[gmx-users] Heat of vap

[David van der Spoel]

On 2011-04-04 01.32, Justin A. Lemkul wrote:
>
>
> Elisabeth wrote:
>>
>>
>>
>>
>> Elisabeth wrote:
>>
>> Dear David,
>>
>> I followed your instructions and calculated Heat of vaporization
>> of my alkane once with one molecule in gas phase (no cutoff) and
>> once with equivalent number of molecules as in liquid phase as
>> Justin suggested. Results are as follows:
>>
>>
>> To get heat of vaporization, you shouldn't be simulating just a
>> single molecule in the gas phase, it should be an equivalent number
>> of molecules as you have in the liquid phase.
>>
>> Hello David and Justin,
>>
>> My explanation was not clear. Below is the results for liquid phase
>> and for gas phase I tried two cases: one single molecule and the other
>> time for equivalent number of molecules as in liquid phase and thats
>> why results are very similar. ( However Justin says one single
>> molecule is not correct. I think when cutoffs is set to zero only
>> bonded terms are
>
> What is not correct is comparing the potential energy of a liquid system
> of many molecules with a "gas phase" of a single molecule. Whether or
> not that was something you did still is not entirely clear, but to be
> very clear, that's what I was saying is incorrect to do. DHvap is based
> on conversion of equivalent systems between liquid and gas.
>
>> treated and even where there are many particles in gas phase to get
>
> This is incorrect. Cutoffs of zero mean that all nonbonded interactions
> are calculated, they are not truncated.
>
>> energies per mole of molecules i.e g_energy -nmol XXX must be used so
>> values should be colse to a single molecules case.. please correct me!
>> Anyway results for gas phase are close and this is not the issue now).
>>
>
> You shouldn't need -nmol for any of this. Simply take the potential
> energy of the two systems (with equivalent numbers of molecules) and
> apply the formula I gave you several emails ago.
NOOOOOOOOOOOOOOOOOOOOOO


1 molecule in the gas phase  -> Epot(g) in your case 59.2 kJ/mol
N molecules in the liquid phase -> Epot(l) (since this is per mole you 
DO need the -nmol option) in you case 34.7 kJ/mol
DHvap = Epot(g) + kBT - Epot(l) = 59.2+2.5-34.7 = 27 kJ/mol which is 
quite close to hexane (28.9 kJ/mol).

>
> -Justin
>
>> Liquid phase:
>>
>> Energy Average Err.Est. RMSD Tot-Drift
>> -------------------------------------------------------------------------------
>>
>> LJ (SR) -27.3083 0.01 0.296591 -0.0389173 (kJ/mol)
>> Coulomb (SR) 6.00527 0.0074 0.122878 0.00576827 (kJ/mol)
>> Coul. recip. 5.59559 0.0032 0.0557413 0.00316957 (kJ/mol)
>> Potential *34.6779 * 0.025 1.03468 -0.11177 (kJ/mol)
>> Total Energy 86.4044 0.026 1.44353 -0.112587 (kJ/mol)
>>
>>
>>
>>
>> *one single molecule in gas phase*
>>
>>
>> Energy Average Err.Est. RMSD Tot-Drift
>> -------------------------------------------------------------------------------
>>
>> LJ (SR) -2.24473 0.073 1.292 0.342696 (kJ/mol)
>> Coulomb (SR) 11.5723 0.55 2.17577 -2.33224 (kJ/mol)
>> Potential * 59.244 * 0.94 10.9756 6.35631 (kJ/mol)
>> Total Energy 106.647 1 15.4828 6.78792 (kJ/mol)
>>
>> *equivalent number of molecules as in liquid* ( large box 20 nm)
>>
>> Statistics over 1000001 steps [ 0.0000 through 2000.0000 ps ], 4
>> data sets
>> All statistics are over 100001 points
>>
>> Energy Average Err.Est. RMSD Tot-Drift
>> -------------------------------------------------------------------------------
>>
>> LJ (SR) -2.16367 0.053 0.171542 0.374027 (kJ/mol)
>> Coulomb (SR) 11.2894 0.23 0.49105 -1.44437 (kJ/mol)
>> Potential * 63.2369 * 1.1 2.47211 7.69756 (kJ/mol)
>> Total Energy 114.337 1.1 2.65547 7.72258 (kJ/mol)
>>
>>
>> Since pbc is set to NO molecules leave the box and I dont know
>> if this all right. I hope the difference is acceptable...!
>>
>>
>> For "pbc = no" there is no box.
>>
>>
>> 0- I am going to do the same calculation but for some polymers
>> solvated in the alkane. For binary system do I need to look at
>> nonboded terms? and then run a simulation for a single polymer
>> in vacuum?
>>
>> Can you please provide me with a recipe for Delta Hvap of the
>> solute in a solvent?
>>
>>
>> The method for calculating heat of vaporization is not dependent
>> upon the contents of the system; it is a fundamental thermodynamic
>> definition. Heat of vaporization is not something that can be
>> calculated from a solute in a solvent. You can calculate DHvap for
>> a particular system, but not some subset of that system.
>>
>> Thanks Justin. I am interested in the energy required to vaporize
>> the solute in a particular solvent not the whole DHvap of the
>> mixture. do you think this can be achieved by calculating nonbonded
>> energies between solute and solvent? ( defining energy groups ..)
>>
>>
>>
>>
>> 1- If I want to look at nonboded interactions only, do I have to
>> add Coul. recip. to [ LJ (SR) + Coulomb (SR) ] ?
>>
>>
>> The PME-related terms contain both solute-solvent, solvent-solvent,
>> and potentially solute-solute terms (depending on the size and
>> nature of the solute), so trying to interpret this term in some
>> pairwise fashion is an exercise in futility.
>>
>> What you mean is when one uses PME interaction energies between
>> components can not be decomposed? So the energy groups I defined to
>> extract nonbonded energies are not giving correct values? Sofar I
>> have been defining energy groups to calculate nonbonded terms
>> between components _A-A A_B... I hope I have not been doing thing
>> wrongly!
>>
>>
>> Please help me out!
>>
>> Thanks,
>>
>


-- 
David van der Spoel, Ph.D., Professor of Biology
Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205.
spoel at xray.bmc.uu.se    http://folding.bmc.uu.se