does the concentration of an ion-pairing agent have any effect on retention time? I have been using ACN: water (90:10) and an ion-pairing agent( N-Octane sulfonic acid sodium salt; 1mM) for the analysis of a basic drug. The retention time is 1.6 min. When I used 0.1mM ion-pairing agent the retention time is 6.7 min with a broad and tailed peak. I am not able to understand why did it happen? Please expalin it to me.

Yes generally ion pairing agent concentration should have effect on the retention. More the ion pairing agent more will be retention.

Check the following link, it describes retention of polar compounds without ion-pairing reagent. LC/MS compatible conditions, simple mobile phase, good peak shape and efficiency

http://www.chromatographyonline.com/lcgc/data/articlestandard/lcgc/102004/87498/article.pdf

Regarding Figure 3 of that link, I predict the TFA would drastically reduce sensitivity for the acids in LC/MS (negative ion), versus formic or acetic acid modifiers. If you've done the experiment and shown otherwise, I'd be interested to hear it.

Dear MG,
TFA is one of several additives you can use with our columns. You can also use ammonium formate, and ammonium acetate at pH 3-4, formic acid, acetic acid etc (different concentrations). Mobile phase can be from 0% organic to 100% organic. Our mostly common mobile phases are ACN-water 40/60 with 10-50 mmol of ammonium formate (pH=3), ACN-water-TFA=40/60/0.1 and ACN-water-HCOOH=40/60/0.1

Now I am really confused, having considered that dearly hated TFA as an ion pairing agent all this time (in protein applications, etc.).

Also, aren´t these mixed mode phases a little like those vacuum cleaners which are convertible to a drink mixer (or so): Don´t work correctly in either function??

AllsepTech, your link was to a chromatogram using TFA, with the statement that the separation is LC/MS compatible. I think the LC/MS compatibility is debatable in this case. From my experience with reverse phase, I would not expect the chromatogram to look the same with HCOOH in place of TFA. Otherwise, why not just use formic acid instead of TFA if LC/MS compatibility is desired?

HW Mueller, I recall some debate on this board in the past as to whether TFA is an ion pair reagent.

DEar MG,
Please download our Primesep Brochures to see explanation for mechanisms and few examples of LC/MS compatible conditions (ammonium acetate, ammonium formate, formic acid). The method in the link was one of our first methods and now we are trying to redevelop some of our methods of "more friendly" LC/MS conditions (without TFA)

Dear HW Mueller,
We think that if people know how to use the mixed mode columns they have almost no limitations in separations. Once everybody realize the science behind our columns the method development becomes very fast and easy, we have developed several methods for our customers in very very short time (a few hours) and people struggled with these methods for months using conventional C18 or ion-exchange columns

Missing link for the brochures
http://www.allsep.com/Brochures_Home.php

Dear HW Mueller
Every HPLC column is a mixed-mode column. RP or ion-exchange is just a matter of quantity not absolute characteristic. We all know that silica based RP columns have ion-exchange properties and polymer based ion-exchange columns have reverse phase properties. You will not find on the market a single column that has clear only one type of interaction with all analytes possible.
Using your analogy every column that not so called “mixed-mode” is a like full size vacuum cleaner which is convertible to a toy drink mixer (or so).
For mixed-mode columns I like another analogy if you mix bread with cheese and little tomato souse you will get pizza. If you take only a bread and very little cheese you will be able to eat it, but you would need something else for the diner. If you have only cheese and very little bread combination is also eatable but probably not ideal (beer requires). Even if you mix bread, cheese and tomato in a bowl it still is not a pizza. It required certain proportion, way of cooking, proper ingredients, and skillful cook to make a good pizza. The same true for Primesep mixed-mode columns. It is not just mechanical mixture of RP and IC packing material. There is some cooking involved. Instead of having uncontrollable second mode interaction which you fight by adding modifiers and it still show up one way or another with different analytes you will be better of having build in secondary interaction which easy to control, well defined, and predictable.
In the end, every chiral column is a pure mixed-mode column. I never heard a complain that interaction there is not pure RP or IC or polar. The simple answer is - it will not work if it would be only one interaction in chiral separation. It works because different type of interaction play different role. The same is true for non-chiral separation, different modes play different roles, but result usually is not achievable with only one separation mode.

Actually, this is an interesting question: why do you need to cook stuff instead of just mixing a reversed-phase with an ion-exchanger?

Mainly, if you just mix stuff, you will get something like those vacuum cleaners which are convertible to a drink mixer that do not work correctly in either function as properly notice HW Mueller.

Well, this answer is neither enlightened nor enlightening. I don't even know, if your material is a mixed bonded phase or just a mixture of an ion-exchanger and a reversed-phase packing. After I take apart the column, I will of course know the difference. But how would I be able to know the difference without taking apart the column?

I think that sevral companies made attempts to make mixed mode material by mixing ion-exchange and reverse phase materials but run into problem with packing and reproducibility from lot to lot. Neither I am aware of any MM colums made by mixing or tried any of Primesep columns. I think that somebody in my department has one or two Primesep columns.

Nick

If you go on www.primesep.com you will find plenty information on stationary phase chemistry and columns properties. By short, every column is made of a ligand bonded to silica surface. And every ligand has both ion-exchange functional group and long alkyl chain.

I can not see a reproducibility problem from lot to lot. You weigh out the stuff, then you mix it. Unless you can't read the numbers on the balance, the reproducibility should not be problem.
I would think that the reproducibility problem would be greater with a mixed surface chemistry, either with a single step bonding process or a multiple step bonding process.
I am still wondering, why Primesep is making such complicated chemistry.
Zelechonok, what is the reason?

Dear Uwe,
I just wondering if you ever tried to pack a mixture of reverse phase and ion exchange silicas and achieve good peak shape and good plate count (80K/meter)on a constant bases. If you did, I am ready to reproduce your procedure and compare our Primesep columns against "mixed" material.

Dear Uwe, you should be familiar with problem of getting several reproducible lots of the same RP material. Pharm companies often use a reserved lot for validated methods or they buy columns in bulk. If Waters does not have this problem then everybody else dose. Multiply this problem by factor of four when you going to mix two materials. Since each material has own variations in surface chemistry, particle distribution, pores size, etc.
Problem with packing too.
But most importantly the mechanical mixing is just simple summation of two columns retention. You don’t need even to mix two packing. Take two columns and connect one after another.
When you make one ligand with both properties the result is totally different. It is synergy phenomena.
Most obvious results are:
Columns show no silanol interaction with any analytes without endcapping.
Columns are stable from pure water (no dewetting) to pure hexane (no swelling)
The same column works in reverse mode, normal mode, ion-exchange mode, HILIC mode, ion-ecsclusion mode, ligand-exchange mode.
There are over 100 methods developed on these materials. And every one of them is a challenge for ether RP or IC columns or combination of both.

All I was trying to say is that I have a problem, theoretically initially, with stuffing diametrically opposed phenomena into one column and expecting it to be optimal for anything: If one makes a C-18 column you have, of course diff. adsorption phenomena + SiOH interactions. Now you add another function, either mechanically or via cooking. Now you will have diluted the interactions of the pure C-18 column....that is certainly "deoptimalized" (by definition, even).
By far the largest problem of HPLC seems to be resolution (remember the case of finding 100 substances, via GC, in a homogeneous HPLC peak). If you need the optimum resolution one should do two anals with two columns.
I see a use for mixed mode if you are operating far from the limits of HPLC.

There is no dilution of any kind. Hydrophobic interaction stays as it is in any RP column. Neutral compounds will be separated in order of their hydrophobicity. The weak silanol interaction replaced in MM with much stronger and well defined ion-exchange. Situation is not more complex, it is more predictable (assuming it is true MM phases, not mechanical mix).
Interesting phenomena observed with neutral compounds. Because of presence of ion-exchange the phase can be loaded with contra-ion such as amines, ammonia, metals (silver, sodium, potassium, cupper, lead, calcium etc.) and new contra ions often provide slightly different selectivity when neutral compounds are analyzed. Silver, for example, is good for cis-trans separation. This is another way of tuning RP selectivity. Instead of screening multiple columns for desirable selectivity one can screen cationic modifiers on a single cation-exchange RP column.

To AllsepTech: I do not see any fundamental problem with packing mixed beds. There is not a lot of high-quality column packing literature around that could help a small company like yours to solve packing problems. To the amatuer, column packing is like voodoo, but it is routine for major column suppliers.
To Zelechonok: while I agree that you get different results from mixing two beds compared to mixed chemistry, I am wondering how much of your technology can be achieved by simply mixing things. "Totally different" (as stated by you) is not necessarily always good.

Uwe, we are not trying to pack two different materials in one column. It is useless exercise. As I said, two columns connected will give you the same result without hassle of packing. The same time there is no problem to pack true MM phase, even literature, as you said, is not available. Primesep routinely produced with 80K plate/meter.
Check any mechanically mixed column from very few suppliers (they are not small company though) and you will find plate count way below of typical RP columns measured by neutral compounds. With charged analytes the efficiency is even lower. Peak symmetry also is not amazing.

To elaborate on the efficency of Primesep columns: A lot of column suppliers are selling so called "5 micron" columns which in real packed with smaller particles (4-4.5). We are using "5 micron" Kromasil which is in real life is 5.8-6.2 microns and still achieve good plate count.
Also we are trying to convince people that efficency is not everything - our statement is "selectivity" (same as in chiral chromatography, nobody ever complained of low plate count in chiral separations)

We don't think that we are " deluting" RP interaction, because every site in our columns is a reverse phase site and ion-exchange.

On "diluting": One has a given amount of room (surface). An optimal column covers almost all of this with a wanted functionality. No room for a second functionalty that is supposed to deliver an optimal 2nd function.

AllsepTech, on your "efficiency" and "selectivity".... That´s what I said above, if you can do with suboptimal.....
Also, if I want to have selectivity in the above examples (between comps of low polarity and ionic ones) I will probably use solvent extraction prior to HPLC and get rid of a lot of goo (etc) at the same time.

Finally, the last few years of some 40 years of professional activity, have been strongly characterized by the confrontation with rediscoverers of the world.

C18 chain is much heavier and bulkier than small ion-exchange group. At 13% of carbon load about 80% of carbons are in pure hydrophobic alkyl chain. The dilution is really insignificant.

I can not see a reproducibility problem from lot to lot. You weigh out the stuff, then you mix it. Unless you can't read the numbers on the balance, the reproducibility should not be problem.
-Uwe

If the above were true, the pharma industry wouldn't have need of granulators, roller compactors, hot-melt extrusion or myriad other technologies. Every solid dosage form could be dry blended and directly compressed ito tablets. I assure you that this is not the case.

It sounds like they've essentially replaced C8/18 with a zwitterionic species bonded to conventional silica phase.

Allsep tech - please send free column to... :)

To Anonymous:

Of course one does not let the mixture sit there and do nothing. Of course it must be mixed. And of course these technologies exist already.

What did I say? You weigh out the stuff, and then you MIX it. Column packing works by making a slurry of the packing in a suitable solvent, which, among other things, requires mixing a solvent with a packing... :-)

To Zelechonok:

As you are aware, mixed columns do exist.

If you are dealing with an isocratic separation, a blending of two packings is not needed, and one can just connect two different column (for example, our old application on explosives). However, if you are running gradient separations, a mixing of the chemistries is needed.

Regarding the reproducibility of packings, and what can beachieved today, I recommend that you look at my publication in J. Chromatogr. 849 (1999) 87-100. You will realize that the reproducibility of a packing can exceed the measurement reproducibility of a test procedure, if the manufacturing procedures are designed properly.

I do not understand your comment on swelling. Why would a silica-based column swell when you put it into an organic solvent? Why are you stressing the non-swelling feature of your packing?

re: your argument with Hans: Unless you invent a phlogiston-type space, the addition of another ligand or a side chain will reduce the coating level of the hydrophobic part of the stationary phase. This may not be a problem, but why deny it?

I never deny. Let me say it again: C18 chain is much heavier and bulkier than small ion-exchange group. At 13% of carbon load about 80% of carbons are in pure hydrophobic alkyl chain. The dilution is really insignificant.

13% carbon is a low carbon content for a C18 on a packing with the surface area of Kromasil. What is the surface coverage in micromoles/squaremeter?

There is no need to have higher carbon load. Half of the commercial RP columns have similar even lower carbon load

Carbon load is not the important part. I am surprised that you do not know the surface coverage of your packing in micromoles/squaremeter.
You must be aware that this number is the value of a bonded phase that is important for selectivity and reproducibility, not the % carbon.

I know the numbers and aware of their importance.

I don’t understand though how it relates to our discussion of mixed-mode chromatography.

I have a little bit different question:
I would like to ask about Luna C18(2) column (Phenomenex). This column is so hydrophopic, that when I analyse some weak acids (pKa about 2-3, e.g. 3,5-dinitrobenzoic acid), they are retarded strongly even without any modifying acid added.
Surprisingly, diminishing pH (by adding HCOOH for example) is diminishing the retention factor!!! Usually it should be in opposite way. At lower pH the analytes should be less ionized and more retarded on the column.
Can you tell me what makes this column so exeptional?

To anon.
Most likely the phenomena is due to an ion-exchange interaction of your acidic analytes with residual basic functional groups on the column. Specially, if the column based on amide chemistry. If I am right, then adding of an acid to the mobile phase will effect ion-exchange in the way you observed. There can be other explanations too.

Well... Luna C18(2) phase is made of pure silica and encapped, so there is really diminished amout of silanols. For sure less than is older era columns (e.g. Lichrosphere type on which diminishing of pH increases the retention of dinitrobenzoic and other acids - as should be expected)

Wow, Zelechonok, you really know your competition. Luna C18(2) with an embedded polar group. The people at Phenomenex are rolling around the floor and hold their bellies loughing hard.

Uwe,
Would you please try to explain the Luna phenomena. Zelechonok gave his explanation which is probably not true for the particular case (Luna)

Ye, Uwe, since you know everything, why don’t you help to the Anon to resolve his question instead of laughing hard with Phenomenex people.

With all the respect Uwe, Luna C18(2) has never been described as with embedded polar groups. I don't understand what is going on... Please explain.

Uwe,
You should know better that Luna has no embedded groups, I thought that by now you would know it.

To Zelechonok and the Anonymous above who asked the question about Luna:

I do not know the answer to the question by Anonymous about Luna C18 (2). Therefore I think it is wise to say nothing, instead of saying something stupid. I'll be willing to discuss the question in private, if the anonymous above wants to contact me.

Yes, I will contact you Uwe, the topic is very interesting.

I wonder if Bharat is satisfied with the replies?

I just wondering if Phenomenex people can explain the phenomena to us. As I understand some people did not like explanation by Zelechonok.

Zelechonok,

Going back to the design of your packings. Since you did not tell us, what the surface coverage is, I became curious and I made some estimates myself. You gave the following information: They are based on Kromasil, which has a specific surface area of about 350 m2/g. Your carbon coverage is 13%. About 80% of the carbon is in the C18 chain, which would put the total number of carbons in a ligand to be around 23 or so. Taking into account that your molecule contains at least the sulfonic acid group, but potentially also a linker, the estimate of the surface coverage results in about 1.7 micromoles/m2. This confirms what I had expected: the surface coverage is much lower than that of a high-quality C18, which is in the order of 3.2-3.5 micromoles/m2 before endcapping.

Uwe,
Can I make an assumption for coverage based on the retention time of toluene, benzoin or benzonitrile (or any other neutral compound). I am comparing a few C18 columns. On one of the C18 column the retention is 70% of rertention for same neutral compound obtained on a Zorbax C18 column. Can I make a statement that coverage is 70% of Zorbax?
Thank you in advance(I am going to borrow you book from the library if you will refer to it in you answer)

No, it is not that easy. Retention gives you at best an idea of the amount of C18 in the column, but you can not use it to guess the surface coverage, unless you have other information on the material.

To Uwe,
People in Allsep were rolling around the floor and hold their bellies laughing hard when they red your conclusion.
Let me explain you this one more time. We are not making good quality reverse phase columns (there are too many of them), we are making good quality mixed-mode columns. For the same reason we do not do any endcapping.

Having taken time out on the weekend it´s surprising to find this is still going. Ok Zelechonok, like AllsepTech you finally recognize the facts, you do not make a good qulity RP column. Now I may require a top quality RP column to do optimum (presently) work. Next time I want to analyze bases, acids, zwitters, triglyzerides, steroids, terpenes, DNA, Proteins, sugars, hydrocarbons all in one run, I might get your mixed mode.

But, the long professional experience has allowed the formulation of the following:

Tenacity of peddler = k/product quality

where k is a near constant depending on ones research style (life philosophy).

Dear HW Mueller,
First of all, we never claimed anywhere that we produce RP –whether good or bad- columns. Ours are classified as "mixed mode".
Second, we never stated that we have C18 phase, just because we do not.
Third, if you define the quality of a column by plate count, symmetry, stability, loadability, and ability to retain a wide range of compounds, then our columns are exceptionally good (plate count is 75-80K on a 6 micron silica gel, symmetry is close to 1, no tailing, and loadability is the same as with many reverse phase columns).
And finally, just wondering how yearly advertising budgets and number of sales persons of major column manufacturers fit your tenacity vs. quality formula. It’s the lack of information that leads to people making wrong choices, and you shouldn’t be making any judgments without even trying to evaluate the product yourself.

AllsepTech, Zelechonok,
This is slowly ceasing to be fun.
I didn´t judge, regarding the formular, I stated extensive experience, if you don´t fit into this than there is no need to be excited.
As I understand it you both corrected earlier statements which suggested that one can or has created a column which is optimal for several separation mechanisms. That being off the table I have no further quips.

Hans

Uh, guys, this thread on the verge of getting out of hand. May I suggest that if you want to continue,you do it via private e-mail, off line.

And now, back to our regular programming.

can any one explain the chemistry of ion pairing agent
what type of reaction increase the retention time of acid or basic group in z-ions

I don't know what you mean by z-ions but the short answer to your question (it's a lot more complex than the following) is that an ion pair reagent induces a reversable ion exchange process through interaction of the ion pair reagent and the stationary phase and an electrostatic interaction between your analyte and the ion pair reagent. You get more retention with increasing ion pair reagent concentration (to a point) and increased retention when your analyte is ionized with a charge opposite that of the ion pair reagent.

is it true:
Acidic compounds have increased retention at pH val- ues below their pKa and basic compounds retain longer at pH values above their pKa
if yes please explain

Looks like a valid question for a good test...

abc

That depends entirely upon your retention mode. In reversed phase chromatography, your statement is true. This is because in reversed phase chromatography, neutral solutes are invariably less polar and more highly retained than the ionized form of the molecule. The guidelines you have stated above will result in solutes being predominantly in the neutral form, thus increasing retention. In ion pair chromatography, it's the opposite, however. This is because, in ion pair chromatography only the ionized form of the molecule contributes to retention (for the most part).