ACN/MeOH mixtures [April 21, 2004]
Posted: Wed Aug 25, 2004 4:43 am
By Luca Corte on Wednesday, April 21, 2004 - 08:12 am:
Dear folks,
On developing HPLC methods for some Penicillin mixtures (mainly, Pen. G Procaine/Benzathine and Cloxacillin Benzathine), I found out that working at pH 3.5, phosphate buffer 0.04 M, with ACN/MeOH 3/2 as organic modifier, gives me very good results in terms of separation and peak simmetry. For some products, IPRs may be needed; anyway, the problem does not lie in the usage of IPRs, but in the fact that, sequence after sequence, I can notice a steady decrease in peak symmetry. I have "blamed" it on the MeOH - acid medium, which can cause deterioration of the column (I use either an Eclipse XDC C18 or a Purospher Star C18E), but I am wondering why in the European Pharmacopeia or in application notes such as Waters', acid buffers with MeOH as eluents are in use (to be precise, they use a trinary phase made up of water, acid P buffer of variable molarity from - 0.01 to 0.01 - and Methanol).
Shouldn't deterioration happen all the same or is this phenomenon enhaced by the mixture ACN/MeOH?
I know it can sound as an entangled issue, I hope there's someone out there who can give me a bit of advice, since I have not been able to find any reference on it.
Thanks
Luca
-------------------------------------------------------------------------------------------------------
By Anonymous on Wednesday, April 21, 2004 - 11:00 am:
And I would have a followup question to this: which is more damaging to an RP column, acid + water, or acid + methanol?
-------------------------------------------------------------------------------------------------------
By bill tindall on Wednesday, April 21, 2004 - 06:12 pm:
If this solvent mix is less than about 50% organic, then one can get a reliable measure of hydrogen ion activity by simply meausring the pH with a pH electrode calibrated in water. It is kind of complicated why this is so , but various factors cancel out under these condtions which enable a measurement to be made which is easy to interpret.
So, make up the solutions of interest and measure their pH and report back the results. Presuming there is an acid catalyzed column degredation reaction, then the pH measurement will tell you which solution is more acidic and hence faster rate. I am curious about the results you find.
I have done many separations in pH 2.2 phosphoric acid and methano/water for aromatic acids and columns last for years.
The answer to the 2nd question is more complicated and depends on the nature of the acid and whether we are talking about pure water vs pure methanol or some mixture. Perchloric acid in methanol is more acidic than perchloric acid in water, becasue methanol is a weaker base than water(hence protonated methanol is a stronger acid). But, things like dihydrogen phosphate are stronger bases in methanol than water, so phosphate buffers are more basic(or less acidic) in methanol than in water.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Thursday, April 22, 2004 - 12:20 am:
It would also be interesting to know whether extensive washing restores symmetry. My guess is that it should, as I also think that a pH of 3.5 would not attack your column noticeably.
On predicting reactivity, see also Bill´s contribution in the message board "pH/pKas". The dielectric property can dramatically influence the activation state of a reaction (an ionic activsation state is at a much lower energy in a high dielectric....). That means the kinetics of a reaction are also important, not only the thermodynamic states. But this complicated matter needs not concern you, your problem should lie elsewhere.
-------------------------------------------------------------------------------------------------------
By Luca Corte on Thursday, April 22, 2004 - 06:18 am:
Thanks Bill and thanks HW!
First time I need to reuse my P Buffer pH 3.5 ACN/MeOH 3/2 eluent, whose volume ratio is 65 Buffer : 35 ACN/MeOH (3/2), I will measure the pH of the final mixture and see what the real value is.
Knowing that with acidic/methanolic mixtures columns last for a long time relievs me. I am wondering if ACN does play a role in my problem.
I have to add that, in order not to leave the column in MeOH/ACN, when the sequence is over I wash it with H2O/ACN to remove the buffer, then with pure ACN. I know that it's safer to store your columns in pure ACN; moreover, analyses of products involving the use of ACN/MeOH are not done on routine in my Lab, and since I have dedicated columns for that, I prefer storing them in ACN.
The loss in symmetry is not dramatic from one sequence set to another, but you can see that if you plot a trend of the analyte TF over time.
In addition to that, if I use those columns to perform a chromatography of a product for which only ACN is needed as organic modifier, I can see a temporary slight loss in symmetry in it too. That one tends to revert over time, but I have not many data on this specific aspect.
I will post follows up after getting pH measurements.
By the way, I have been told that you can use MeOH to wash your columns when they look to be "spent" in some way, so that you can actually regenerate them, to some extent, but I have also been warned this procedure might be dangerous if done too often. Do you have any information on that? So, why shipping new columns in water/methanol instead of pure ACN?
Thanks
Luca
-------------------------------------------------------------------------------------------------------
By RH on Thursday, April 22, 2004 - 07:17 am:
I used mixtures of MeOH/ACN in buffer with different C18 phases and didn`t see adverse effects. One important difference between the two solvents, besides other physicochemical properties, is the protic nature of methanol.
As your peak symmetry changes slightly with analysis time this could be a problem of accumulation of matrix impurities.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Thursday, April 22, 2004 - 03:25 pm:
I agree with RH, that the most likely reason for the chage in peak shape are the components of the sample.
Bill, the pH of a phosphate buffer of 3.5 in water is going to drop, when you add organic solvent. In addition, how is going to help you, if the only information on column stability that you can get is the pH stability with the pH measured in water?
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Thursday, April 22, 2004 - 03:28 pm:
Geee -- the pH of the phosphate is going to INCREASE when you add an organic solvent ....
See Bill, what you made me do...
-------------------------------------------------------------------------------------------------------
By HW Mueller on Friday, April 23, 2004 - 01:29 am:
Luca, where do you measure Symmetry (Asymm...) at baseline or further up? Maybe you are overinterpreting?
Some manufacturers are recommending the more inert ACN for storage, because they apparently think in years... you can wash all you want with MeOH. Maybe there are some exceptions? I have not seen any. This high reactivity of MeOH toward C-18, etc., phases is apparently surviving from very old days??
-------------------------------------------------------------------------------------------------------
By Luca Corte on Friday, April 23, 2004 - 02:16 am:
Hi again!
I have seen that some of you are guessing about accumulation of matrix impurities on the column. I think you mean impurities from the sample... Well, I have very simple products to analyze which totally dissolve in the mobile phase, or, better, in the corresponding water/orgainc solvent mixture. Sometimes, in water alone. Moreover, I always filter samples through a 0.22 um filter prior to injection. Solvents and Milli-Q grade water are 0.22 um filtered as well.
So, according to me, therre shouldn't be any matter accumulation on the column. I should also notice pressure increase, which I don't.
Symmetry and TF are measured by means of the HPLC 1100 software (extended perfomance report). I look at the USP tailing and Symmetry, as you can see in the report format. They are complementary indexes, so that - ideally - if symmetry is 0.95, for instance, TF is 1.05. This way I can have info on column overloading and pH effects on the quantity of injected sample. Picky as I am, I am always striving to get ideal symmetry...
I must say that I am relieved to hear there's someone else out there using ACN/MeOH mxitures...
Luca
-------------------------------------------------------------------------------------------------------
By Alex on Friday, April 23, 2004 - 06:16 am:
Hi Luca,
We are using also MeOH/MeCN mixtures frequently. With water and no buffer. There is no reason that the mixture should be more agfressive then the solvents alone.
To your last posting: Do you isocratic elution? If so, more hydrophobic substances may stick to the column, even if the sample solution looks clean, is filtered.... Try to included a cleaning step with 100% organic from time to time.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Friday, April 23, 2004 - 08:00 am:
Ok, so you use the default position for measuring Symmetry, which is probably above the baseline. If you let it be determined at baseline you will have problems (the start and end of a peak is somewhat difficult to find, discussed before).
Why don´t you tell us what your matrix comes from? It is dissolved material which will change your column characteristics.
-------------------------------------------------------------------------------------------------------
By A.Mouse on Friday, April 23, 2004 - 07:30 pm:
Contamination of the column and pressure increase do not always correlate.
-------------------------------------------------------------------------------------------------------
By Luca Corte on Monday, April 26, 2004 - 01:22 am:
Hi again!
First off, a couple of questions to Rolf and Alexander, who have been using ACN/MeOH mixtures.
When you are done with your analyses, do you store the column in ACN/MeOH alone? As I said before, I used to store the column in pure ACN, thinking that long-term storage in a MeOH mixture could possibly damage the hydrophobically-modified silica layer, but after reading your comments, it may not be the case. It could be that, removing ACN from the column and using the ACN/MeOH mixture upon ensuing sequences, may lead to a small and temporary loss in symmetry because of variation in the eluents chemical-physical properties. In short, the column may need longer equilibration times.
Anyway, I used to do final washing in ACN alone as I also wanted to get rid of Methanol in pump and injector assembly. I have just a single HPLC instrument and not all of my analyses require ACN/MeOH mixtures, most of them are done with Buffer/ACN. I wanted to have the machine ready to operate in the "standard" conditions, so to speak, and I was also afraid that MeOH, in contact with steel and plastic parts, could be more aggressive than ACN.
That leads me to column contamination... It could also be that small debris, like metal particles could come off from the tubing, or pump, and contaminate the column without affecting the internal pressure. This effect could me more visible with the Purospher Star column - which is said to be metal-ion free and capable of giving more straight peaks. So, if metal contamination occurs, it should be more evident than that observed with a "standard" column.
Maybe a precolumn would do the trick.
To answer HW, my sample matrix is made up of the sample alone, that is: pure Penicillin Benzathine, or Penicillin Procaine, or Cloxacillin Benzathine. I just analyze raw materials, without excipients of any sort.
To answer Alex, every time a sequence is over, I carry out a final washing with pure organic solvent.
Thanks again to you all, you have been very helpful! More comments are welcomed!
Luca
-------------------------------------------------------------------------------------------------------
By HW Mueller on Monday, April 26, 2004 - 03:17 am:
Pure is a relative term. "Normally" pure samples and mobile phases should give hundreds of injections without significant symmetry trends.
I don´t use piston pumps, but would expect the seals to be happier with MeOH than with ACN.
To expound on A.Mouse and my former statements: Generally, particulate matter does not change surface sharacteristics, they may cause plugging, whereas materials that change surface characteristics seldomly cause plugging (in my experience, only proteins have done both).
Thus, it seems either you don´t know something about the materials you use (or the software, like where the Sym is calculated) or you are overinterpreting.
-------------------------------------------------------------------------------------------------------
By Luca Corte on Monday, April 26, 2004 - 08:08 am:
Well, I think I have a sound knowledge of the materials I analyze, since they come from customers who have already done their analyses (so I can compare notes on results); they are products of synthesis and recrystallizazion, plain powders. I don't think the loss in symmetry has to do with the product, otherwise I would notice the same operating with ACN, and I don't. As I stated before, symmetry is calculated via the USP tailing factor. 5% height, is that right? Maybe I am overinterpreting... but there's something different with respect to eluting with ACN alone.
Luca
-------------------------------------------------------------------------------------------------------
By Anonymous on Tuesday, May 18, 2004 - 05:33 pm:
i want to known whether the physicochemical properties of cloxacillin can affects its bioavailability when concorrent used with chloroquine
-------------------------------------------------------------------------------------------------------
By Anonymous on Wednesday, May 19, 2004 - 02:04 pm:
the column contamination theorists (and myself) are probably interested in the results where a corresponding guard column is used..
what if the contaminant isn't observed at your detection wavelength? has the sample been examined by eg. NMR, or have you tried examining the problem sample with a 'universal' detection technique like ELS or MS?
-------------------------------------------------------------------------------------------------------
By gregor scheipl on Tuesday, June 1, 2004 - 01:20 am:
Does anybody know a method how to analyse organic powders like CuPc, ZnPc, Alq3 etc. at HPLC, and a technique how to bring these powders in solution, because i have to analyse these powders prepurified!!!
Dear folks,
On developing HPLC methods for some Penicillin mixtures (mainly, Pen. G Procaine/Benzathine and Cloxacillin Benzathine), I found out that working at pH 3.5, phosphate buffer 0.04 M, with ACN/MeOH 3/2 as organic modifier, gives me very good results in terms of separation and peak simmetry. For some products, IPRs may be needed; anyway, the problem does not lie in the usage of IPRs, but in the fact that, sequence after sequence, I can notice a steady decrease in peak symmetry. I have "blamed" it on the MeOH - acid medium, which can cause deterioration of the column (I use either an Eclipse XDC C18 or a Purospher Star C18E), but I am wondering why in the European Pharmacopeia or in application notes such as Waters', acid buffers with MeOH as eluents are in use (to be precise, they use a trinary phase made up of water, acid P buffer of variable molarity from - 0.01 to 0.01 - and Methanol).
Shouldn't deterioration happen all the same or is this phenomenon enhaced by the mixture ACN/MeOH?
I know it can sound as an entangled issue, I hope there's someone out there who can give me a bit of advice, since I have not been able to find any reference on it.
Thanks
Luca
-------------------------------------------------------------------------------------------------------
By Anonymous on Wednesday, April 21, 2004 - 11:00 am:
And I would have a followup question to this: which is more damaging to an RP column, acid + water, or acid + methanol?
-------------------------------------------------------------------------------------------------------
By bill tindall on Wednesday, April 21, 2004 - 06:12 pm:
If this solvent mix is less than about 50% organic, then one can get a reliable measure of hydrogen ion activity by simply meausring the pH with a pH electrode calibrated in water. It is kind of complicated why this is so , but various factors cancel out under these condtions which enable a measurement to be made which is easy to interpret.
So, make up the solutions of interest and measure their pH and report back the results. Presuming there is an acid catalyzed column degredation reaction, then the pH measurement will tell you which solution is more acidic and hence faster rate. I am curious about the results you find.
I have done many separations in pH 2.2 phosphoric acid and methano/water for aromatic acids and columns last for years.
The answer to the 2nd question is more complicated and depends on the nature of the acid and whether we are talking about pure water vs pure methanol or some mixture. Perchloric acid in methanol is more acidic than perchloric acid in water, becasue methanol is a weaker base than water(hence protonated methanol is a stronger acid). But, things like dihydrogen phosphate are stronger bases in methanol than water, so phosphate buffers are more basic(or less acidic) in methanol than in water.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Thursday, April 22, 2004 - 12:20 am:
It would also be interesting to know whether extensive washing restores symmetry. My guess is that it should, as I also think that a pH of 3.5 would not attack your column noticeably.
On predicting reactivity, see also Bill´s contribution in the message board "pH/pKas". The dielectric property can dramatically influence the activation state of a reaction (an ionic activsation state is at a much lower energy in a high dielectric....). That means the kinetics of a reaction are also important, not only the thermodynamic states. But this complicated matter needs not concern you, your problem should lie elsewhere.
-------------------------------------------------------------------------------------------------------
By Luca Corte on Thursday, April 22, 2004 - 06:18 am:
Thanks Bill and thanks HW!
First time I need to reuse my P Buffer pH 3.5 ACN/MeOH 3/2 eluent, whose volume ratio is 65 Buffer : 35 ACN/MeOH (3/2), I will measure the pH of the final mixture and see what the real value is.
Knowing that with acidic/methanolic mixtures columns last for a long time relievs me. I am wondering if ACN does play a role in my problem.
I have to add that, in order not to leave the column in MeOH/ACN, when the sequence is over I wash it with H2O/ACN to remove the buffer, then with pure ACN. I know that it's safer to store your columns in pure ACN; moreover, analyses of products involving the use of ACN/MeOH are not done on routine in my Lab, and since I have dedicated columns for that, I prefer storing them in ACN.
The loss in symmetry is not dramatic from one sequence set to another, but you can see that if you plot a trend of the analyte TF over time.
In addition to that, if I use those columns to perform a chromatography of a product for which only ACN is needed as organic modifier, I can see a temporary slight loss in symmetry in it too. That one tends to revert over time, but I have not many data on this specific aspect.
I will post follows up after getting pH measurements.
By the way, I have been told that you can use MeOH to wash your columns when they look to be "spent" in some way, so that you can actually regenerate them, to some extent, but I have also been warned this procedure might be dangerous if done too often. Do you have any information on that? So, why shipping new columns in water/methanol instead of pure ACN?
Thanks
Luca
-------------------------------------------------------------------------------------------------------
By RH on Thursday, April 22, 2004 - 07:17 am:
I used mixtures of MeOH/ACN in buffer with different C18 phases and didn`t see adverse effects. One important difference between the two solvents, besides other physicochemical properties, is the protic nature of methanol.
As your peak symmetry changes slightly with analysis time this could be a problem of accumulation of matrix impurities.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Thursday, April 22, 2004 - 03:25 pm:
I agree with RH, that the most likely reason for the chage in peak shape are the components of the sample.
Bill, the pH of a phosphate buffer of 3.5 in water is going to drop, when you add organic solvent. In addition, how is going to help you, if the only information on column stability that you can get is the pH stability with the pH measured in water?
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Thursday, April 22, 2004 - 03:28 pm:
Geee -- the pH of the phosphate is going to INCREASE when you add an organic solvent ....
See Bill, what you made me do...
-------------------------------------------------------------------------------------------------------
By HW Mueller on Friday, April 23, 2004 - 01:29 am:
Luca, where do you measure Symmetry (Asymm...) at baseline or further up? Maybe you are overinterpreting?
Some manufacturers are recommending the more inert ACN for storage, because they apparently think in years... you can wash all you want with MeOH. Maybe there are some exceptions? I have not seen any. This high reactivity of MeOH toward C-18, etc., phases is apparently surviving from very old days??
-------------------------------------------------------------------------------------------------------
By Luca Corte on Friday, April 23, 2004 - 02:16 am:
Hi again!
I have seen that some of you are guessing about accumulation of matrix impurities on the column. I think you mean impurities from the sample... Well, I have very simple products to analyze which totally dissolve in the mobile phase, or, better, in the corresponding water/orgainc solvent mixture. Sometimes, in water alone. Moreover, I always filter samples through a 0.22 um filter prior to injection. Solvents and Milli-Q grade water are 0.22 um filtered as well.
So, according to me, therre shouldn't be any matter accumulation on the column. I should also notice pressure increase, which I don't.
Symmetry and TF are measured by means of the HPLC 1100 software (extended perfomance report). I look at the USP tailing and Symmetry, as you can see in the report format. They are complementary indexes, so that - ideally - if symmetry is 0.95, for instance, TF is 1.05. This way I can have info on column overloading and pH effects on the quantity of injected sample. Picky as I am, I am always striving to get ideal symmetry...
I must say that I am relieved to hear there's someone else out there using ACN/MeOH mxitures...
Luca
-------------------------------------------------------------------------------------------------------
By Alex on Friday, April 23, 2004 - 06:16 am:
Hi Luca,
We are using also MeOH/MeCN mixtures frequently. With water and no buffer. There is no reason that the mixture should be more agfressive then the solvents alone.
To your last posting: Do you isocratic elution? If so, more hydrophobic substances may stick to the column, even if the sample solution looks clean, is filtered.... Try to included a cleaning step with 100% organic from time to time.
-------------------------------------------------------------------------------------------------------
By HW Mueller on Friday, April 23, 2004 - 08:00 am:
Ok, so you use the default position for measuring Symmetry, which is probably above the baseline. If you let it be determined at baseline you will have problems (the start and end of a peak is somewhat difficult to find, discussed before).
Why don´t you tell us what your matrix comes from? It is dissolved material which will change your column characteristics.
-------------------------------------------------------------------------------------------------------
By A.Mouse on Friday, April 23, 2004 - 07:30 pm:
Contamination of the column and pressure increase do not always correlate.
-------------------------------------------------------------------------------------------------------
By Luca Corte on Monday, April 26, 2004 - 01:22 am:
Hi again!
First off, a couple of questions to Rolf and Alexander, who have been using ACN/MeOH mixtures.
When you are done with your analyses, do you store the column in ACN/MeOH alone? As I said before, I used to store the column in pure ACN, thinking that long-term storage in a MeOH mixture could possibly damage the hydrophobically-modified silica layer, but after reading your comments, it may not be the case. It could be that, removing ACN from the column and using the ACN/MeOH mixture upon ensuing sequences, may lead to a small and temporary loss in symmetry because of variation in the eluents chemical-physical properties. In short, the column may need longer equilibration times.
Anyway, I used to do final washing in ACN alone as I also wanted to get rid of Methanol in pump and injector assembly. I have just a single HPLC instrument and not all of my analyses require ACN/MeOH mixtures, most of them are done with Buffer/ACN. I wanted to have the machine ready to operate in the "standard" conditions, so to speak, and I was also afraid that MeOH, in contact with steel and plastic parts, could be more aggressive than ACN.
That leads me to column contamination... It could also be that small debris, like metal particles could come off from the tubing, or pump, and contaminate the column without affecting the internal pressure. This effect could me more visible with the Purospher Star column - which is said to be metal-ion free and capable of giving more straight peaks. So, if metal contamination occurs, it should be more evident than that observed with a "standard" column.
Maybe a precolumn would do the trick.
To answer HW, my sample matrix is made up of the sample alone, that is: pure Penicillin Benzathine, or Penicillin Procaine, or Cloxacillin Benzathine. I just analyze raw materials, without excipients of any sort.
To answer Alex, every time a sequence is over, I carry out a final washing with pure organic solvent.
Thanks again to you all, you have been very helpful! More comments are welcomed!
Luca
-------------------------------------------------------------------------------------------------------
By HW Mueller on Monday, April 26, 2004 - 03:17 am:
Pure is a relative term. "Normally" pure samples and mobile phases should give hundreds of injections without significant symmetry trends.
I don´t use piston pumps, but would expect the seals to be happier with MeOH than with ACN.
To expound on A.Mouse and my former statements: Generally, particulate matter does not change surface sharacteristics, they may cause plugging, whereas materials that change surface characteristics seldomly cause plugging (in my experience, only proteins have done both).
Thus, it seems either you don´t know something about the materials you use (or the software, like where the Sym is calculated) or you are overinterpreting.
-------------------------------------------------------------------------------------------------------
By Luca Corte on Monday, April 26, 2004 - 08:08 am:
Well, I think I have a sound knowledge of the materials I analyze, since they come from customers who have already done their analyses (so I can compare notes on results); they are products of synthesis and recrystallizazion, plain powders. I don't think the loss in symmetry has to do with the product, otherwise I would notice the same operating with ACN, and I don't. As I stated before, symmetry is calculated via the USP tailing factor. 5% height, is that right? Maybe I am overinterpreting... but there's something different with respect to eluting with ACN alone.
Luca
-------------------------------------------------------------------------------------------------------
By Anonymous on Tuesday, May 18, 2004 - 05:33 pm:
i want to known whether the physicochemical properties of cloxacillin can affects its bioavailability when concorrent used with chloroquine
-------------------------------------------------------------------------------------------------------
By Anonymous on Wednesday, May 19, 2004 - 02:04 pm:
the column contamination theorists (and myself) are probably interested in the results where a corresponding guard column is used..
what if the contaminant isn't observed at your detection wavelength? has the sample been examined by eg. NMR, or have you tried examining the problem sample with a 'universal' detection technique like ELS or MS?
-------------------------------------------------------------------------------------------------------
By gregor scheipl on Tuesday, June 1, 2004 - 01:20 am:
Does anybody know a method how to analyse organic powders like CuPc, ZnPc, Alq3 etc. at HPLC, and a technique how to bring these powders in solution, because i have to analyse these powders prepurified!!!