-
- admin
- Site Admin
-
- Posts: 647
- Joined: Wed Aug 11, 2004 12:19 am
By yotsa pom on Wednesday, June 9, 2004 - 10:43 am:
I have trouble reading some papers about perfusion chromatography. What is the definition of this term and what's the benefit I will get from it?
I don't know that it's one of the modes of chromatography. Please answer,Thank you.
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Saturday, June 12, 2004 - 05:17 pm:
I'm not surprised that you are confused inasmuch as the original promotion of this technology was in the context of considerable marketing BS. A rough definition of perfusion chromatography is: separation under conditions where a significant component of eluent flow is through the particles as opposed to flowing around the particles as is normally the case. I haven't done the calculation myself but I once participated in a discussion session with Jack Kirkland and Fred Regnier on this topic. Jack stated that he had calculated the actual flow rate through the particles in the case of the original perfusion media (supplied by Perceptive Biosystems) as being around 1% that of the flow around the particles. While this might not sound like much but for molecules with slow diffusion characteristics, this can result in significant improvement of chromatographic performance when operating under gradient conditions. Actually, one could consider monolithic media to be the ultimate perfusion material as it typically exhibits even better chromatographic performance under high linear velocity conditions.
-------------------------------------------------------------------------------------------------------
By Anonymous on Sunday, June 13, 2004 - 10:22 am:
This is a good discussion because it gets at issues most chromatographers don't understand very well.
Chris: I agree with your basic defenition that perfusion particles have channels that allow mobile phase to pass through the particles. But I think the benefit of this is supposed to be the shorter diffusion paths, between the regions of bulk flow and regions of static flow, as compared with totally porous particles. Hence there is a smaller stationary phase mass-transfer term. You lost me when you spoke of the differences in flow rate throught and around the particles as being the critical factor. Could you expand on that.
I also do not understand why the textbooks generally say that you can use larger paticles with perfusion chromatography. It seems to me the basic principle should still apply that you want the interparticle channels to be as narrow as possible. And of course this is accomplished by using small diameter particles.
Any insight would be appreciated.
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Tuesday, June 15, 2004 - 12:09 pm:
My comments concerning to the importance of the relative flow rate through the particles compared to the flow rate around the particle are really just another way to look at the situation you describe. If the flow rate through the particle is zero (or approximately zero) as is commonly the case in conventional HPLC media, then transport in and out of the particle will be diffusion limited. As the flow rate through the particle increases due to the presence of larger pores, a packed bed begins to look more and more like the primary particles from which each particle of stationary phase is constructed (i.e. shorter diffusion path lengths and smaller effective particle diameters result in improved chromatographic performance and a flatter HETP versus flow rate profile). If you're interested in reading more about the theory, you might want to read the details in patent #5,019,270.
Regarding the point about recommending larger particle size media for perfusion chromatography, I think the information has been turned around backward. It's not that you need to use larger particle size materials in order to operate under perfusion conditions. Perfusion can theoretically be accomplished in any size media. The problem is that if the particle size is too small and the flow rate through the particle is minimal, the pressure drop will be too extreme for practical applications under "perfusion conditions". For example, assuming one had a column packed with 5 micron particles of perfusion media. The pressure would be pretty extreme if you wanted to use this material at 10 ml per minute in a 4.6 mm ID column. So, I think you can see that this leads to the general consensus that for practical reasons you are better off using larger particle diameters when working under "perfusion conditions".
-------------------------------------------------------------------------------------------------------
By Anonymous on Tuesday, June 15, 2004 - 06:09 pm:
Hey
Thanks for the patent reference. I will defenitely followup on that. If I understand what you've said thus far, the fact that the flow through the particles is ~1% of the flow around the particles is an indication that perfusion media doesn't really do what it claims (only if the mobile phase had easy access through the particles would there be any possibility to benefit from the perfusion design). Is that why you referred to it as BS? Have there been any publications showing success with this or has it been an abject failure. It seems like a solvable problem (make the pores bigger ??); but then I don't really know of where I speak.
Regarding the paticle diamter issue I will also try to capture what you've said. Is it that perfusion media are generally 'sold' on the concept of faster separations; and therefore larger particles are typically used to allow for higher flow rates. This seems a little odd to me since, in many cases, one would attempt to increase separation speed by simultaneously reducing the particle diameter and the column length. It's especially perplexing given that perfusion particles seem to be advertised for large molecule separations where one never uses high flow rates (or so I thought).
Any further response would be of interest and appreciated.
-------------------------------------------------------------------------------------------------------
By Anonymous on Saturday, June 19, 2004 - 07:02 am:
Nobody wants in here?
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Saturday, June 19, 2004 - 12:20 pm:
I don't think it's appropriate to be overly negative regarding perfusion chromatography (if I seemed so in my previous posts, it wasn't my intention). I've seen a number of demonstrations of the utility of this technique. If you check the Journal of Chromatography in the early nineties you should be able to find a number of examples of successful implementation of this technique.
It's just that the way it was originally pitched, it was subject to considerable exaggerations as to the magnitude of the improvement achieved by the technique.
In answer to your second question, the point is that until you get the intra particle flow rate to be higher than the diffusion velocity for a given solute, you can't expect to see any benefit from using such materials. Generally, this means operating at flow rates greater than 2 ml per minute with a 4 mm column and more often at flow rates greater than 5 ml per minute. The pressures observed in smaller particle media under these conditions will generally be rather near the operating pressure limit of the instrument. For this reason, larger particles are generally recommended.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Sunday, June 20, 2004 - 12:13 pm:
In perfusion chromatography, the access of slowly diffusing proteins to the interior of a particle is accelerated through a slow flow through the particle. This is done by using large pores that approach the size of the channels between the particles. It has been clearly demonstrated to work for unretained peaks only. For retained peaks, one is left with comparative chromatograms, but nothing that demonstrates the theory, or that the effects are absent on particles with a standard 50 to 100 nm pore size.
-------------------------------------------------------------------------------------------------------
By yotsa on Wednesday, June 23, 2004 - 02:27 am:
I'm thank all of you who interested and answer this message, I understand it better now. But something I can't really understand is its application. It's seems that this kind of chromatography is designed for a preparative scale of protein purifacation, but the publications that I've been through only mention about analytical purpose for this. Is anybody ever use this mode to do preparative scale of protein purification?
Another thing is in this perfusion condition, is it necessary to do it only in packed bed? Because right now I'm interest in Expanded bed chromatography which can combinded many steps in purification process in one step. Could EBA can run in perfusion condition?
-------------------------------------------------------------------------------------------------------
By Anonymous on Wednesday, June 23, 2004 - 08:37 am:
In order to flow through the pores, you need pressure. It won't work in EBC.
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Saturday, June 26, 2004 - 02:47 pm:
yotsa,
The lack of any published data on the preparative scale purification using perfusion chromatography is connected to another issue with the media as it was originally marketed. I was present at more than one oral presentation on the economic benefits of using perfusion for process scale applications. The argument made by the presenter in each case was that while the media was roughly 10 times more expensive than conventional gel based media commonly used for preparative separation of proteins but that this cost disadvantage could be offset by using a smaller volume of media and a proportionate increase in the number of preparative cycles. The claim was that perfusion is so much faster than a conventional separation that one can easily scale down the size of the column was out sacrificing throughput. On at least two occasions a member of the audience pointed out the fallacy of this proposal: each preparative cycle would still have to be independently validated, greatly increasing the manufacturing cost and eliminating any potential savings. Presumably the lack of significant numbers of published articles using this approach is an indication that most did not find the idea of dealing with the constraints of perfusion compelling for preparative applications.
-------------------------------------------------------------------------------------------------------
By James on Sunday, June 27, 2004 - 04:38 pm:
The comment made above by Uwe ('that the basic concept of perfusion particles has only been shown to work for non-retained species') propels me to ask another question. Shouldn't the same then by true of monolithic columns. As pointed out above, monolithic columns are basically an extension of the same concept.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Thursday, July 22, 2004 - 05:12 pm:
James,
The performance of some of the monolithic columns has been established for the most part with small molecules in isocratic chromatography under conditions of retention. Please see the publications by Tanaka et al, or probably best the review article by Tanaka et al in Analytical Chemistry. There is no problem with these assessments.
-------------------------------------------------------------------------------------------------------
By Dell Farnan on Tuesday, July 27, 2004 - 05:02 pm:
It has been a while since I looked at the perfusion issue, so
I am not up to date on any very recent work.....
Firstly, please distinguish the manner in which mass
transport (mx) mechanisms and rates are determined and
the manner in which a protein separation is normally
conducted. The MX experiments are typically done
isocratically or frontaly, where as protein elutions are
typically done using step changes or gradients of the
mobile phase. This distinction may through some light on
wether the mass transport phenomena will impact the day
to day usage of the phase.
Another observation is that while the mass transport within
the particles may be supplemented at higher flowrates, the
same perfusive column may still have a higher number of
thoeretical plates at a much lower flowrate. If you are doing
preparative work you may be more concerned about
concentration than time (within reason of course).
It is interesting in the above discussion that there is the
notion making it easier to use higher and higher flow rates
by using larger particle sizes. This may not necessarily help
achieve convection within the particle. It should be
remembered that in order for convection within the particle
to occur, there should be a sufficient pressure drop across
the particle to generate the intraparticle flow. The larger
the particle, by necessity the larger the nominal pore
diameter required for intraparticle convection. There was a
paper by in 1993 by Doug Frey, BioTech.Prog. 9(3) 273,
which dealt with convection within porous particles. If I
recall correctly, in that paper, it was concluded that the
nominal pore diameter had to be greater than about 1/20
of the particle size in order to achieve convection within the
particle. For a 4000A pore size that would suggest an
upper limit of about 8micron to expect signficant
intraparticle convection.
As to the question as to how appropriate is the
demonstration of the intraparticle convection with non-
retained species. In some data that I measured many years
ago I remember it was quite stunning how quickly the plate
count for protein columns dropped off when you start
retaining the compounds. Indeed there were so few plates
that it was impossible to determine the plate count with
any precision . While in a few occasions some protein
columns can be run with really shallow gradients to give
stunning results, e.g. the ProPac, most protein columns are
run with moderate/steep gradients where the gradient
focussing could have just as much impact on peak shape as
the mass transfer.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Wednesday, July 28, 2004 - 03:14 pm:
It is quite possible to measure plate counts and theoretical plate heights from gradients. However, it is not as simple as pressing a button in a software program. The principle on how to do this is described in my book on pages 76 to 79, and an example that demonstrates that this works for protein samples is given on page 79.
Except for my own work, I have never seen anybody yet to adopt the technique. The problem may very well be that one needs to have a rock-solid understanding of several aspects of the separation mechanism in order to do this properly. I have also done it on perfusion columns. Maybe one of these days I will get around to publish these results...
I have trouble reading some papers about perfusion chromatography. What is the definition of this term and what's the benefit I will get from it?
I don't know that it's one of the modes of chromatography. Please answer,Thank you.
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Saturday, June 12, 2004 - 05:17 pm:
I'm not surprised that you are confused inasmuch as the original promotion of this technology was in the context of considerable marketing BS. A rough definition of perfusion chromatography is: separation under conditions where a significant component of eluent flow is through the particles as opposed to flowing around the particles as is normally the case. I haven't done the calculation myself but I once participated in a discussion session with Jack Kirkland and Fred Regnier on this topic. Jack stated that he had calculated the actual flow rate through the particles in the case of the original perfusion media (supplied by Perceptive Biosystems) as being around 1% that of the flow around the particles. While this might not sound like much but for molecules with slow diffusion characteristics, this can result in significant improvement of chromatographic performance when operating under gradient conditions. Actually, one could consider monolithic media to be the ultimate perfusion material as it typically exhibits even better chromatographic performance under high linear velocity conditions.
-------------------------------------------------------------------------------------------------------
By Anonymous on Sunday, June 13, 2004 - 10:22 am:
This is a good discussion because it gets at issues most chromatographers don't understand very well.
Chris: I agree with your basic defenition that perfusion particles have channels that allow mobile phase to pass through the particles. But I think the benefit of this is supposed to be the shorter diffusion paths, between the regions of bulk flow and regions of static flow, as compared with totally porous particles. Hence there is a smaller stationary phase mass-transfer term. You lost me when you spoke of the differences in flow rate throught and around the particles as being the critical factor. Could you expand on that.
I also do not understand why the textbooks generally say that you can use larger paticles with perfusion chromatography. It seems to me the basic principle should still apply that you want the interparticle channels to be as narrow as possible. And of course this is accomplished by using small diameter particles.
Any insight would be appreciated.
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Tuesday, June 15, 2004 - 12:09 pm:
My comments concerning to the importance of the relative flow rate through the particles compared to the flow rate around the particle are really just another way to look at the situation you describe. If the flow rate through the particle is zero (or approximately zero) as is commonly the case in conventional HPLC media, then transport in and out of the particle will be diffusion limited. As the flow rate through the particle increases due to the presence of larger pores, a packed bed begins to look more and more like the primary particles from which each particle of stationary phase is constructed (i.e. shorter diffusion path lengths and smaller effective particle diameters result in improved chromatographic performance and a flatter HETP versus flow rate profile). If you're interested in reading more about the theory, you might want to read the details in patent #5,019,270.
Regarding the point about recommending larger particle size media for perfusion chromatography, I think the information has been turned around backward. It's not that you need to use larger particle size materials in order to operate under perfusion conditions. Perfusion can theoretically be accomplished in any size media. The problem is that if the particle size is too small and the flow rate through the particle is minimal, the pressure drop will be too extreme for practical applications under "perfusion conditions". For example, assuming one had a column packed with 5 micron particles of perfusion media. The pressure would be pretty extreme if you wanted to use this material at 10 ml per minute in a 4.6 mm ID column. So, I think you can see that this leads to the general consensus that for practical reasons you are better off using larger particle diameters when working under "perfusion conditions".
-------------------------------------------------------------------------------------------------------
By Anonymous on Tuesday, June 15, 2004 - 06:09 pm:
Hey
Thanks for the patent reference. I will defenitely followup on that. If I understand what you've said thus far, the fact that the flow through the particles is ~1% of the flow around the particles is an indication that perfusion media doesn't really do what it claims (only if the mobile phase had easy access through the particles would there be any possibility to benefit from the perfusion design). Is that why you referred to it as BS? Have there been any publications showing success with this or has it been an abject failure. It seems like a solvable problem (make the pores bigger ??); but then I don't really know of where I speak.
Regarding the paticle diamter issue I will also try to capture what you've said. Is it that perfusion media are generally 'sold' on the concept of faster separations; and therefore larger particles are typically used to allow for higher flow rates. This seems a little odd to me since, in many cases, one would attempt to increase separation speed by simultaneously reducing the particle diameter and the column length. It's especially perplexing given that perfusion particles seem to be advertised for large molecule separations where one never uses high flow rates (or so I thought).
Any further response would be of interest and appreciated.
-------------------------------------------------------------------------------------------------------
By Anonymous on Saturday, June 19, 2004 - 07:02 am:
Nobody wants in here?
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Saturday, June 19, 2004 - 12:20 pm:
I don't think it's appropriate to be overly negative regarding perfusion chromatography (if I seemed so in my previous posts, it wasn't my intention). I've seen a number of demonstrations of the utility of this technique. If you check the Journal of Chromatography in the early nineties you should be able to find a number of examples of successful implementation of this technique.
It's just that the way it was originally pitched, it was subject to considerable exaggerations as to the magnitude of the improvement achieved by the technique.
In answer to your second question, the point is that until you get the intra particle flow rate to be higher than the diffusion velocity for a given solute, you can't expect to see any benefit from using such materials. Generally, this means operating at flow rates greater than 2 ml per minute with a 4 mm column and more often at flow rates greater than 5 ml per minute. The pressures observed in smaller particle media under these conditions will generally be rather near the operating pressure limit of the instrument. For this reason, larger particles are generally recommended.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Sunday, June 20, 2004 - 12:13 pm:
In perfusion chromatography, the access of slowly diffusing proteins to the interior of a particle is accelerated through a slow flow through the particle. This is done by using large pores that approach the size of the channels between the particles. It has been clearly demonstrated to work for unretained peaks only. For retained peaks, one is left with comparative chromatograms, but nothing that demonstrates the theory, or that the effects are absent on particles with a standard 50 to 100 nm pore size.
-------------------------------------------------------------------------------------------------------
By yotsa on Wednesday, June 23, 2004 - 02:27 am:
I'm thank all of you who interested and answer this message, I understand it better now. But something I can't really understand is its application. It's seems that this kind of chromatography is designed for a preparative scale of protein purifacation, but the publications that I've been through only mention about analytical purpose for this. Is anybody ever use this mode to do preparative scale of protein purification?
Another thing is in this perfusion condition, is it necessary to do it only in packed bed? Because right now I'm interest in Expanded bed chromatography which can combinded many steps in purification process in one step. Could EBA can run in perfusion condition?
-------------------------------------------------------------------------------------------------------
By Anonymous on Wednesday, June 23, 2004 - 08:37 am:
In order to flow through the pores, you need pressure. It won't work in EBC.
-------------------------------------------------------------------------------------------------------
By Chris Pohl on Saturday, June 26, 2004 - 02:47 pm:
yotsa,
The lack of any published data on the preparative scale purification using perfusion chromatography is connected to another issue with the media as it was originally marketed. I was present at more than one oral presentation on the economic benefits of using perfusion for process scale applications. The argument made by the presenter in each case was that while the media was roughly 10 times more expensive than conventional gel based media commonly used for preparative separation of proteins but that this cost disadvantage could be offset by using a smaller volume of media and a proportionate increase in the number of preparative cycles. The claim was that perfusion is so much faster than a conventional separation that one can easily scale down the size of the column was out sacrificing throughput. On at least two occasions a member of the audience pointed out the fallacy of this proposal: each preparative cycle would still have to be independently validated, greatly increasing the manufacturing cost and eliminating any potential savings. Presumably the lack of significant numbers of published articles using this approach is an indication that most did not find the idea of dealing with the constraints of perfusion compelling for preparative applications.
-------------------------------------------------------------------------------------------------------
By James on Sunday, June 27, 2004 - 04:38 pm:
The comment made above by Uwe ('that the basic concept of perfusion particles has only been shown to work for non-retained species') propels me to ask another question. Shouldn't the same then by true of monolithic columns. As pointed out above, monolithic columns are basically an extension of the same concept.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Thursday, July 22, 2004 - 05:12 pm:
James,
The performance of some of the monolithic columns has been established for the most part with small molecules in isocratic chromatography under conditions of retention. Please see the publications by Tanaka et al, or probably best the review article by Tanaka et al in Analytical Chemistry. There is no problem with these assessments.
-------------------------------------------------------------------------------------------------------
By Dell Farnan on Tuesday, July 27, 2004 - 05:02 pm:
It has been a while since I looked at the perfusion issue, so
I am not up to date on any very recent work.....
Firstly, please distinguish the manner in which mass
transport (mx) mechanisms and rates are determined and
the manner in which a protein separation is normally
conducted. The MX experiments are typically done
isocratically or frontaly, where as protein elutions are
typically done using step changes or gradients of the
mobile phase. This distinction may through some light on
wether the mass transport phenomena will impact the day
to day usage of the phase.
Another observation is that while the mass transport within
the particles may be supplemented at higher flowrates, the
same perfusive column may still have a higher number of
thoeretical plates at a much lower flowrate. If you are doing
preparative work you may be more concerned about
concentration than time (within reason of course).
It is interesting in the above discussion that there is the
notion making it easier to use higher and higher flow rates
by using larger particle sizes. This may not necessarily help
achieve convection within the particle. It should be
remembered that in order for convection within the particle
to occur, there should be a sufficient pressure drop across
the particle to generate the intraparticle flow. The larger
the particle, by necessity the larger the nominal pore
diameter required for intraparticle convection. There was a
paper by in 1993 by Doug Frey, BioTech.Prog. 9(3) 273,
which dealt with convection within porous particles. If I
recall correctly, in that paper, it was concluded that the
nominal pore diameter had to be greater than about 1/20
of the particle size in order to achieve convection within the
particle. For a 4000A pore size that would suggest an
upper limit of about 8micron to expect signficant
intraparticle convection.
As to the question as to how appropriate is the
demonstration of the intraparticle convection with non-
retained species. In some data that I measured many years
ago I remember it was quite stunning how quickly the plate
count for protein columns dropped off when you start
retaining the compounds. Indeed there were so few plates
that it was impossible to determine the plate count with
any precision . While in a few occasions some protein
columns can be run with really shallow gradients to give
stunning results, e.g. the ProPac, most protein columns are
run with moderate/steep gradients where the gradient
focussing could have just as much impact on peak shape as
the mass transfer.
-------------------------------------------------------------------------------------------------------
By Uwe Neue on Wednesday, July 28, 2004 - 03:14 pm:
It is quite possible to measure plate counts and theoretical plate heights from gradients. However, it is not as simple as pressing a button in a software program. The principle on how to do this is described in my book on pages 76 to 79, and an example that demonstrates that this works for protein samples is given on page 79.
Except for my own work, I have never seen anybody yet to adopt the technique. The problem may very well be that one needs to have a rock-solid understanding of several aspects of the separation mechanism in order to do this properly. I have also done it on perfusion columns. Maybe one of these days I will get around to publish these results...
