Volume decrease in NPT simulation with increase in temperature.

Hi, I am working on the MD simulations of a crystal structure where there I am trying to replicate the experimental values. I am simulating from 123K, the volume consistently increased with temperature till 233 but then it started to decrease till 253 and then increased. I am trying to solve it but couldn’t know what the route cause is. Any help on this would be much appreciated. Please find the input file setting and the graph for understanding.

In the graph, the red line is experimental
black line - npt
green - npt 6 component change(constant stress ensemble)

pastedImage.jpg

ScreenShot2019-10-29at7.21.45pm.jpg

input_npt (1).MDRun (1.8 KB)

Purely from a statistical physics point of view; if your volume decreases, you probably went into a more densely packed crystalline state. This can happen upon a temperature increase if the increased thermal energy allows your structure to anneal. Is your starting point defect-free or not? If you ramp temperature up and down repeatedly, do you trace out the same curve or does it change? I would definitely look at a simulation trajectory to see what is going on at the volume change.

pastedImage.jpg

pastedImage.jpg

ScreenShot2019-10-29at7.21.45pm.jpg

input_npt (1).MDRun (1.8 KB)

(Attachment ATT00001.txt is missing)

ATT00002.txt (171 Bytes)