI detector. The first thing to check is that are you allowing sufficient time to equilibrate. Binary mixtures of ACN and water have wierd behaviour in terms of fluid properties. There is a density or a viscosity maximum at a certain composition (perhaps 25% ACN). Such changes will give rise to refractive index changes. Do you see the same behaviour with PDA?

Secondly, is the composition of your sample solvent nearly same as the eluent? That may be worth a try.

M. Farooq

[lube chemist]

M. Farooq,

Thanks for replying. I should have mentioned that I am seeing this phenomenon at 205nm on the PDA detector. The RI detector doesn't seem to pick as much up. I see it within the same sample set - it gets bigger before I change solvent.

My samples are currently in 100% ACN, so I would guess that when I am running 100% ACN I shouldn't see anything more than perhaps the solvent front, right?

[danko]

If I understand it correctly, you’re running some strange kind of normal phase chromatography (100% ACN!) on a C18 column. The Atlantis’ partial affinity for polar compounds and maybe the silica support its self might be the explanation of why you get your compound/s retained, but I think you’d be better off if you tried the Atlantis HILIC column instead, which also would provide you an opportunity for controlling the retention and thus the selectivity by adding various amounts of water to the eluent.
Regarding the increasing peak, I assume some of the sample’s component isn’t eluting quite willingly, until it saturates the column and begins eluting after several consecutive injections.

Best Regards

[lube chemist]

danko,

How I ended up even trying 100% ACN is because I was so frustrated I decided to start there and work my way back adding in water. I expected that nothing would be retained with 100% ACN and that as I added in water I would see things start to separate. Because I could never see my ester peak, I was thinking that it was not being retained but perhaps it was being too well retained?

Unfortunately, I don't think that I can buy another column since we just got ths one. My motivation for doing this is to save money on sending samples out for GC-MS and try to get an idea of the purity in-house.

Incidentally on my final set of runs today, I got nice big peaks on the RI detector and the mystery peak only showed on the PDA.

[M. Farooq]

Hello,

As you say RI does not pick up the peak but the PDA does at 205 nm. The eluent is 100% ACN and so is the sample solvent. Your sample is a reaction mixture.

1. What is the peak shape of your large peak. Is it tailed or fronted. This might give a clue whether it is a component that is overloading the column or it is an artefact of the instrument.

At times the injector gives rise to peaks (as in my case), as a artefact.

2. Make sure your instrument is clean first. The column manual should have cleaning procedure. In order to be systematic and to avoid making assumptions about column "dirt", start with a cleaned up machine and column.


3. Try to inject, a dilute sample of individual reactant(s) and see if a peak comes out at the same position as this anomalous peak.

4. You should not see the solvent peak if the eluent and and solvent for samples are identical. Solvents peaks are good for GC. If you have a different solvent for the "reaction mixture" then it might be the case.

[Uwe Neue]

The retention of your analyte in 100% acetonitrile is rather unusual. It woudl require to be a purely and strongly hydrophobic compound, if only RP would be active. Please tell us a bit more about your analyte: approximate molecular weight, what functional groups there are etc. Then we may be able to design a better method.

[grzesiek]

i understand your problem - it's all the confusion, and i must say that i would be confused too if i were in your shoes right now

i don't know if it is good idea to even write something that could help you cos you already got a hundred responses and still not sure what to do right?

i would agree with the latest post - start with telling us everything you know about the analytes

[aceto_81]

How about doing step by step your method development again and post your results here?

What I suggest: cos your running an UV, you should be able to run gradients. Assuming you have a non ionic component, run twice a gradient from 0% ACN to 100%ACN in water. Now take a look at this chroms, are they reproduceble in terms of retention time, and peaks?
If so, do you see where your analyte is comming thru the column?
Now you can go further and take a look wether its possible to make your method isocratic, so you can make use of your RI.

One thing you should be carefull: if you dissolve your sample in 100% ACN, this is a strong solvent compared with your RP mobile phase, try to dilute in your mobile phase, or even weaker. So if it's possible to dilute in water, that should be great when running the above gradient.

Good luck

Ace

[lube chemist]

Thanks for all the replies - I really appreciate the help here. I'll summarize here what I worked on yesterday that prompted my post. This morning I was able to go through everything and I am getting fairly reproducible results with both detectors, except for my mystery peak.

I am making a diester (MW ~ 400) from 3-methyl-1,5-pentanediol and n-nonanoic acid. I'm using the nonanoic acid in excess. I want to try to quantify the purity of my product using HPLC in order to save $$ sending the sample out for GC-MS which is what the customer requires - I want to be pretty sure it is pure before sending it out.

I was using a very old Novapak 3.9 x 150mm C18 column (60 A, 4 um) but thought that my product was not being retained. I purchased a new Waters Atlantis T3 column. My system is Waters 2695 equipped with 410 RI detector and 996 PDA.

I started out yesterday trying to develop a method from scratch thinking that I'd try 100% ACN and then slowly add in water to get separation. Here is my first run. In all of these the blue is my product (the ester), black is the nonanoic acid and green is the ACN blank:

RI:


PDA (205nm) you can see the mystery peak starting to show up:


Then I switched to 90% ACN, 10% water and got this:

RI:


PDA (205nm) mystery peak start to show up more prominently at around 5 minutes:


Back to 100% ACN:

RI:



PDA (205 nm) Mystery peak is taking over:


Sorry that this is so long with so many images, but I'm not sure how else to describe it. And also confused about why my sample is taking so long to elute in 100% ACN.

[grzesiek]

surely your ester is very well retained

i think you should go for decreasing its retention - stronger solvent (THF?), low hydrophobicity column (C4, CN?)

[KarenJ]

The first thing that I would do when you are trying to increase the amount of water in your mobile phase is to increase the run time. 15 minutes is a pretty short run time. It is possible that your analyte is coming off at about 20 minutes but since you are stopping the run at 15 minutes, you see your analyte in the next injection at 5 min. Try extending your run time to about 30 minutes while you are doing development work. You can shorten it again later when you get this worked out.

What happens if you inject several blanks in a row (e.g. 5). Does your mystery peak go away? I would make sure that I have a clean blank response before trying to optimize the standards. If it doesn't go away, then it is most likely a mobile phase artifact.

Good luck,

Karen