Interesting retention time for C18 column

[Anthony_Ng]

Today my groupmate gave me a compound to check by LC-MS.

We use Zorbax XDB C18 column.

She said it has 2-3 amide groups, some amine groups and 1 benzene ring.

I run it with 80:20 MeOH:H2O isocratic and RT is 25min. and a bit broad-tailing peak.

Hence I change to 90:10 isocratic, hoping that the peak will come earlier and sharper. Interestingly RT is longer (35 min) and even wider.

Then I change to 70:30 isoctatic, RT is faster to 20min. and peak is sharper.

For my limited knowledge, the more the organic modifier, the faster the compound will come out in reverse phase. But today's result is totally reversed.

Can I rationalize her cpd is a salt so that have above result in RP column?

[Uwe Neue]

Silanols?

[Anthony_Ng]

She synthesize it in the lab. And purified by flash column. So I got only 1 peak in the chromatogram.

MS showed the molecular weight for her compound is correct. (MW:588)

[sassman]

When you change the water/solvent ratio, you are changing the effective pKa of the amine groups. More water will make the compound more easily ionized. More water also will cause residual silanols to become more ionized and have a stronger interaction with the amine group. Try using a buffered mobile phase (perhaps 10mM NH4OAc at a pH of 3?) and see if you get the same effect.

[Kostas Petritis]

I would just use 0.1% formic acid in order to minimize the residual silanols interactions and take it from there. This will help to (hopefully) improve the peak shapes...

[Anthony_Ng]

This is the structure. My colleague said if she didn't add dilute HCl in the work-up, the polarity of the cpd is higher (and the cpd is not stable itself). After adding HCl in work-up process, the polarity become smaller and easier for her to purify in flash column with silica gel.

Hence I can see MW:588 for molecular ion and also [M+Cl]- adduct in the MS.

[lmh]

Just a couple of comments and questions:
Firstly, I think you may have a minor slip-up in your structure; is the left hand side supposed to look like the right?
Secondly, are you doing ESI? If so, I'm a little surprised about the chloride adduct. I find that even when a sample has been prepared with HCl, by the time it's been through chromatography the HCl is long since diluted to an extent it doesn't much affect ionisation. I would expect, at first glance, that this ought to give quite a reasonable positive ion. Did it? What was the mass? Does the chloride adduct have the right isotope peaks?

[Anthony_Ng]

Yes, left hand side is the same as right hand side, if my colleague draws it correctly.

Yes, we have Agilent single quard 6120 linked to LC. It can spend 50% time in +ve and 50% in -ve. Its spray chamber is designed for mutil-mode ionization. i.e. ESI and APCI.

It has positive ion [M+1]+ (major), [M+Na]+; and negative ion (also with chloride adduct M+35). Isotope ratio for chloride adduct is also correct.

Does it mean that the "long lasting" chloride should not be there after her flash column purification?

[lmh]

No, it sounds like you've got all the right ions present. I was slightly surprised that the chloride stuck around through chromatography, but I was really just wondering whether there's something special about this molecule that gives it a high affinity for chloride. And obviously anything special about this molecule might help us guess why it has such odd RTs.

If it were me, I'd wonder if I'd topped up the bottles with the wrong solvents, or whether I'd got the lines mixed up. But I don't dare admit such unprofessional things have ever happened in an instrument associated with me!

On the drawing, sorry, my mistake. I automatically drew the RHS with the sidechain on the N. Otherwise you have an N that ought to be an NH... but it makes no difference to mass.

[Anthony_Ng]

Here are the chromatograms:

90:10 MeOH:H2O (RT 32 min)


80:20 MeOH:H2O (RT 23.9min)



70:30 MeOH:H2O (RT 21.4min)



When I added 0.1% Formic acid to mobile phase, the peak become very sharp and come out before the break-through peak:


There I think sassman is correct. Water changed the ionization of the compund and the residual silanol group of C18 column.

Also, when Formic acid is added, the compound is fully ionized and hence there is not much interaction with C18, making it come out even faster than break-through peak.

Please let me know if my rationale is correct or not. Thanks!!!

By the way, what is the pros and cons of using NH4OAc versus Formic Acid as an additive?

[HW Mueller]

This is another case where we are not informed systematically and sufficiently. You state above that the compound is not stable unless HCl is added, then you do chromatography without acid. What are you chromatographing??
You furthermore state that you see an M+Cl peak, but I don´t see what type of sample you are using to do the MS. Is it a HCl-crystal or a HPLC eluate?
Furthermore, what is a breakthrough peak in your opinion? If you think the compund (did you do a LC-MS here besides the UV?) is ahead of tm (to) you might be wrong.
One uses a buffer to get a stable pH. The pH is optimized for optimum retention time, etc.
Also, is this another one of those compounds which need to be analyzed in connection with human use?

[Anthony_Ng]

Sorry that I mention the situation unclearly.

I ask my colleague again this morning. This compound is somewhat like "guanine". You find that the left and right hand side have some portion of the structure similar to guanine.

She said if dilute HCl is not added, it is relatively unstable, that is easy to hydrolyse. Usually they add dil. HCl to make it become guanine hydrochloride. Hence, take this for their own flash chromatography.

What I got after their own purification is 0.1mg sample in a vial (usually white solid without solvent). I added 1ml HPLC MeOH to dissovle it and run by LC-MS. The mobile phase is 80:20 MeOH:H2O (without acid in both mobile phase). Hence I got broad-tailing peak as posted some time before.

Then I tried to add 0.1% formic acid in mobile phase H2O, and got a sharp peak.

This is the positive spectrum (without acid in mobile phase). [M+1] is dominant, with some [M+Na].


This is the negative spectrum of the same run (again no acid in mobile phase). Here [M-1], and mass: 623, which I think [M+Cl] are present. Note that ratio of 623:625 is about 3:1. So I guess the 625 is come from isotope of 37Cl.


Here is the positive spectrum (with 0.1% formic acid added in H2O mobile phase). Again [M+1] is dominant.


And this is negative spectrum (with 0.1% formic acid in H2O mobile phase). This time [M-1] disappeared, 623 still there but a new mass 633 appeared. I guess this is [M+HCOO] which come from formic acid.


The spectra 3 and 4 are indeed come before breakthrough peak (injection peak). (Note the top left corner of the RT, which is the same as the chromatogram). I am not sure..... if the compund is ionic/salt form, it will have less interaction with C18 column, so it come out very fast?

For this compound, I am not sure if it is for human use because we are working in a syntheic group in a university. My colleague doesn't disclose too much to me.... (I am just a "small potato" technician)

[HW Mueller]

I am stumbed. not seeing from where the Cl is coming (except in the case of the formic acid . . ., where everything comes out front), nor the Na. My suggestion is to follow the advice already given by some, namely optimize the pH with a buffer and forget about the MeOH-H2O chroms. I would try to collect enough of a peak at a useful retention time to characterize it with NMR and possibly other methods.

[lmh]

If you're worried about the adduct peaks, you may be able to reduce their intensity by careful choice of source parameters. The formate and chloride adducts are effectively cluster ions ([M-H] + HCl, [M-H]+formic acid), and will (probably!) be discouraged by conditions that are more drying or more fragmenting.

About where the chloride comes from: if your colleague has supplied a dried white solid, it may not have been dried very thoroughly. In any case, if it has an affinity for chloride, it will be a chloride salt, and although usually the chloride is washed away during chromatography, a thing with a very high affinity will still manage to scavenge a chloride ion in ionisation. Also, if a chemist says they used dilute HCl, it may not mean 1-2mM. It may mean "a-few-drops-in-a-tube-this-big" and turn out to be in the order of 1M!

[HW Mueller]

This is -N: + HCl >> -N:H+ + Cl-, where is the high affinity for Cl-? If I have seen this correctly there have been cases of coelution of ion pairs in aprotic media?? Of course, at to it is not surprising to have Cl-.
I am affraid that Anthony might see traces of everything.