C8 vs C18 columns

[adam]

Here's a question.

We all know that C8 columns give somewhat less retention than C18. And that C8 columns also show less tailing because the analytes have more 'equal access' to the active sites: vs the situation with C18 where only a fraction of the molecules can interact with the active sites.

But what about differences in stationary mass transfer. It would seem that this should be less with C8 columns. But in all the reading I have done over the years - quite a bit - I've never heard this claim made.

Does anyone have any knowledge to share about this point.

Much Thanks
Adam

[dave.cheng]

Here's a question.

We all know that C8 columns give somewhat less retention than C18. And that C8 columns also show less tailing because the analytes have more 'equal access' to the active sites: vs the situation with C18 where only a fraction of the molecules can interact with the active sites.

But what about differences in stationary mass transfer. It would seem that this should be less with C8 columns. But in all the reading I have done over the years - quite a bit - I've never heard this claim made.

Does anyone have any knowledge to share about this point.

Much Thanks
Adam

Hello,

I think it is right that C8 show less tailing than C18. but I think the cause is the retention on C8 is shorter than C18, the analytes and interface on C8 stationary phase interaction sites less than on C18, not more than C18.
Good Luck!

[philippem]

Hi Adam,

During my PhD I did a lot of work on C8 and C18 materials in respect to retention behaviour and tailing.

We prepared different lots of C8 and C18 (silica based )material by varying the nature of the siloxane to which the C8 or C18 chain was coupled . Therefor we could bonded different amounts of C8 or C18 to the silica.


Tailing has two aspects: (residual ) active sites (Si-OH, acidic sites) and the column packing procedure.

If you are looking at pure reversed phase mechanism one can deactivate these active sites using the proper chemistry, so tailing can be avoided.

Proper column packing is also important in order to achieve high efficient column and symmetrical peaks. In our research results we did not saw difference using C8 or C18 if the proper packing procedures are used.

Mass transfer is indeed important, but look at the "mass" : The more material bound onto the silica the higher the retention, so a high loaded C8 could generate more retention than a low loaded C18.

Hope this info helps

regards

Philippe

[Wayne Way]

FYI


C18 does not always have greater retention than C8.

[tom jupille]

I had a chance to look at the retention data that was used for the PQRI column selectivity database generated by Snyder and Dolan. There were something like 180 C18 and 80 C8 columns (this is from memory, so the numbers are approximate). I couldn't find any statistically significant difference between C8 and C18 columns. On average the C8s were less retentive than the C18s, but there was so much variation within those classes that the difference was insignificant (I think the difference in the means was around 1 sigma).

Conclusion: variations in bonding chemistry and packing quality have much more of an effect than C8 vs C18.

[Alfred88]

Dear all:
As I remember, both C8 and C18 columns work by hydrophobic interactions (you can read: no interaction). Tom stated that, "On average the C8s were less retentive than the C18s, but there was so much variation within those classes that the difference was insignificant." Nonetheless, I think that the difference could be significant.

Here is a "thought experiment": I put two identical, spherical balls on two almost identical sheep (except one has longer hair). The ball on the sheep with longer hair should stay much longer (before falling to the ground), because the longer hair would make the ball to move slower, so the ball will fall off later.

A C18 column is similar to the sheep with longer hair, so it should give a higher retention.

Alfred

[JA]

Here is a "thought experiment": I put two identical, spherical balls on two almost identical sheep (except one has longer hair). The ball on the sheep with longer hair should stay much longer (before falling to the ground), because the longer hair would make the ball to move slower, so the ball will fall off later.

A C18 column is similar to the sheep with longer hair, so it should give a higher retention.

What if the sheep with longer hair is actually balding, or in other words, that the hair is less dense?

Quite a few questions now;
Is it fair to summise that C8 bonding, on average, affords higher bonding densities?

Do the alkyl chains point 'straight up'? In which case, shouldn't the maximum bonding densities be roughly similar? Alternatively, are the substitutued siloxanes sufficiently close together to actually intertwine, and if so, do they?

Do manufacturers use an excess of silating reagent in creating the C8/C18 phase? What then determines the bonding density, steric effects alone, or physicochemical properties like the acidity of the silica (residual metals and impurities, etc)?

Thanks

[adam]

Interesting comments. But it's kind of getting away from my initial question.

"Is there a difference in stationary phase mass transfer, and hence, efficiency; with C8 vs C18"

Thanks

[Uwe Neue]

The answer to all question is "Yes", but sometimes also "No".

There are many different options, and the results will vary with the options. For example, with Kirkland's sterically hindered silanes, the amount that you can put on the surface depends on the steric hindrance around the base of the silane, and the coating level would in this case be independent of the chain length. The same is true for the case of hybrid packings, where only part of the surface is covered with silanols. There the amount of silanols on the surface determines how much you can bond, and the surface coverage of a C18 and a C8 is identical.

With respect to steric hindrance, you should look at a pore as the inside surface of a ball. If the pore size is small enough, the tips of your C18 chains will touch earlier than the tips of your C8 chains, and you get a higher coverage for a C8. On the other hand, if the ball is large enough then this does not play a role any more, than the surface coverage of C8 and C18 is identical, even when you go for maximum coverage. The last thought gets me to another point: everything said above applies to a situation, where you try to achieve maximum coverage. There are packings around (e.g. the Atlantis T3), where the surface coverage with the primary C18 silane is kept low on purpose, for example to achieve a full wettability in water. In this case, silanols are removed through the endcapping, which is anyway the common second step.

The other element that plays a role in retention is the ratio of surface area to pore volume. Most of us are aware that a C18 on a large pore size gives less retention than a C18 on a small pore size. But the pore volume itself can be varied, and a packing with a small pore volume gives much more retention than the %C indicates. %C is not a valid measure for retention.

So, some packings are made like sheep with a full head of hair, and other packings are made like bolding sheep, and then you have little lambs as well.

[tom jupille]

Tom stated that, "On average the C8s were less retentive than the C18s, but there was so much variation within those classes that the difference was insignificant." Nonetheless, I think that the difference could be significant.

For a long time I thought the same thing. However the fact is that looking at a sample of 180-odd C18 and 80-odd C8 columns, all run under the same conditions with the same sample, there was more variability within each category that there was difference between them. In other words, the difference in retention was not statistically significant. All the thought experiments in the world don't change that.

Uwe has given a much more detailed explanation, but the bottom line is that C8 vs C18 is only one one of the parameters involved.

[adam]

Thank you.