Viscous Heat Generation in UPLC

[unmgvar]

Uwe

from one the articles i have read, the ratio at 2 ml/min looks more 2-3 and around 1 for 0.5 ml/min (Chromatographyonline issue of May 2005, The use of acquity uplc in pharmaceutical development, figure 1 page 16), and the optimal flow range starts at 1ml/min.

my remark regarding column lenght limitation was on the base that out of the 3 parameters of column i.d., particle size and column lenght only the later would be changed. under those conditions there should be a maximum column lenght in theory.

the article actually "focuses more on another problem" (sorry, due to lack of a better term since it is not the articles main subject) of small i.d column which is their small internal volume and the effect to early eluting peaks.
this got me "trap" in contradiction:

the optimal range for the column according to the Van-Deemter plot starts at 1ml/min up, but the greater my flow rate then the earlier my peaks elute and then i have to deal with that broadning effect, then i have to find a second way to retain my peaks, which would be first of all to go and increase the buffer concentration. small i.d. of tubing, very small volumes of system, probably not a good idea in the long run for the system.
second option is too increase column lenght!
so know it looks to me that the smaller the column i.d. then columns must have a minimal optimal lenght and a maximum optimal lenght, in order to prevent peak broadning (eventhou the causes are different at both extremes).

[Uwe Neue]

I do not have the articles here at home, so I do not know what you are talking about. However, the optimum flow rate depends on the analyte, the mobile phase and the temperature. According to my models, for real analytes and real flow rates, the UPLC system does not have any obvious limitations. I will read the article to see where the misunderstanding lies.

The column length limitation is a function of the diffusion coefficient of the analyte. At the given pressure of the UPLC system, it may be at 20 cm for a molecular weight of 100, but it is more along the lines of 40 cm for a molecular weight of 250. For a peptide sample, one is with a 15 cm column completely in the normal comfortable range of applications.

Extra-column bandspreading is an issue for early eluting peaks on a small i.d. column. However, precolumn bandspreading is irrelevant for gradient separations, and most UPLC separations that I have seen have been gradient separations. In addition, there are other ways in which a sample can be focused on column, if there is a need to do so. In all of the applications work that I have seen on 2 mm columns, extra-column bandspreading was not an issue for isocratic separations.

[HW Mueller]

OK, if the heating is negligible in the pump then there is only friction in the column left for causing heating.

I was talking about a flow grade within columns, not an applied flow rate change. The pressure is near atmospheric at the column outlet, the compressed liquid expands throughout its traverse of the column, thus, the flow rate is differnt throughout (slower at the head of the column than at the outlet).

[Uwe Neue]

I do not understand, why you would worry about such things. Even if it were not perfect, you will always be able to calibrate your detector response. If this were not so, GC would not exist...

[unmgvar]

Uwe,

the figure in the article did show that if working with a 2 mm i.d. column the effects were negligible, and the author of the artilce did stress that, yet on the other hand doesn't it mean that smaller particle sizes of silica and smaller i.d. of column are not possible due to that, in the future?
unless some modification are made and also are available to first generation instruments.

[Uwe Neue]

unmgvar

Extra-column contributions are always more of an issue, if one reduces the column size. However, precolumn bandspreading effects can be completely eliminated in gradient elution, or with closely associated injection techniques in isocratic chromatography. Therefore I do not see a fundamental impediment to run 1 mm columns on an UPLC instrument. Bottom line: people are doing this.

[Kostas Petritis]

Unmgvar,

I agree with Uwe. For example in our lab we have used capillary columns with an ID as small as 15 um for packed capillaries and 10 um for monolithic columns. Also in the case of packed capillaries we have gone down to 0.8 um particle size (porous particles) and as high as 20000 psi pressures.

[unmgvar]

Unmgvar,

I agree with Uwe. For example in our lab we have used capillary columns with an ID as small as 15 um for packed capillaries and 10 um for monolithic columns. Also in the case of packed capillaries we have gone down to 0.8 um particle size (porous particles) and as high as 20000 psi pressures.

At which flow rates if i may ask?

what then are the important points to take into consideration when transporting an application from HPLC range to the UPLC range?
after all they are many issues to take into accounts.
as you saw just from reading articles on the matter, it raised many questions. if today i had to do such a transport i would certainly do it wrong and come out with the misbelief conclusion that the investment is not worth the time, money or effort.

[Uwe Neue]

unmgvar:

I am not sure if you are aware of column scaling principles. You can change the particle size and the column length in the same proportion to get the same backpressure per unit analysis time. At the same time, the linear velocity for the maximum performance increases in inverse proportion to the change in the particle size.

There is a roughly 4-fold gap between the pressure limit of an UPLC instrument and the optimum linear velocity for a standard pharmaceutical molecule. If I apply the scaling rule mentioned above to the fullest degree, I can use a 1.25 cm 0.45 micron column on the same system and still get optimal results. Of course, such scaling does not leave you any flexibility. However, before columns with this type of dimensions and particle sizes become available, UPLC will have moved to UUPLC, or even to UUUPLC, or whatever it will be called...

[HW Mueller]

To get back at that flowrate gradient along the column: Since chromatography is still empirical to large extent it is, of course, not absolutely necessary to think about it. But it would be interesting to know whether heating is different in different parts of the column. The other question in the back of my head, about flow rates which could be far from ideal in parts of the column, seem to have been answered by Uwe´s mentioning of column scaling regarding particle size and column length.

[Kostas Petritis]

Unmgvar,

There is nothing really tedious on tranfering methods in UPLC. I guess that the only disadvantages are a premium that has to be paid for the aquitition of a UPLC and the limited availability of stationary phases with <2 um stationary phases (in order to take full advantage of the very high pressure LC).

Uwe, are you implying that there is a UUPLC on it's way?

[Uwe Neue]

Kostas,

That depends what you call "on its way".

It took some 25 years between the time that 10 micron particles became available to the time where you could buy commercial columns with 2.5 micron particles. Considering that I was talking about a similar 4x step in particle size, I suppose that I will have retired by the time we are using sub-0.5 micron particles.

[unmgvar]

Sorry for the delay,
Not had time to log until now.


Uwe,

my question was like for tips on going from analytical to prep, only on the other side. also like you gave in your exemple you didn't take into account that we do not wish to shorten column lenght like you describe because we want to improve performance. otherwise why would Waters bother devoloping a 15 cm 2mm i.d. column, if no to imporve performance?
also i was asking my question for all aspect of the instrument.
i do know that for example the detector has to work at faster sampling rates, which increases noise and has to be taken into account for impurities for example.

tubing effects after the column become even greater as we go down in column i.d.
sample loading also is an issue.
basically like they are rules to move up, i was kind of hopping you would give some place where you could find the rules on how to move down.

Kostas,

i do not have on hands experience on the matter, but i did investigate and ask on the matter, and i did not receive your feedback. on the contrary the opinions were that method transfer was not straight forward at all, and the costs are not only in the initial cost but also in the maintenance of the equipment and the very short time of the consumables, higher cost per injection for sample. true there is a lot less more mobile phase being used, but coluns clogg a lot more faster in pratice.

like you your self say, stationnary phases are the thing in absorption separation.
we were able to show by comparison that it is easier and simpler to develop a nice simple HPTLC method. this was mainly due to the fact that we posses a lot more stationnary phase and with HPTLC you can better use the NP. true for the same phase UPLC is better, but with HPTLC the flexibility of a lot more phases, gives way more separation power in the end.
we got better cost/efficient. faster separation, a lot smaller initial costs, a lot smaller maintenanace costs. more samples for less time.
true you do not get the same level of RSD and precision. but as we looked at it. did we really have the need for it? no, we are far with in regulation demands, but we do have the need for faster simpler, cheaper separation power. and for the price of a regular HPLC for the CAMAG system you get a a PDA scanner, which of course is way cooler

still i believe that something was made wrong with the use of the UPLC, because the method transfer was made too straight forward. i do believe that UPLC should have behaved better. my main question regarding UPLC is to see if it is has "cool" as capillary electrophoresis (which in the end did not "deliver the goods").

also after Uwe's words on the speed of evolution of the system, you do actually convince me to have a more indepth look at the bore/ nano cap range. why wait in the in-between with UPLC?
if to invest then trully for the same money nano cap technology is better more often then not in comparison to UPLC.

[Uwe Neue]

unmgvar,

If you give me sufficient information about an existing separation, I can propose UPLC conditions that meet the separation power of your separation in a shorter time or exceed the separation power that you have in either the same time or a shorter time. No tricks other than playing with column length, particle size and pressure. More specifically, I will do this using 1.7 micron UPLC columns.

What I need from you in order to do this is the complete information on how the classical separation is carried out. I need column length, column diameter, particle size, flow rate, temperature of the separation, mobile phase composition (e.g. 30:40:30 water:methanol:acetonitrile) and the molecular weight range of your analytes. I will then estimate the performance that you get from your column, and I will then estimate, what you could get from UPLC under the same mobiel phase conditions with the same samples. I will specify the run time that you will get for the same performance on an UPLC column, and I will specify the maximum performance that you could get from an UPLC column under the same or faster conditions than you have specified.

If you are running a gradient, I will need the same information as above, plus the mobile phase composition at the beginning and the end of the gradient. If your gradient is not linear, but a bit more complicated, this is a good first guess, which in my experience translates well to more complex gradients.

I am looking forward to a challenge.

(Maybe I should say that these predictions will be based on verifiable science, i.e. fundamental HPLC theory, and are not random nonsense.)

[Kostas Petritis]

Unmgvar,

Assuming that you sent me an e-mail (I think that I got one asking about the waters UPLC), I didn't answer as we do not have a water UPLC in house (at least not yet).

We worked at very high pressures LC way before there is something available so we have several systems (>10) working 7/7 24 h/24h.

Also smaller particle size and smaller ID columns are two different things all together so I do not think that UPLC is an in-between step. The Waters UPLC is a first step towards a commercial solutions that take advantage of what higher pressures can offer (i.e. make possible the use of lower particle size stationary phases which itself provides several advantages).

Time will show how the UPLC will evolve but I wouldn't compare it with capillary electrophoresis. It is the next logical step if you want to use smaller particle size packed columns.