- Posts: 7
ethylene oxide (EO) bead interaction with antifreeze
- aelahi4
-
Topic Author
- Offline
- Fresh Boarder
Less
More
4 years 6 months ago #8295
by aelahi4
ethylene oxide (EO) bead interaction with antifreeze was created by aelahi4
Hi.
In the newest itp file for ethylene oxide (EO) beads (martini_v2.0_PEO_PS_CNP.itp) based on the following paper, there are some confusing parameters for me.
link to file: www.cgmartini.nl/images/parameters/polym..._v2.0_PEO_PS_CNP.itp
reference paper: J. Phys. Chem. B 2018, 122, 7436−7449
in this file, the interaction between EO beads and water particles (P4), is different from their interaction with antifreeze particles (BP4) as below.. I believe this contradicts with the goal of definition antifreeze particles.
EO BP4 sigma=4.3 eps=3.15
EO P4 sigma=4.7 eps=3.5
I appreciate if you can explain the reason another interaction set of parameter is chosen for them? were they (EO-BP4) got tuned to reproduce a certain parameter?
Just want to know is there any physical interpretation behind that or not.
thank you so much for your time and response
In the newest itp file for ethylene oxide (EO) beads (martini_v2.0_PEO_PS_CNP.itp) based on the following paper, there are some confusing parameters for me.
link to file: www.cgmartini.nl/images/parameters/polym..._v2.0_PEO_PS_CNP.itp
reference paper: J. Phys. Chem. B 2018, 122, 7436−7449
in this file, the interaction between EO beads and water particles (P4), is different from their interaction with antifreeze particles (BP4) as below.. I believe this contradicts with the goal of definition antifreeze particles.
EO BP4 sigma=4.3 eps=3.15
EO P4 sigma=4.7 eps=3.5
I appreciate if you can explain the reason another interaction set of parameter is chosen for them? were they (EO-BP4) got tuned to reproduce a certain parameter?
Just want to know is there any physical interpretation behind that or not.
thank you so much for your time and response
Please Log in or Create an account to join the conversation.
- fgrunewald
-
- Offline
- Admin
Less
More
- Posts: 17
4 years 6 months ago #8297
by fgrunewald
Replied by fgrunewald on topic ethylene oxide (EO) bead interaction with antifreeze
Hi,
Your conclusion is indeed correct. It seems to be a mistake in the itp file and I'll fix it. So the interactions between BP4 should be the same as with P4.
However, generally speaking there is no need to add BP4 into a simulation with PEG. Unless you work at quite low temperatures or very low concentrations of PEG, it should prevent MARTINI water from freezing on it's own. Also be careful not to add a lot of BP4 with PEG, because the cavity cost in water might change significantly, which will change the expected behavior.
If you want to discuss the influence of BP4 onto your simulations with PEG in more detail, please feel comment here. I'm happy to give advise, if I can. Also thanks for reporting this issue!
Cheers,
Fabian
Your conclusion is indeed correct. It seems to be a mistake in the itp file and I'll fix it. So the interactions between BP4 should be the same as with P4.
However, generally speaking there is no need to add BP4 into a simulation with PEG. Unless you work at quite low temperatures or very low concentrations of PEG, it should prevent MARTINI water from freezing on it's own. Also be careful not to add a lot of BP4 with PEG, because the cavity cost in water might change significantly, which will change the expected behavior.
If you want to discuss the influence of BP4 onto your simulations with PEG in more detail, please feel comment here. I'm happy to give advise, if I can. Also thanks for reporting this issue!
Cheers,
Fabian
Please Log in or Create an account to join the conversation.
- aelahi4
-
Topic Author
- Offline
- Fresh Boarder
Less
More
- Posts: 7
4 years 6 months ago #8298
by aelahi4
Replied by aelahi4 on topic ethylene oxide (EO) bead interaction with antifreeze
thank you much Fabian for your response
well yes I still need to learn more about this issue. I appreciate your advises.
currently I am working on PEO4600, that is having specific volume of around 0.84 cm3/g in water at 1% mass fraction based on literature in room temperature. without antifreeze particles, using the model refereed in the message above, I came up with 25% of error. and the density of PEG-Water is so close to Water density. my impression was that there is no salvation of PEG in water, and water particles do not move to see PEG particles to interact with, so if I add BP4. the water particles will move and more probably can solvate PEG(is this intuition correct?). But by adding 10% of BP4 and using the above itp file, I got a very very close result. although this itp file includes some wrong values as we discussed, it's interesting that it is reproducing correct results as experiment. I am wondering why? Using same EO:P4 and EO:BP4, I am ending up with around 10% off. still acceptable but got worse
I want to make sure whether my understanding about BP4 particles is correct or not in overall
thank you so much for your consideration
well yes I still need to learn more about this issue. I appreciate your advises.
currently I am working on PEO4600, that is having specific volume of around 0.84 cm3/g in water at 1% mass fraction based on literature in room temperature. without antifreeze particles, using the model refereed in the message above, I came up with 25% of error. and the density of PEG-Water is so close to Water density. my impression was that there is no salvation of PEG in water, and water particles do not move to see PEG particles to interact with, so if I add BP4. the water particles will move and more probably can solvate PEG(is this intuition correct?). But by adding 10% of BP4 and using the above itp file, I got a very very close result. although this itp file includes some wrong values as we discussed, it's interesting that it is reproducing correct results as experiment. I am wondering why? Using same EO:P4 and EO:BP4, I am ending up with around 10% off. still acceptable but got worse
I want to make sure whether my understanding about BP4 particles is correct or not in overall
thank you so much for your consideration
Please Log in or Create an account to join the conversation.
- fgrunewald
-
- Offline
- Admin
Less
More
- Posts: 17
4 years 6 months ago #8299
by fgrunewald
Replied by fgrunewald on topic ethylene oxide (EO) bead interaction with antifreeze
Hi,
This is an interesting problem. Let's see:
1) BP4 should only be added to the simulation in order to stop water from freezing. If it is frozen or not, you can assess by computing the MSD as function of time. That should be linear. You can also check if the radius of gyration of PEG fluctuates. For PEO4600 as in 100 monomers (?) you should see quite some fluctuation (I'd say at least 0.5nm in the end-to-end distance). If instead you get a very staccato looking time series, it can mean 2 things: either you start from an initial coil that is too dense and it gets kinetically trapped or water freezes.
2) You need to make sure that you have the correct real masses, when computing the density and concentration. However, the PEG repeat unit mass is so close to the real values (i.e. 44 vs 45 amu per repeat unit) I don't expect this to make a large difference. But better safe than sorry.
3) MARTINI in general is not parametrized to reproduce densities accurately. So in that sense it would not be surprising, if there was a difference of some 10%. 25% seems a bit large though especially considering, how well the model reproduces small oligomer densities. On the other hand a value of 0.84 cm3/g in comparison to small molecule PEG water mixtures, seems rather low to me, especially since PEO4600 at RT is still quite far from the phase separation point. Can you send me the reference?
4) Finally there is the question how polydisperse PEG is in the experiment and how to best mimic the experimental setup. I think it would anyways be important to have a couple of PEG chains in water and perhaps of different length. How many chains do you use in your simulation?
Generally speaking I'd not fix any property by adding BP4 to the simulation. Also I would verify my system against scattering data rather than specific densities, if that's available. I find that a bit more trust-worthy. I hope this helps you a bit. But without knowing, where you want to go, it's hard to come up with anything more specific.
Cheers,
Fabian
This is an interesting problem. Let's see:
1) BP4 should only be added to the simulation in order to stop water from freezing. If it is frozen or not, you can assess by computing the MSD as function of time. That should be linear. You can also check if the radius of gyration of PEG fluctuates. For PEO4600 as in 100 monomers (?) you should see quite some fluctuation (I'd say at least 0.5nm in the end-to-end distance). If instead you get a very staccato looking time series, it can mean 2 things: either you start from an initial coil that is too dense and it gets kinetically trapped or water freezes.
2) You need to make sure that you have the correct real masses, when computing the density and concentration. However, the PEG repeat unit mass is so close to the real values (i.e. 44 vs 45 amu per repeat unit) I don't expect this to make a large difference. But better safe than sorry.
3) MARTINI in general is not parametrized to reproduce densities accurately. So in that sense it would not be surprising, if there was a difference of some 10%. 25% seems a bit large though especially considering, how well the model reproduces small oligomer densities. On the other hand a value of 0.84 cm3/g in comparison to small molecule PEG water mixtures, seems rather low to me, especially since PEO4600 at RT is still quite far from the phase separation point. Can you send me the reference?
4) Finally there is the question how polydisperse PEG is in the experiment and how to best mimic the experimental setup. I think it would anyways be important to have a couple of PEG chains in water and perhaps of different length. How many chains do you use in your simulation?
Generally speaking I'd not fix any property by adding BP4 to the simulation. Also I would verify my system against scattering data rather than specific densities, if that's available. I find that a bit more trust-worthy. I hope this helps you a bit. But without knowing, where you want to go, it's hard to come up with anything more specific.
Cheers,
Fabian
Please Log in or Create an account to join the conversation.
- aelahi4
-
Topic Author
- Offline
- Fresh Boarder
Less
More
- Posts: 7
4 years 6 months ago #8300
by aelahi4
Replied by aelahi4 on topic ethylene oxide (EO) bead interaction with antifreeze
Thank you Fabian for your detailed response
1- That is correct, but it is really affecting the density of the system as expected. it reduces the density of water around 10%, and PEG-WATER system for more than 20 %.
2-I used both, 45 and 44, although the terminals for PEG s using the stated model should be 31. density in both are same.
3-I did not understand this part by " 0.84 cm3/g in comparison to small molecule PEG water mixtures". At this T, specific volume of water is around 1. at the same temperature, it's less for PEG, due ordered water particles around it and high degree of solvation. what can be an estimate of volume for PEG and why?
this is the paper I'm using, look at table 1.
pubs.acs.org/doi/pdf/10.1021/jp036963c
4- I'm using 14 chains, equal to 2% of mass fraction in 12.5 nm box side size.
I'm trying to determine how transferable this model is over temperature, how good it can capture temperature effect on behavior of PEG in water. I did not find any paper reporting scattering result like radius of gyration. this is the only thing I found for this purpose over wide range of temperature
thank you so much
4-
1- That is correct, but it is really affecting the density of the system as expected. it reduces the density of water around 10%, and PEG-WATER system for more than 20 %.
2-I used both, 45 and 44, although the terminals for PEG s using the stated model should be 31. density in both are same.
3-I did not understand this part by " 0.84 cm3/g in comparison to small molecule PEG water mixtures". At this T, specific volume of water is around 1. at the same temperature, it's less for PEG, due ordered water particles around it and high degree of solvation. what can be an estimate of volume for PEG and why?
this is the paper I'm using, look at table 1.
pubs.acs.org/doi/pdf/10.1021/jp036963c
4- I'm using 14 chains, equal to 2% of mass fraction in 12.5 nm box side size.
I'm trying to determine how transferable this model is over temperature, how good it can capture temperature effect on behavior of PEG in water. I did not find any paper reporting scattering result like radius of gyration. this is the only thing I found for this purpose over wide range of temperature
thank you so much
4-
Please Log in or Create an account to join the conversation.
- fgrunewald
-
- Offline
- Admin
Less
More
- Posts: 17
4 years 6 months ago #8301
by fgrunewald
Replied by fgrunewald on topic ethylene oxide (EO) bead interaction with antifreeze
Hi,
I had some time to review the paper and data. I think that the data is not very suited to assess temperature transferability, because it puts demands on the density of water, PEG and their mixtures that are unrealistic for us to reproduce. In the low concentration regimes density deviations in the order 5%, which is very acceptable for MARTINI2, leads to deviations actually much larger than 25%. If you want instead you could look directly at the density and see if the trends are at least preserved.
In fact, instead I would recommend taking a look at this paper: pubs.acs.org/doi/abs/10.1021/acs.jctc.7b00560 . In this paper you will find radius of gyration data from simulation but with a reasonably good model. They also discuss experimental verification such as Kirkwood-Buff integrals, osmotic pressure and viral coefficients as function of temperature. Bottom line is: unless you go to long chains (Mn>300) or very high temperatures temperature effects are generally quite small. But this data should be better suitable to assess this.
I hope this helps you further.
Best,
Fabian
I had some time to review the paper and data. I think that the data is not very suited to assess temperature transferability, because it puts demands on the density of water, PEG and their mixtures that are unrealistic for us to reproduce. In the low concentration regimes density deviations in the order 5%, which is very acceptable for MARTINI2, leads to deviations actually much larger than 25%. If you want instead you could look directly at the density and see if the trends are at least preserved.
In fact, instead I would recommend taking a look at this paper: pubs.acs.org/doi/abs/10.1021/acs.jctc.7b00560 . In this paper you will find radius of gyration data from simulation but with a reasonably good model. They also discuss experimental verification such as Kirkwood-Buff integrals, osmotic pressure and viral coefficients as function of temperature. Bottom line is: unless you go to long chains (Mn>300) or very high temperatures temperature effects are generally quite small. But this data should be better suitable to assess this.
I hope this helps you further.
Best,
Fabian
Please Log in or Create an account to join the conversation.
- Pim
-
- Offline
- Expert Boarder
Less
More
- Posts: 105
4 years 6 months ago #8302
by Pim
Replied by Pim on topic ethylene oxide (EO) bead interaction with antifreeze
Slightly off-topic as it is unrelated to Martini and antifreeze, but I enjoyed this recent contribution to the whole PEG solubility mystery from the Stratingh institute newsletter.
If somebody would ask you, which compound is more soluble in water, polyethylene glycol ([-CH2CH2-O-]n (PEG) or polyoxymethylene ([-CH2-O-]n (POM), it is highly likely you would say POM and not PEG. But this is not the case, PEG is almost infinitely soluble, whereas POM is completely insoluble in water. Ensing and Woutersen (Univ. of Amsterdam), Hunger and Bonn (Max Planck Inst., Mainz, Germany) with six coworkers have now provided an explanation for this apparent mystery. Using femtosecond vibrational and GHz dielectric spectroscopy with complimentary ab initio calculations and molecular dynamic simulations, strong evidence was obtained that the dynamics of water molecules solvating PEG and POM is highly different and depending on their C/O composition. Previously it was commonly believed that steric effects were dominant, but now it was shown that the partial charge on the O atoms depends on the number of C atoms by which they are separated. So the different solubilities are mainly resulting from inductive effects, which determine the different partial charges on the O atoms in the two molecules. There is a larger partial charge on the O atoms of the PEG-like polyethers. These effects are enhanced by the anchoring of the water molecules hydrating the neighboring hydrophobic ether groups, which also explains the experimentally observed difference in water dynamics.
Ensing, B., Tiwari, A., Tros, M., Hunger, J., Domingos, S.R., Perez, C., Smits, G., Bonn, M., Bonn, D., Woutersen, S., Nature Comm. 2019, DOI 10.1038/s41467-019-10783-z
If somebody would ask you, which compound is more soluble in water, polyethylene glycol ([-CH2CH2-O-]n (PEG) or polyoxymethylene ([-CH2-O-]n (POM), it is highly likely you would say POM and not PEG. But this is not the case, PEG is almost infinitely soluble, whereas POM is completely insoluble in water. Ensing and Woutersen (Univ. of Amsterdam), Hunger and Bonn (Max Planck Inst., Mainz, Germany) with six coworkers have now provided an explanation for this apparent mystery. Using femtosecond vibrational and GHz dielectric spectroscopy with complimentary ab initio calculations and molecular dynamic simulations, strong evidence was obtained that the dynamics of water molecules solvating PEG and POM is highly different and depending on their C/O composition. Previously it was commonly believed that steric effects were dominant, but now it was shown that the partial charge on the O atoms depends on the number of C atoms by which they are separated. So the different solubilities are mainly resulting from inductive effects, which determine the different partial charges on the O atoms in the two molecules. There is a larger partial charge on the O atoms of the PEG-like polyethers. These effects are enhanced by the anchoring of the water molecules hydrating the neighboring hydrophobic ether groups, which also explains the experimentally observed difference in water dynamics.
Ensing, B., Tiwari, A., Tros, M., Hunger, J., Domingos, S.R., Perez, C., Smits, G., Bonn, M., Bonn, D., Woutersen, S., Nature Comm. 2019, DOI 10.1038/s41467-019-10783-z
Please Log in or Create an account to join the conversation.
Time to create page: 0.100 seconds