-
- admin
- Site Admin
-
- Posts: 647
- Joined: Wed Aug 11, 2004 12:19 am
By Anonymous on Tuesday, May 25, 2004 - 03:35 am:
I'm using a LiChrospher® 100 RP-18e
column. Mobile phase is 4% ACN/96% water.
What is the best sample to evaluate t0, the column dead time?
-------------------------------------------------------------------------------------------------------
By HW Mueller on Tuesday, May 25, 2004 - 04:27 am:
T2O
-------------------------------------------------------------------------------------------------------
By Anonymous on Tuesday, May 25, 2004 - 04:47 am:
T2O?? I'm not familiar with that name. Can you please write the name of this compound by full?
-------------------------------------------------------------------------------------------------------
By Anonymous on Tuesday, May 25, 2004 - 05:29 am:
Thanx. I already found the name af the compound, but we haven't got it here. Maybe anyone has got other suggestions?
-------------------------------------------------------------------------------------------------------
By Renata on Tuesday, May 25, 2004 - 05:47 am:
I know that the most common is to inject uracil.
Actually, I use a formula to calculate it. And it's the right time to ask if I'm doing it right using the formula or if it's really necessary to perform injections to have the right volume time.
What I do:
Void volume (mL) = d^2 * Pi * L (cm) * Pore Volume /4
Void time (min) = Void volume (mL)/Flow rate (mL/min)
The pore volume for Lichrospher is 0,65
So, considering: Lichrospher RP 18 = 250 x 4 mm and flow rate 1 mL/min
Void volume = 4^2 * 3,14159 * 0,250 * 0,65 /4 = 2,04 mL
Void time = 2,04/1 mL/min = 2,04 min.
Renata
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Tuesday, May 25, 2004 - 03:10 pm:
You can't use the "pore volume" to do the calculation (I assume that you want to use the "specific pore volume" given in the catalogue). However, "0.65" is a good factor for most reversed-phase columns. Note that no manufacturer that I know of gives you the specific pore volume for a bonded phase.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Wednesday, May 26, 2004 - 12:06 am:
Dolan, Snyder, Troubleshooting LC Systems...., give:
to ~ 0.5Ld^2
where L = column length
d = column diameter
this is a rough approximation for porous, spherical particle silica columns at a flow rate of 1mL/min.
Sorry that I didn´t list T2O as tritiated water, or di-tritium oxide.
-------------------------------------------------------------------------------------------------------
By Anonymous on Wednesday, May 26, 2004 - 02:37 am:
Why not use any organic solvent to calculate T0 like acetone , methanol etc.??????
-------------------------------------------------------------------------------------------------------
By Tim on Wednesday, May 26, 2004 - 07:19 am:
You need a compond which has a low affinity for the stationary phase of the column, but will give a UV absorption. T20 and Uracil are generally considered good compounds for this. Acetone is retained by a reverse phase column.
You can also do an injection of say your blank solution and look for the initial deflection of the baseline, which should be around the theoretical value (but need to account for additional dead volume that the tubing in the HPLC system gives). Some analyses don't give a very clear deflection, so could try altering the wavelength to something lower which will increase the deflection you see.
-------------------------------------------------------------------------------------------------------
By Einar Pontén - SeQuant AB on Wednesday, May 26, 2004 - 02:55 pm:
One may speculate that with such high content of water you may find a smaller to (lower volume) for the void than when the RP phase is wetted by a high content of solvent.
Thus, the mobile phase is excluded from the pores and the accessable volume for uracil decreases. For a method with a fixed mobile phase the operational approach is the most practical and straightforward procedure, but to is not a fixed value for the column.
-------------------------------------------------------------------------------------------------------
By A.Mouse on Wednesday, May 26, 2004 - 03:04 pm:
HW:
My old-fashioned calculator says that
0.65*pi/4 ~ 0.5
Your estimate is identical to the one by Renata.
-------------------------------------------------------------------------------------------------------
By greg on Wednesday, May 26, 2004 - 10:42 pm:
use of a formula to calculate the void volume does seem adequate but does this approach take into account the actual tubing associated with the fluid path?? If not, then the calculated void will be (i imagine) slightly lower than the "true" void.
As mentioned, for C18, uracil is a common cpd for void volume determination of the entire chromatographic system as opposed to the column alone.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Thursday, May 27, 2004 - 12:19 am:
Tim,
initial deflections can sometimes be due to a pressure shock (or whatever you want to call this) that may be considerably ahead of to.
Einar,
one reason to pay attention to to (arghh, tm would be better) is to rule out that you are developing a method under dewetting conditions. But it appears that many modern columns seldomly do this, anyway.
A.Mouse,
probably the two estimations are based on the same assumptions.
Greg and Tim,
it might be wise to drastically reduce your extra-precolumn volume if it increases to measurably?
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Saturday, May 29, 2004 - 11:00 am:
I know that many people use the "initial deflection" as a measure of V0. Since this is not defined at all, I am strongly opposed to this. If one needs to calculate a retention factor, a well defined V0 marker needs to be used. Otherwise you get retention factors that nobody can duplicate. If you want to write a publication or a procedure that is used in another department, and you need to calculate retention factors, you must measure V0 properly.
-------------------------------------------------------------------------------------------------------
By tom jupille on Sunday, May 30, 2004 - 04:47 pm:
When we teach our method development coursess, this whole topic comes up frequently; it is quite a "can of worms".
We generally advise people to use the initial "refractive index anomaly" or "schlieren peak" or "solvent front" (no standard terminology of this!) that most UV detectors will show when the composition change caused by injection gets through the column -- but observe some "common sense" caveats:
- estimate the dwell volume (using the 0.5 x L x dc^2 approximation -- which is good to +/- 10-15% or so) as a reality check; that usually will prevent misidentification of a pressure pulse as t0.
- if the sample is dissolved in the mobile phase and/or if the detector is running on a non-sensitive scale (e.g., for quantitating a major component), a "spectator" compound (UV-absorbing but non-retained) such as uracil, as suggested by Renata, may be a good idea.
The use of a spectator compound opens up its own questions, of course, in that different compounds will have different t0, depending on how much of the pore structure they can penetrate (this is how size exclusion or ion exclusion work -- each analyte elutes at its characteristic t0). Mueller's suggestion in the second post is probably the "most correct" (D2O also works, by the way), but even that opens up questions on close examination if you accept the model that reversed-phase involves the formation of an organic-solvent-enriched layer at the surface of the stationary phase.
From a practical standpoint, of course, the above paragraph is akin to medieval theologians pondering the question of how many angels can dance on the head of a pin. Consistency is more important than accuracy in this case.
-------------------------------------------------------------------------------------------------------
By Einar Pontén - SeQuant AB on Tuesday, June 1, 2004 - 03:12 pm:
Nobody mention nitrate as a to-marker when using a UV detector? In fact, uracil has retention on some hydrophilic RP phases.
Nitrate should have no retention and it exhibits good UV absorbtion. That assure a clear "deflection" for a wide range of RP phases and sensitivity settings. Is anyone aware of a reason for not using nitrate?
By the way, the simplest way to determine the extra-column void volume should be to replace the column with a "zero-volume" union before the injection of any absorbing specie, e.g., nitrate.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Tuesday, June 1, 2004 - 04:08 pm:
If one is doing fundamental work on retention mechanisms, column characterization, mobile phase selectivity and related things, one must use a well defined marker for the measurement of the column dead volume. For the work common in the practice of chromatography, the definition of a retention time or a retention time window is perfectly adequate. For getting an impression on the retention factor of a compound, one can use the calculation rules mentioned above, but an estimate is not good enough for determining a retention factor. What is not acceptable at all, is to define a retention time by an undefined number such as the "initial deflection". Such a value is not acceptable in science, nor should it be acceptable in any other environment.
Also, Einar is correct that for solid scientific work, the extra-column contribution to V0 should be subtracted. However, I have seen only few publications that pay attention to this.
-------------------------------------------------------------------------------------------------------
By Anonymous on Wednesday, June 2, 2004 - 05:38 am:
if we're calculating k' do we also have to correct the observed analyte retention time for the extracolumn contribution?
-------------------------------------------------------------------------------------------------------
By HW Mueller on Wednesday, June 2, 2004 - 07:32 am:
Well, decide for yourself, for instance, if you connect the inj. valve to a 4.6x250mm column with a 0.1x50mm tube you may have a Vo ~ 3000µL, whereas the extracolumn volume ~ 4µL.
-------------------------------------------------------------------------------------------------------
By Einar Pontén - SeQuant AB on Wednesday, June 2, 2004 - 01:10 pm:
It is rarely the case that the extra column volume (Vex) is only 0.1% of Vo. The injection volume itself may be >1%.
On the contrary, our experience is that too large Vex sometimes becomes a problem for unexperienced users since, as you know, it ruins the effiency, especially for columns with smaller particles and small ID's (relative effect).
Still, it does not make sense to correct for Vex in calculation of k' (time spent on the stationary phase/time spent in mobile phase), although k' will decrease measurable for large extra volumes and small columns.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Wednesday, June 2, 2004 - 03:44 pm:
On a standard system, the extra-column volume is about 70 to 80 microL. As Einar has pointed out, this is not a problem if you are running a 4.6 x 150 mm column. The influence of the extra-column volume on the determination of the retention factor is about 5%. This error is too small for most people to worry about. However, if I run a 2.1 x 150 mm column, the error in the determination of the V0 is now about 25%. That is a big deal isn't it? I can see the department director sitting on his desk wondering why the reported retention factors are so much different on the 2.1 mm column...
Well, I suppose it's these lousy column manufacturers that don't know how to pack their columns properly...
-------------------------------------------------------------------------------------------------------
By syx_gf on Wednesday, June 23, 2004 - 05:24 pm:
If we use uracil or other available marker, how could we determine the dead time, using retention time or start point of the peak?
-------------------------------------------------------------------------------------------------------
By tom jupille on Friday, June 25, 2004 - 03:49 pm:
The retention time.
-------------------------------------------------------------------------------------------------------
By Jimmy on Friday, August 6, 2004 - 07:14 pm:
To Einar and Uwe,
Is the sample loop volume part of the extra column volumn? If yes, then the extra-column volume wouldn't be 80 uL on a standard system.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Saturday, August 7, 2004 - 06:49 am:
For practical HPLC, it is not very exciting, but one needs to be aware of this extra-column volume. For thorough theoretical work, the extra-column volume should be subtracted (as it is done in my recent publications)
-------------------------------------------------------------------------------------------------------
By tom jupille on Monday, August 9, 2004 - 10:52 am:
To Jimmy: if the injector is properly configured, on ly the volume actually injected counts as part of the extra-column volume. In a properly-configured injector, the flow direction in the loop during the injection process is opposite that during the loading process (think of this as "last in, first out") so that the sample does not traverse the entire length (volume) of the loop.
I'm using a LiChrospher® 100 RP-18e
column. Mobile phase is 4% ACN/96% water.
What is the best sample to evaluate t0, the column dead time?
-------------------------------------------------------------------------------------------------------
By HW Mueller on Tuesday, May 25, 2004 - 04:27 am:
T2O
-------------------------------------------------------------------------------------------------------
By Anonymous on Tuesday, May 25, 2004 - 04:47 am:
T2O?? I'm not familiar with that name. Can you please write the name of this compound by full?
-------------------------------------------------------------------------------------------------------
By Anonymous on Tuesday, May 25, 2004 - 05:29 am:
Thanx. I already found the name af the compound, but we haven't got it here. Maybe anyone has got other suggestions?
-------------------------------------------------------------------------------------------------------
By Renata on Tuesday, May 25, 2004 - 05:47 am:
I know that the most common is to inject uracil.
Actually, I use a formula to calculate it. And it's the right time to ask if I'm doing it right using the formula or if it's really necessary to perform injections to have the right volume time.
What I do:
Void volume (mL) = d^2 * Pi * L (cm) * Pore Volume /4
Void time (min) = Void volume (mL)/Flow rate (mL/min)
The pore volume for Lichrospher is 0,65
So, considering: Lichrospher RP 18 = 250 x 4 mm and flow rate 1 mL/min
Void volume = 4^2 * 3,14159 * 0,250 * 0,65 /4 = 2,04 mL
Void time = 2,04/1 mL/min = 2,04 min.
Renata
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Tuesday, May 25, 2004 - 03:10 pm:
You can't use the "pore volume" to do the calculation (I assume that you want to use the "specific pore volume" given in the catalogue). However, "0.65" is a good factor for most reversed-phase columns. Note that no manufacturer that I know of gives you the specific pore volume for a bonded phase.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Wednesday, May 26, 2004 - 12:06 am:
Dolan, Snyder, Troubleshooting LC Systems...., give:
to ~ 0.5Ld^2
where L = column length
d = column diameter
this is a rough approximation for porous, spherical particle silica columns at a flow rate of 1mL/min.
Sorry that I didn´t list T2O as tritiated water, or di-tritium oxide.
-------------------------------------------------------------------------------------------------------
By Anonymous on Wednesday, May 26, 2004 - 02:37 am:
Why not use any organic solvent to calculate T0 like acetone , methanol etc.??????
-------------------------------------------------------------------------------------------------------
By Tim on Wednesday, May 26, 2004 - 07:19 am:
You need a compond which has a low affinity for the stationary phase of the column, but will give a UV absorption. T20 and Uracil are generally considered good compounds for this. Acetone is retained by a reverse phase column.
You can also do an injection of say your blank solution and look for the initial deflection of the baseline, which should be around the theoretical value (but need to account for additional dead volume that the tubing in the HPLC system gives). Some analyses don't give a very clear deflection, so could try altering the wavelength to something lower which will increase the deflection you see.
-------------------------------------------------------------------------------------------------------
By Einar Pontén - SeQuant AB on Wednesday, May 26, 2004 - 02:55 pm:
One may speculate that with such high content of water you may find a smaller to (lower volume) for the void than when the RP phase is wetted by a high content of solvent.
Thus, the mobile phase is excluded from the pores and the accessable volume for uracil decreases. For a method with a fixed mobile phase the operational approach is the most practical and straightforward procedure, but to is not a fixed value for the column.
-------------------------------------------------------------------------------------------------------
By A.Mouse on Wednesday, May 26, 2004 - 03:04 pm:
HW:
My old-fashioned calculator says that
0.65*pi/4 ~ 0.5
Your estimate is identical to the one by Renata.
-------------------------------------------------------------------------------------------------------
By greg on Wednesday, May 26, 2004 - 10:42 pm:
use of a formula to calculate the void volume does seem adequate but does this approach take into account the actual tubing associated with the fluid path?? If not, then the calculated void will be (i imagine) slightly lower than the "true" void.
As mentioned, for C18, uracil is a common cpd for void volume determination of the entire chromatographic system as opposed to the column alone.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Thursday, May 27, 2004 - 12:19 am:
Tim,
initial deflections can sometimes be due to a pressure shock (or whatever you want to call this) that may be considerably ahead of to.
Einar,
one reason to pay attention to to (arghh, tm would be better) is to rule out that you are developing a method under dewetting conditions. But it appears that many modern columns seldomly do this, anyway.
A.Mouse,
probably the two estimations are based on the same assumptions.
Greg and Tim,
it might be wise to drastically reduce your extra-precolumn volume if it increases to measurably?
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Saturday, May 29, 2004 - 11:00 am:
I know that many people use the "initial deflection" as a measure of V0. Since this is not defined at all, I am strongly opposed to this. If one needs to calculate a retention factor, a well defined V0 marker needs to be used. Otherwise you get retention factors that nobody can duplicate. If you want to write a publication or a procedure that is used in another department, and you need to calculate retention factors, you must measure V0 properly.
-------------------------------------------------------------------------------------------------------
By tom jupille on Sunday, May 30, 2004 - 04:47 pm:
When we teach our method development coursess, this whole topic comes up frequently; it is quite a "can of worms".
We generally advise people to use the initial "refractive index anomaly" or "schlieren peak" or "solvent front" (no standard terminology of this!) that most UV detectors will show when the composition change caused by injection gets through the column -- but observe some "common sense" caveats:
- estimate the dwell volume (using the 0.5 x L x dc^2 approximation -- which is good to +/- 10-15% or so) as a reality check; that usually will prevent misidentification of a pressure pulse as t0.
- if the sample is dissolved in the mobile phase and/or if the detector is running on a non-sensitive scale (e.g., for quantitating a major component), a "spectator" compound (UV-absorbing but non-retained) such as uracil, as suggested by Renata, may be a good idea.
The use of a spectator compound opens up its own questions, of course, in that different compounds will have different t0, depending on how much of the pore structure they can penetrate (this is how size exclusion or ion exclusion work -- each analyte elutes at its characteristic t0). Mueller's suggestion in the second post is probably the "most correct" (D2O also works, by the way), but even that opens up questions on close examination if you accept the model that reversed-phase involves the formation of an organic-solvent-enriched layer at the surface of the stationary phase.
From a practical standpoint, of course, the above paragraph is akin to medieval theologians pondering the question of how many angels can dance on the head of a pin. Consistency is more important than accuracy in this case.
-------------------------------------------------------------------------------------------------------
By Einar Pontén - SeQuant AB on Tuesday, June 1, 2004 - 03:12 pm:
Nobody mention nitrate as a to-marker when using a UV detector? In fact, uracil has retention on some hydrophilic RP phases.
Nitrate should have no retention and it exhibits good UV absorbtion. That assure a clear "deflection" for a wide range of RP phases and sensitivity settings. Is anyone aware of a reason for not using nitrate?
By the way, the simplest way to determine the extra-column void volume should be to replace the column with a "zero-volume" union before the injection of any absorbing specie, e.g., nitrate.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Tuesday, June 1, 2004 - 04:08 pm:
If one is doing fundamental work on retention mechanisms, column characterization, mobile phase selectivity and related things, one must use a well defined marker for the measurement of the column dead volume. For the work common in the practice of chromatography, the definition of a retention time or a retention time window is perfectly adequate. For getting an impression on the retention factor of a compound, one can use the calculation rules mentioned above, but an estimate is not good enough for determining a retention factor. What is not acceptable at all, is to define a retention time by an undefined number such as the "initial deflection". Such a value is not acceptable in science, nor should it be acceptable in any other environment.
Also, Einar is correct that for solid scientific work, the extra-column contribution to V0 should be subtracted. However, I have seen only few publications that pay attention to this.
-------------------------------------------------------------------------------------------------------
By Anonymous on Wednesday, June 2, 2004 - 05:38 am:
if we're calculating k' do we also have to correct the observed analyte retention time for the extracolumn contribution?
-------------------------------------------------------------------------------------------------------
By HW Mueller on Wednesday, June 2, 2004 - 07:32 am:
Well, decide for yourself, for instance, if you connect the inj. valve to a 4.6x250mm column with a 0.1x50mm tube you may have a Vo ~ 3000µL, whereas the extracolumn volume ~ 4µL.
-------------------------------------------------------------------------------------------------------
By Einar Pontén - SeQuant AB on Wednesday, June 2, 2004 - 01:10 pm:
It is rarely the case that the extra column volume (Vex) is only 0.1% of Vo. The injection volume itself may be >1%.
On the contrary, our experience is that too large Vex sometimes becomes a problem for unexperienced users since, as you know, it ruins the effiency, especially for columns with smaller particles and small ID's (relative effect).
Still, it does not make sense to correct for Vex in calculation of k' (time spent on the stationary phase/time spent in mobile phase), although k' will decrease measurable for large extra volumes and small columns.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Wednesday, June 2, 2004 - 03:44 pm:
On a standard system, the extra-column volume is about 70 to 80 microL. As Einar has pointed out, this is not a problem if you are running a 4.6 x 150 mm column. The influence of the extra-column volume on the determination of the retention factor is about 5%. This error is too small for most people to worry about. However, if I run a 2.1 x 150 mm column, the error in the determination of the V0 is now about 25%. That is a big deal isn't it? I can see the department director sitting on his desk wondering why the reported retention factors are so much different on the 2.1 mm column...
Well, I suppose it's these lousy column manufacturers that don't know how to pack their columns properly...
-------------------------------------------------------------------------------------------------------
By syx_gf on Wednesday, June 23, 2004 - 05:24 pm:
If we use uracil or other available marker, how could we determine the dead time, using retention time or start point of the peak?
-------------------------------------------------------------------------------------------------------
By tom jupille on Friday, June 25, 2004 - 03:49 pm:
The retention time.
-------------------------------------------------------------------------------------------------------
By Jimmy on Friday, August 6, 2004 - 07:14 pm:
To Einar and Uwe,
Is the sample loop volume part of the extra column volumn? If yes, then the extra-column volume wouldn't be 80 uL on a standard system.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Saturday, August 7, 2004 - 06:49 am:
For practical HPLC, it is not very exciting, but one needs to be aware of this extra-column volume. For thorough theoretical work, the extra-column volume should be subtracted (as it is done in my recent publications)
-------------------------------------------------------------------------------------------------------
By tom jupille on Monday, August 9, 2004 - 10:52 am:
To Jimmy: if the injector is properly configured, on ly the volume actually injected counts as part of the extra-column volume. In a properly-configured injector, the flow direction in the loop during the injection process is opposite that during the loading process (think of this as "last in, first out") so that the sample does not traverse the entire length (volume) of the loop.
