[AMBER] radial distribution function tails for liquids using GAFF

From: conor parks <coparks2012.gmail.com>
Date: Fri, 10 Jun 2016 21:51:35 -0400

Hello everyone,

I am trying to prepare a metastable liquid for some very expensive
nucleation simulations I am about to perform with acetaminophen in water.
As such, I am trying to make sure EVERYTHING is good to go before
proceeding.

I have a general question about cut off lengths, radial distribution
function tails, and checking that a system has reached a liquid state when
using the GAFF force field. In my work, acetaminophen is being simulated
with GAFF parameters, and water is being represented with the TIP3P force
field. I first ran a 6.0 ns NVT equilibration run for a system consisting
of 2500 acetaminophen molecules, and 33934 tip3p water molecules, at
approximately 0.4 g/cm3 concentration (box length ~12 nm), at 380 K. I
checked the radial distribution function of the methyl carbon-methyl carbon
(c3-c3), and noticed longer tails than I expected. I thought, perhaps
naively, that the RDF should converge to 1 at 12.0 nm, as I believe this is
the cut off for the GAFF force field. However, after this 6.0 ns run, it
appears the RDF doesn't converge to 1.0 until approximately 3.0 nm. The
plot is attached large_simulation_NVT.pdf

I checked the pressure of the NVT run, and it had equilibrated at ~800 atm.
To check if the long tail was an artifact of equilibrating in the NVT
ensemble, without having relaxed the box length prior, I prepared a much
smaller simulation of 100 acetaminophen molecules in 1398 tip3p molecules,
again at ~0.4 g/cm3, with a box length of 4.0 nm. I ran an NPT ensemble
equilibration for 4.5 ns at 380 K and 1 atm. I then gathered RDF info for
25 ps, at 250 fs intervals. The plot of the n-n RDF is attached, and is
entitle small_simulation_NPT.pdf. As you can see, the RDF for this run
hasn't quite converged to 1.0 within the 1.2 nm cutoff.

So as a general question, should I even expect the RDF's of a liquid system
to converge to 1.0 by the 1.20 nm cutoff value, or are these long tails
possible for liquids using GAFF? My other possible interpretation, is that
the long tails are indicative of long range correlations, and thus
potentially partial solidification, although the temperature of the run
(380 K) is quite high for solidification to occur on these time scales.
Could someone comment if I am being overly anal, or if these long tails are
indicative of a possible misstep in my equilibration procedure?




Regards,

Conor Parks
B.S.E in Chemical Engineering, University of Michigan, 2012
PhD candidate in Chemical Engineering, Purdue University


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber

Received on Fri Jun 10 2016 - 19:00:02 PDT
Custom Search